Technical Report: River basin and Flood Modeling and Flood ...

1JUNE 2014

2

THE PROJECT TEAM

Designation Name AffiliationStudy Team Leader ENRICO C. PARINGIT, Dr. Eng. UP TCAGPFlood Hazard Expert ALFREDO MAHAR FRANCISO LAGMAY, PhD UP NIGSSub-Team Leader for Land Cover CZAR JAKIRI P. SARMIENTO, MSRS UP TCAGPHydrographic Survey Consultant LOUIE P BALICANTA, MAURP UP TCAGPResearch Associates RAQUEL FRANCISCO

JULIUS NOAH SEMPIOFRA ANGELICO VIRAY

UP TCAGP

© Copyright 2014. All rights reserved.

Any part of this document may be used and reproduced, provided proper acknowledgement is made.

Published by the Climate Change Commission under the CCC-UNDP-Australian Government Project - Project Climate Twin Phoenix

For inquiries, please contact:

Climate Change Commission2U LPLP Building, JP Laurel StMalacañang Compound, Manila City Email: [email protected]

3

ACKNOWLEDGEMENTS

We are very grateful to the following individuals and agencies for their untiring support and for sharing theirgenerous time during the conduct of our field surveys and other activities undertaken in this study:

1. Mindanao State University – Iligan Institute of Technology (MSU – IIT), especially Engr. Dan Mostrales, for providing the datasets.

2. Xavier University, especially Prof Dexter Lo, for providing some background information on their studies made regarding the

3. The Department of Public Works and Highways Regional Office X (DPWH-10), especially Engr. Aldrin Albano, for the exchange of information.

4. City Planning and Development Office (CPDO) of Cagayan de Oro City for the coordination work during the long-term flow measurements in Cagayan de Oro and Iponan Rivers.

5. City Planning and Development Office (CPDO) of Iigan City for the coordination work during the long-term flow measurements in Mandulog and Iligan Rivers.

6. Advanced Science and Technology Institute (ASTI) for the use of their High-performance Computing Facility.

7. DREAM staff who helped in the field surveys—Jeremy Acosta Regine Faelga, Joana Patricia Decilos, Cara Punay, JMSon Calalang and the rest of the field team

8. To the technical and administrative staff of the CCC-Project Climate Twin Phoenix led by Ms Susan Rachel G. Jose, Ms Julie Amoroso, Mr. Bayani Arcenas, Ms. Marie France Balawitan and Mr Ramon Enrico Punongbayan, Mr Jim Tangonan, the PCTP Liason Officer in Cagayan de Oro City for the assistance.

i

ii

TABLE OF CONTENTS

Chapter(1( 1

INTRODUCTION( 11.1#Background# 1

1.2#Scope#of#this#Study# 2

1.3#Expected#Outputs#and#Deliverables# 5

1.4#Professional#StafCing#and#Implementation# 6

1.5#Structure#of#this#report# 6

Chapter(2( 7

RIVER(BASIN(CHARACTERISTICS( 72.1#General#Characteristics# 7

2.2#SubJwatersheds#and#tributaries# 12

Chapter(3( 13

METHODOLOGY( 133.1#Research#for#existing#data# 13

3.2#Sites#Reconnaissance# 14

3.3#Watershed#RainfallJRunoff#Modeling# 143.3.1$Image$Classi.ication$and$Processing$ 143.3.2$Determination$of$Hydrological$Parameters$SCS=CN$Determination$ 153.3.3$Land$Cover$Change$Detection$ 153.3.4$Discharge$Modeling$using$HEC=HMS$ 163.3.5$HEC$HMS$Rainfall=Runoff$Hydrologic$Model$Components$ 16

3.4#DEM#Generation#from#LIDAR# 17

3.5#River#Measurements# 183.5.1$Cross$Section$and$Pro.ile$Measurement$ 183.5.1.2$River$Pro.ile$Survey$ 183.5.2$Data$Processing$ 213.5.3$Hydrometry$ 213.5.4$River$bathymetry$ 24

3.6#Spot#Mapping#of#Flooded#Areas# 24

3.7#Rainfall#Statistics# 25

ii

iii

3.8#Roughness#from#Land#Cover#Map# 25

3.9#Flood#Inundation#Modeling# 263.9.1$Brief$Model$Description$ 263.9.2$Pre=Requisite$Data$Files$ 273.9.3$Programming$Code$ 273.9.4$Cases$Supplied$ 283.9.5$Model$pre=processing$ 29

Chapter(4( 33RESULTS#OF#FIELD#SURVEY#DATA#PROCESSING#AND#ANALYSIS# 33

4.1#Gathered#CrossJsection#Survey#Points# 33

4.2#Generated#DEM#of#River#Basins# 36

4.3#Gathered#River#ProCile#Survey#Points# 38

4.4#Merged#Elevation#Data#Points#from#Field#Surveys#and#Other#Sources# 40

4.5#River#Bed#Characteristics# 40

Chapter(5( 46

RESULTS(OF(RIVER(BASIN(MODELING( 465.1#Modeling#domain# 46

5.2#HEC#HMS#Model#Preparation# 485.2.1$Model$Pre=processing$ 515.2.2$Land$Cover$Change$Projections$ 52

5.3#Actual#Rainfall#Events# 62

5.4#Hypothetical#Rainfall#Events# 64

5.5#RainfallJRunoff#Simulations#using#HECJHMS# 665.5.1$Simulated$Runoff$from$Cagayan$de$Oro$River$ 675.5.2$Simulated$Runoff$from$Iponan$River$ 735.5.3$Simulated$Runoff$from$Mandulog$River$ 785.5.4$Simulated$Runoff$from$Iligan$River$ 85

Chapter(6( 91

RESULTS(OF(FLOOD(MODEL(SIMULATIONS( 916.1#Simulated#Flood#Maps#and#Analysis# 91

6.1.1$Cagayan$de$Oro$Simulated$Flooding$ 916.1.2$Iponan$Simulated$Flooding$ 1106.1.3$Mandulog$River$Simulated$Flooding$ 1296.1.3.1$Mandulog$River$Flooding:$2013$Land$Cover$Condition$ 1296.1.3.2$Mandulog$River$Flooding:$2020$Land$Cover$Condition$ 135

iii

iv

6.1.3.3$Mandulog$River$Flooding:$2050$Land$Cover$Condition$ 1376.1.4$ligan$River:$Simulated$.looding$ 140

6.2#Combined#Flood#Hazard#Map# 1596.2.1$Cagayan$de$Oro$City$(Cagayan$and$Iponan$Rivers)$ 1596.2.2$Iligan$City(Iligan$and$Mandulog$River)$ 164

6.3#Comparison#of#Flood#Depths# 1736.5$Results$of$Flood$Inundation$Height$Validation$ 180

6.5#Simulated#Velocity#Maps# 1836.5.1$Cagayan$de$Oro$ 1836.5.2$Iponan$Simulated$Flooding$ 1966.5.3$Mandulog$Velocity$Maps$ 2096.5.4$Iligan$Velocity$Maps$ 216

Chapter(7( 229

DISCUSSIONS( 2297.1#Enhancements#introduced#in#the#Flood#Hazard#Maps# 229

7.2#Factors#Aggravating#the#Flooding#Problem# 2307.2.1$Changes$in$land$use/land$cover$conditions$ 2307.2.2$Sedimentation$and$Flooding$ 2317.2.3$Urban$Development$$Aspects$of$Flooding$ 233

Chapter(8( 235

CONCLUDING(REMARKS( 2358.1#Summary# 235

8.2#Recommendations# 235

8.3#Concluding#Remarks# 236

REFERENCES( 237

iv

v

LIST OF TABLES

Table&1.&Sub+watersheds&of&Cagayan&de&Oro&River&Basin& 12

Table&2.&Curve&Number&Values&Adapted&For&Rainfall+Runoff&Model.& 15

Table&3.&Land&Cover&Threshold&for&Change.& 15

Table&4.&&HEC+HMS&models&selected&to&&constitute&&the&&three&&components&of&&the&rainfall+runoff&model.& 16

Table&5.&Sample&Mandulog&Bridge&Hydrometry&Dataset&dated&April&13,&2013& 22

Table&6.&Land&Cover&Roughness&Values&Utilized& 26

Table&7.&Input&data&and&parameters&used&in&the&Tlood&modeling&using&Gerris&Flow&Solver& 27

Table&8.&Code&and&Respective&Descriptions.& 28

Table&10.&Summary&of&river&characteristics&based&on&the&results&of&the&Tield&surveys.& 41

Table&11.&Land&Cover&Threshold&for&Change.& 52

Table&12.&ClassiTication&of&antecedent&moisture&conditions&(AMC)&for&the&runoff&curve&number&method& 55

Table&13.&Curve&Number&(II)values&adapted&for&the&rainfall+++runoff&model&(Source:&NRCS,&1986).& 55

Table&14.&Rainfall&Intensity&Frequency&Duration&(RIDF)&data&generated&by&PAGASA&for&Cagayan&de&Oro.& 64

Table&15.&Rainfall+++Intensity&Frequency&Duration&(RIDF)&data&generated&by&PAGASA&for&Lumbia& 65

Table&16.&Rainfall+++Intensity&Frequency&Duration&(RIDF)&data&generated&by&PAGASA&for&MSU.& 65

Table&17.&Simulated&OutTlow&Volume&and&Peak&OutTlow&Rate&at&Cagayan&De&Oro&River&for&the&5+&Year&Rainfall&Event.& 71

Table&18.&Simulated&OutTlow&Volume&and&Peak&OutTlow&Rate&at&Cagayan&De&Oro&River&for&the&25+&Year&Rainfall&Event& 71



Table&21.&Simulated&outTlow&volume&and&peak&outTlow&rate&at&Iponan&River&for&the&5+year&rainfall&event.& 77

Table&22.&Simulated&outTlow&volume&and&peak&outTlow&Rate&at&Iponan&River&for&the&25+year&rainfall&event.& 77

Table&23.&Simulated&OutTlow&Volume&and&peak&outTlow&rate&at&Iponan&River&for&the&50+&Year&Rainfall&Event& 77

Table&24.&Simulated&outTlow&volume&and&peak&outTlow&rate&at&Iponan&River&for&the&100+year&rainfall&event& 77

Table&25.&Simulated&OutTlow&Volume&and&Peak&OutTlow&Rate&at&Mandulog&River&for&the&5+&Year&Rainfall&Event.& 84

Table&26.&Simulated&OutTlow&Volume&and&Peak&OutTlow&Rate&at&Mandulog&River&for&the&25+&Year&Rainfall&Event& 84

Table&27.&Simulated&OutTlow&Volume&and&Peak&OutTlow&Rate&at&Mandulog&River&for&the&50+&Year&Rainfall&Event& 84

Table&28.&Simulated&OutTlow&Volume&and&Peak&OutTlow&Rate&at&Mandulog&River&for&the&100+&Year&Rainfall&Event&84

Table&29.&Simulated&OutTlow&Volume&and&Peak&OutTlow&Rate&at&Iligan&River&for&the&5+&Year&Rainfall&Event& 89



Table&32.&Simulated&OutTlow&Volume&and&Peak&OutTlow&Rate&at&Iligan&for&the&100+&Year&Rainfall&Event& 90

v

LIST OF FIGURES

Figure 1. Location and de1inition of the cities of Cagayan de Oro and Iligan in Mindanao, Philippines. 2

Figure 2. Location of cross section, pro1iles and bathymetry along the Cagayan de Oro River as planned for the 1ield surveys 3

Figure 3. Location of cross section, pro1iles and bathymetry along the Iponan River as planned for the 1ield surveys 4

Figure 4. Location of cross section, pro1iles and bathymetry along the Mandulog River as planned for the 1ield surveys 4

Figure 5. Location of cross section, pro1iles and bathymetry along the Iligan River as planned for the 1ield surveys 5

Figure 6. Map of the Cagayan de Oro River Basin superimposed on the city/municipal and barangay boundaries. The numbers indicate sub-‐basins. 8

Figure 7. Map of the Iponan River Basin superimposed on the city/municipal and barangay boundaries. The numbers indicate sub-‐basins. 9

Figure 8. Map of the Mandulog River Basin superimposed on the city/municipal and barangay boundaries. The numbers indicate sub-‐basins. 10

Figure 9. Map of the Iligan River Basin superimposed on the city/municipal and barangay boundaries. The numbers indicate sub-‐basins. 11

Figure 10. Flowchart of overall project methodology. 13

Figure 11. Flow chart of Watershed Rainfall-‐Runoff Model development. 17

Figure 12. HI-‐TARGET VF echosounder with GPS setup in a rubber boat 19

Figure 13. Flow Discharge from the Mandulog Bridge Hydrometry Dataset Dated April 13, 2013. 22

Figure 14. Photographs showing the the team setting up the 1low meter used to collect water velocity 23

Figure 15. Flow meter with 1in and counterweight 24

Figure 16. Spot map of 1looded areas in Iligan City during Typhoon Sendong 25

Table 9. Look-‐up table used to convert the land-‐cover map to Manning’s n values 29

Figure 17. Map showing the location of boundary condition points where time series of water surface elevation and incoming 1low were assigned for every model simulation in the Cagayan de Oro River 30

Figure 18. Map showing the location of the boundary conditions points where time series of water surface elevation and incoming 1low were assigned for every model simulation in the Iponan River. 31

Figure 19. Map showing the location of the boundary conditions points where time series of water surface elevation and incoming 1low were assigned for every model simulation in the Mandulog River. 31

Figure 20. Map showing the location of the boundary conditions points where time series of water surface elevation and incoming 1low were assigned for every model simulation in the Iligan River. 32

Figure 21. Flow chart of river 1lood model development. 32

Figure 22. Map showing the actual river pro1ile, cross-‐section, and bathymetry data gathered from the 1ield survey in the Cagayan de Oro River. 33

Figure 23. Map showing the actual river pro1ile, cross-‐section, and bathymetry data gathered from the 1ield survey in the Iponan River. 34

Figure 24. Map showing the actual river pro1ile, cross-‐section, and bathymetry data gathered from the 1ield survey in the Mandulog River. 34

Figure 25. Map showing the actual pro1ile, cross-‐section, and bathymetry data gathered from the 35

vi

1ield survey in the Iligan River. 35

Figure 26. Map showing the DEMof the Cagayan de Oro River Basin and Flood Plains; and the 1lood model domain boundary 36

Figure 27. Map showing the DEM of the Iponan River Basin and Flood Plains; 37

and the 1lood model domain boundary 37

Figure 28. Map showing the DEM of the Mandulog River Basin and Flood Plains; 37


Figure 29. Map showing the DEMof the Iligan River Basin and Flood Plains; 38


Figure 30. Bed elevation pro1ile of Cagayan de Oro River. The coordinate of the 1irst point is (940436.917 N, 682364.126 E) 39

Figure 31. Bed elevation pro1ile of Iponan River. The coordinate of the 1irst point is (942107.489 N, 677567.738 E) 39

Figure 32. Bed elevation pro1ile of Mandulog River. The coordinate of the 1irst point is (912657.717 N, 637105.214 E) 40

Figure 33. Bed elevation pro1ile Iligan River. The pro1ile data was collected from downstream to upstream. 40

Figure 34. Map showing the narrowest and widest portions of the Cagayan de Oro River. 42

Figure 35. Map showing the narrowest and widest portions of the Iponan River. 43

Figure 36. Map showing the narrowest and widest portions of Mandulog River. 44

Figure 37. Map showing the narrowest and widest portions of the Iligan River. 45

Figure 38. Map showing the 1lood model domain (enclosed in red line) in the Cagayan de Oro River Basin while the subbasin divides are shown in violet line. 46

Figure 39. Map showing the 1lood model domain (enclosed in red line) in the Iponan River Basin while the subbasin divides are shown in violet line. 47

Figure 40. Map showing the 1lood model domain (enclosed in red line) in the Mandulog River Basin while the subbasin divides are shown in violet line. 47

Figure 41. Map showing the 1lood model domain (enclosed in red line) in the Iligan River Basin while the subbasin divides are shown in violet line. 48

Figure 42. The Cagayan de Oro River Basin model generated thru HEC-‐HMS 49

Figure 43. The Iponan River Basin model generated thru HEC-‐HMS 50

Figure 44. The Mandulog River Basin model generated thru HEC-‐HMS 50

Figure 45. The Iligan River Basin model generated thru HEC-‐HMS 51

Figure 46. Land cover distribution in the Cagayn de Oro River Basin 52

Figure 48. Land cover map of the Cagayan de Oro River Basin used for the estimation of the CB and watershed lag parameters of the rainfall-‐runoff model. 56

Figure 49. Land cover map of the Iponan River Basin used for the estimation of the CB and watershed lag parameters of the rainfall-‐runoff model. 56

Figure 50. Land cover map of the Mandulog River Basin used for the estimation of the CB and watershed lag parameters of the rainfall-‐runoff model. 57

Figure 51. Land cover map of the Iligan River Basin used for the estimation of the CB and watershed lag parameters of the rainfall-‐runoff model. 57

Figure 52. Soil map of the Cagayan de Oro River Basin Used for the Estimation of the CN Parameter 58

vii

Figure 53. Soil map of the Iponan River Basin Used for the Estimation of the CN Parameter 58

Figure 54. Soil map of the Mandulog River Basin Used for the Estimation of the CN Parameter 59

Figure 55. Soil map of the Iligan River Basin Used for the Estimation of the CN Parameter 59

Figure 56. Map showing the weighted CN values assigned to each watershed in the Cagayan de Oro River Basin(also called Sub-‐basin in HEC-‐HMS) 60

Figure 57. Map showing the weighted CN values assigned to each watershed in the Iponan River Basin(also called Sub-‐basin in HEC-‐HMS) 60

Figure 58. Map showing the weighted CN values assigned to each watershed in the Mandulog River Basin(also called Sub-‐basin in HEC-‐HMS) 61

Figure 59. Map showing the weighted CN values assigned to each watershed in the Iligan River Basin(also called Sub-‐basin in HEC-‐HMS) 61

Figure 60. Rainfall and hydrograph event recorded at Pelaez Bridge. These rainfall events were used to generate runoff hydrographs to calibrate the Cagayan de Oro River Basin 1lood model 62

Figure 61. Six-‐hourly rainfall event at San Simon Bridge. This rainfall event was used to generate runoff hydrographs to calibrate the Iponan River Basin 1lood model 62

Figure 62. Six-‐hourly rainfall event at Mandulog 2 Bridge. This rainfall event was used to generate runoff hydrographs to calibrate the Mandulog River Basin 1lood model 63

Figure 63. Six-‐hourly rainfall event recorded at Mandulog Bridge. This rainfall event was used to generate runoff hydrographs to calibrate the Iligan River Basin 1lood model 63

Figure 64. Interface of the Cagayan de Oro River Basin HEC HMS Rainfall-‐Runoff Model developed in this project 66

Figure 65. Cagayan de Oro watershed out1low hydrographs for the 5-‐Year Rain return period in 2013 land cover conditions. 67

Figure 66. Cagayan de Oro Watershed simulated out1low hydrographs for the 25-‐Year rain return period with 2013 land cover conditions. 67

Figure 67. Cagayan de Oro Watershed Out1low Hydrographs Generated from the Runoff Hydrographs Produced by the HEC HMS Model for the 50-‐ Year Rainfall Event in 2013 Land Cover. 68


Figure 69. Cagayan de Oro Watershed Out1low Hydrograph for the 5-‐ Year Rainfall Event in 2020 land cover and rainfall pattern from climate change projection. 68

Figure 70. Cagayan de Oro Watershed Out1low Hydrograph for the 25-‐ Year Rainfall Event in 2020 land cover and rainfall pattern from climate change projection. 69







viii

Figure 77. Iponan Watershed Out1low Hydrographs Generated from the Runoff Hydrographs Produced by the HEC HMS Model for the 5-‐ Year Rainfall Event in 2013 Land Cover. 73

Figure 78. Iponan watershed out1low hydrographs for the 25-‐year rainfall event in 2013 land cover. 73











Figure 89. Mandulog Watershed Out1low Hydrographs Generated from the Runoff Hydrographs Produced by the HEC HMS Model for the 5-‐ Year Rainfall Event in 2013 Land Cover. 78












Figure 101. Iligan Watershed Out1low Hydrographs Generated from the Runoff Hydrographs Produced by the HEC HMS Model for the 5-‐ Year Rainfall Event in 2013 Land Cover. 85



ix










Figure 113. Flood Map of the Cagayan de Oro 1lood plain showing the maximum 1lood extent and depths resulting from the 5 year rainfall event for the 2013 land cover overlain over the hillshaded topography. The roads, streets and the barangay names are also superimposed in the inundation map. 93

Figure 114. Estimated Extent of Flooding in Barangays in Cagayan de Oro for the 5-‐Year Rainfall Event, 2013 Land Cover 94

Figure 115. Flood Map of the Cagayan de Oro 1lood plain showing the maximum 1lood extent and depths resulting from the 25 year rainfall event for the 2013 land cover 95

Figure 116. Estimated Extent of Flooding in Barangays in Cagayan de Oro for the 25 Year Rainfall Event, 2013 Land Cover 96

Figure 117. Flood Map of the Cagayan de Oro 1lood plain showing the maximum 1lood extent and depths resulting from the 50 year rainfall event for the 2013 land cover condition. 97


Figure 119. Flood Map of the Cagayan de Oro 1lood plain showing the maximum 1lood extent and depths resulting from the 100 year rainfall event for the 2013 land cover condition. 99


Figure 121. Flood inundation map of the Cagayan de Oro 1lood plain showing the maximum 1lood extent and depths resulting from the 5 year rainfall event for the 2020 land cover. 101






x



Figure 129. Flood Map of the Iponan 1lood plain showing the maximum 1lood extent and depths resulting from the 5 year rainfall event for the 2013 land cover condition. 111

Figure 130. Estimated Extent of Flooding of Iponan River for the 5 Year Rainfall Event, 2013 Land Cover 112












Figure 142. Flood Map of the Iponan 1lood plain showing the maximum 1lood extent and depths resulting from the 25 year rainfall event for the 2050 land cover condition 126



Figure 145. Flood Map of the Mandulog 1lood plain showing the maximum 1lood extent and depths resulting from the 5 year rainfall event for the 2013 land cover condition 131

Figure 146. Estimated Extent of Flooding in Barangays in Mandulog for the 5 Year Rainfall Event, 2013 Land Cover 131





xi











Figure 161. Flood Map of the Iligan 1lood plain showing the maximum 1lood extent and depths resulting from the 5 year rainfall event for the 2013 land cover condition 141

Figure 162. Estimated Extent of Flooding in Barangays in Iligan for the 5 Year Rainfall Event, 2013 Land Cover 142












xii




Figure 177. Flood Map of the Cagayan de Oro and Iponan 1lood plains showing the maximum 1lood extent and depths resulting from the 5 year rainfall event for the 2013 land cover condition 159









Figure 186. Flood Map of the Iligan and Mandulog 1lood plains showing the maximum 1lood extent and depths resulting from the 5 year rainfall event for the 2013 land cover condition 164


Figure 188 Flood Map of the Iligan and Mandulog 1lood plains showing the maximum 1lood extent and depths resulting from the 100 year rainfall event for the 2013 land cover condition 166







Figure 195. Distribution of 1lood depths for various return periods for Cagayan de Oro River for present (2013). 173

Figure 196. Distribution of 1lood depths for various return periods for Iponan River for present (2013). 174

Figure 197. Distribution of 1lood depths from combined effects of Cagayan de Oro River and Iponan River for various return periods for Cagayan de Oro City in present condition (2013). 174

xiii

Figure 198. Distribution of 1lood depths for various return periods for Mandulog River for present (2013). 175

Figure 199. Distribution of 1lood depths for various return periods for Iligan River for present (2013). 175

Figure 200. Distribution of 1lood depths from combined effects of Mandulog River and Iligan River for various return periods for Iligan City in present condition (2013). 176

Figure 201. Comparison of of 1lood depth distribution from combined effects of Cagayan de Oro and Iponan Rivers for 5-‐year rainfall return period for 2013, 2020 and 2050 176

Figure 202. Comparison of 1lood depth distribution from the combined effects of Cagayan de Oro and Iponan Rivers for 25-‐year rainfall return periods for Cagayan de Oro City for present condition (2013) and from future scenario (2020 and 2050). 176

Figure 203. Comparison of 1lood depth distribution from combined effects of Cagayan de Oro and Iponan Rivers for 100-‐year rainfall return periods for Cagayan de Oro City for 2013, 2020, 2050 177

Figure 204. Comparison of 1lood depth distribution from combined effects of Mandulog and Iligan Rivers for 5-‐year rainfall return periods for Iligan City for 2013, 2020, and 2050 178

Figure 205. Comparison of 1lood depth distribution from combined effects of Mandulog and Iligan Rivers for 25-‐year rainfall return periods for Iligan City for 2013, 2020, and 2050 178

Figure 206. Comparison of of 1lood depth distribution from combined effects of Mandulog and Iligan Rivers for 100-‐year rainfall return periods for Iligan City for 2013, 2020, 2050 179

Figure 207. Correlation of 1lood heights during Sendong event for Cagayan de Oro River (n=37). 181

Figure 208. Correlation of 1lood heights during Sendong event for Iponan River (n=35). 181

Figure 209. Correlation of 1lood heights during Sendong event for Mandulog River (n=143). 182

Figure 210. Correlation of 1lood heights during Sendong event for Iligan River (n=35). 182

Figure 211. Simulated 1lood inundation and velocity map of Cagayan de Oro River for a 5-‐year rainfall return period under 2013 land cover conditions. 184












xiv

Figure 223. Simulated 1lood inundation and velocity map of Iponan River for a 5-‐year rainfall return period under 2013 land cover conditions. 197











Figure 234. Simulated 1lood inundation and velocity map of Iponan River for a 100-‐year rainfall return period under 2050 land cover conditions 208

Figure 235. Simulated 1lood inundation and velocity map of Mandulog River for a 5-‐year rainfall return period under 2013 land cover conditions. 209











xv


Figure 247. Simulated 1lood inundation and velocity map of Iligan River for a 5-‐year rainfall return period under 2013 land cover conditions. 217












Figure 259. Photo of Cagayan de Oro River merging (foreground) tributary taken from Bubunawan station, Bubunawan. Bukidnon upstream. 231

Figure 260. View of San Simon Bridge along Iponan River. 232

Figure 261. Paseo de Oro high-‐end shopping and hotel complex in front of Cagayan de Oro River. 233

Figure 262. Sky view of Paseo del Rio right beside Cagayan de Oro River 234

xvi

LIST OF ABBREVIATIONS

ADCP Acoustic Doppler Current ProfilerCADD Computer Aided Drafting and DesignCCC Climate Change Commission CdO Cagayan de OroCST Cross-section TeamDEM Digital Elevation ModelDOST Department of Science and TechnologyGIS Geographic Information System GNSS Global Navigation Satellite SystemGPS Global Positioning System HEC HMS Hydrologic Engineering Center Hydrologic Modeling System LiDAR Light Detection and RangingNAMRIA National Mapping and Resource Information Authority NRCS Natural Resources Conservation ServiceNIWA National Institute of Water and Atmospheric ResearchPAGASA Philippine Atmospheric, Geophysical and Astronomical Services Administration PRS92 Philippine Reference System of 1992PPCS/TM Philippine Plane Coordinate System/ Transverse MercatorRIDF Rainfall Intensity Duration FrequencySCS-CN Soil Conservation Services Curve NumberTGBM Tidal Gauge BenchmarkUNDP United Nations Development ProgrammeUP TCAGP UP Training Center for Applied Geodesy and PhotogrammetryUSDA United States Department of Agriculture

xvii

Chapter 1INTRODUCTION

1.1 Background

The Climate Change Commission (CCC) is implementing Project Climate Twin Phoenix with support from the United Nations Development Program and the Australian government. The component for the Typhoon Sendong-‐affected areas aims to assess the risks and vulnerabilities of the cities of Cagayan de Oro and Iligan to extreme weather events, including the potential impacts of climate change.

For this purpose, CCC partnered with the University of the Philippines Training Center for Applied Geodesy and Photogrammetry (UP-‐TCGAP) to undertake riverbasin and flood modeling study of four riversystems that transverse the two cities, namely, Cagayan de Oro, Mandulog, Iponan and Iligan (Figure 1). The study covered profile and cross-‐section surveys, inflow measurements, flood inundation modeling and, watershed and climate change impact analyses. It likewise incorporates projected rainfall generated by the Philippine Atmospheric, Geophysical, and Astronomical Services Administration (PAGASA).

The results will act as basis for priority mitigation actions like risk assessment, community based and managed early warning systems and, integrated contingency planning, and in the preparation or updating comprehensive land use and development plans to make them climate and disaster risk sensitive.

This comprehensive study complements Project on the Nationwide Operational Assessment of Hazards (Project NOAH) of the Department of Science and Technology (DOST), a project which current hazard maps using advanced technology in line with the disaster response and mitigation efforts of the Philippine Government. The flood modeling will enable the country’s agencies involved in disaster management to have a six-‐hour lead time to warn vulnerable communities against impending floods.

However, this study is different from DOST’s Project NOAH and its Disaster Risk Exposure Assessment and Mitigation (DREAM) component, because it includes the Iponan and Iligan River Basins, which both affected the cities of Cagayan de Oro and Iligan during the 2011 Sendong disaster. An updated (2013) land cover was also utilized to determine the impact of climate change, thus producing better results. Rainfall return periods were simulated based on the predicted changes to the latest land cover data.

RIVER BASIN AND FLOOD MODELING AND FLOOD HAZARD ASSESSMENT OF RIVERS IN THE CITIES OF CAGAYAN DE ORO AND ILIGAN

CCC-UNDP-Australian Government I PROJECT CLIMATE TWIN PHOENIX 1

Figure 1. Location and deHinition of the cities of Cagayan de Oro and Iligan in Mindanao, Philippines.

1.2 Scope of this Study

The preparation of the riverbasin and flood modeling study involved the following activities:1. Conduct measurements and surveys necessary for the 1lood modeling, including but not

limited to reference (horizontal and vertical control) surveys, cross section and pro1ile surveys, and in1low measurements;

2. Process data to convert all surveyed data and existing datasets necessary for the development of the 1lood inundation model;

3. Develop 1lood inundation model including their calibration and validation for each of the four rivers based on the datasets available from the surveys and measurements;

4. Generate 1lood inundation scenarios based on the 1lood inundation model on climate change projections to be provided by PAGASA;



5. Analyze the impact of watershed land cover and changes from rainfall characteristics as a result of climate change; and

6. Convert the inundation results in usable forms such as GIS-‐ready maps and statistics.

Figure 2. Location of cross section, proHiles and bathymetry along the Cagayan de Oro River as planned for the Hield surveys

* Updated during the actual 1ield surveys



Figure 3. Location of cross section, proHiles and bathymetry along the Iponan River as planned for the Hield surveys


Figure 4. Location of cross section, proHiles and bathymetry along the Mandulog River as planned for the Hield surveys




Figure 5. Location of cross section, proHiles and bathymetry along the Iligan River as planned for the Hield surveys


1.3 Expected Outputs and Deliverables

The outputs (including electronic files in CD or DVD) of this study consist of:

a. Tabulated coordinates of the established horizontal and vertical control points that includes WGS84 and PRS92 geographic coordinates, and PPCS/TM grid coordinates;

b. Cross section and pro1ile data and maps using appropriate coordinate system scale and digital 1ile format of choice (e.g., CADD dwg, dxf, shp, txt);



c. Data of in1low input measurements and computations for the four river basins in digital format;

d. Flood inundation model of the four river basins;e. Flood Inundation maps corresponding to 1loods with four return periods for the

present (2013), 2020 and 2050 land cover scenarios in hard copies and GIS shape1iles;

f. Watershed and climate change impact analyses for the four catchments; andg. Report/ documentation of the study.

1.4 Professional StafHing and Implementation

The study team was composed of river hydrographers, watershed modeler, flood modelers and engineers. The team leader was a licensed geodetic engineer and hydrologist who oversaw and managed the execution of all required surveys. The control, cross section and profile surveys and flow measurements were handled by subteam leaders. Survey works were handled by survey aides.

Field surveys were conducted from April-‐May and June-‐July, 2013, followed by data processing and flood modeling.

1.5 Structure of this report

This report consists of eight (8) chapters. Chapter 1 (this chapter) provides the project’s background, scope of work, expected outputs and professional staffing and implementation. Chapter 2 presents the riverbasin characteristics while chapter 4 presents the methodologies employed. Chapter 5 presents the results of the field surveys and data processing. Chapter 5 provides the results of the river basin modeling. Chapter 6 presents the development of the flood model and simulation results. Chapter 7 provides the discussion and chapter 8 provides the concluding remarks.



Chapter 2RIVER BASIN CHARACTERISTICS

2.1 General Characteristics

The Cagayan de Oro River Basin is one of the 18 major river basins in the Philippines. It has an estimated land area of 138 hectares or 1,521 sq. km. and has eight major rivers and along with their respective tributaries, that run across seven municipalities and cities. Those from Bukidnon include Talakag, Baungon, Libona and Pangantucan. Iligan City in Lanao del Norte, Municipality of Bubong in Lanao del Sur, ARMM and Cagayan de Oro are also included. It covers a total of 120 barangays in these areas (Figure 6). The basin starts at its upstream areas of the watersheds of Mt. Kalatungan and Kitanglad Mountain Ranges in Bukidnon. It flows towards Cagayan de Oro before discharging into the Macajalar Bay with a drainage area of 1,374.6 sq. km. The downstream end of the river is relatively flat and easily affected by tidal movements.

Other river basins of interest are Iponan, Mandulog and Iligan. The Iponan River Basin is located both in the Cagayan de Oro and Iligan cities. Its main tributary, the Iponan River, runs 60 km. from its head water in Iligan City before draining into Macajalar Bay, which is very much like the Cagayan de Oro River Basin (Figure 7). It has a drainage area of 407 sq. km. Meanwhile, the Mandulog River Basin runs a land area of 782 sq. km. The basin starts upstream from the Kalatungan range in Bukidnon, running 50 km., before draining into the Iligan Bay (Figure 8). The smallest river basin studied among the four is Iligan River (Figure 9) which runs about 19 km from its source, upstream in Lanao del Sur, towards the north of the Iligan Bay. It has a drainage area of about 243 sq. km.

The four river basins cover three provinces in Region 10 -‐ Misamis Oriental, Bukidnon and Lanao del Norte, and one province in the Autonomous Region in Muslim Mindanao (ARMM) – Lanao del Sur. The Cagayan de Oro River Basin can be found in all three provinces in Region 10, while the Iponan river basin can be found in two -‐ Misamis Oriental and Lanao del Norte. The Iponan river covers Opol, Cagayan de Oro City, El Salvador City, Manticao – all in Misamis Oriental, and Iligan City in Lanao del Norte. Meanwhile, the Mandulog and Iligan River Basins are both located within Misamis Oriental, Lanao del Norte and Lanao del Sur. The Mandulog river covers Iligan City; Tagoloan in Lanao de Norte; Kapai, Tagoloan II, Bubong in Lanao del Sur; and Manticao in Misamis Oriental. In terms of population (2010), Misamis Oriental has a population of 813,856; Bukidnon, 1,299,192; Lanao del Norte, 607,917; and Lanao del Sur, 933,000.



Figure 6. Map of the Cagayan de Oro River Basin superimposed on the city/municipal and barangay boundaries. The numbers indicate sub-‐basins.



Figure 7. Map of the Iponan River Basin superimposed on the city/municipal and barangay boundaries. The numbers indicate sub-‐basins.



Figure 8. Map of the Mandulog River Basin superimposed on the city/municipal and barangay boundaries. The numbers indicate sub-‐basins.



Figure 9. Map of the Iligan River Basin superimposed on the city/municipal and barangay boundaries. The numbers indicate sub-‐basins.



2.2 Sub-‐watersheds and tributaries

The Cagayan de Oro River Basin is composed of 56 sub-‐basins, 54 reaches and 55 junctions. Moreover, it has eight sub-‐watersheds, namely the (1) Bubunawan, Cagayan de Oro Rivers, (2) Tumalaong, Samalawan Rivers, (3) Tagiti River, (4) Kalawaig, Tutoban, Minontay Rivers, (5) Batang, Banongcol, Baylanan, Sangaya, Sagayan Rivers, (6) Tikalaan, Picalin Rivers, (7) Pigcotin, Bulaong Rivers, and (8) Munigui River. More information on the length of the rivers and their tributaries can be found in Table 1.

The Iponan River Basin has 26 sub-‐basins, 25 reaches and 26 junctions. It includes smaller tributaries such as the Domalokdok and Talakag creeks. The Mandulog River Basin, meanwhile, has 26 sub-‐basins, 24 reaches and 33 junctions. It conveys water through the Sardab and Saburan creeks, and the Rogongong, Digkila-‐an and Kapa-‐I rivers, among others, before pouring into the Iligan Bay. The Iligan River Basin has 31 sub-‐basins, 29 reaches and 30 junctions. It carries the Pugaan river, Malindawag creek and other smaller streams into the interior of the bay as well.

Table 1. Sub-‐watersheds of Cagayan de Oro River Basin

Sub-‐Watersheds LocationSub –

Watershed Area (sq. km.)

Length of Rivers (km)

Length of Tributaries

(km.)

Bubunawan, Cagayan de Oro Rivers

Cagayan de Oro, MisOr, Libona & Baungon, Bukidnon

26,875.89 71.93 194.27

Tumalaong, Samalawan Rivers

Baungon, Bukidnon 13,352.12 45.71 112.48

Tagiti River Baungon, Bukidnon 9,255.24 38.06 61.37

Kalawaig, Tutoban,Minontay Rivers

Baungon & Talakag,Bukidnon

19,382.66 84.78 109.30

Batang, Banongcol,Baylanan, Sangaya, Sagayan Rivers

Talakag, PangantucanBukidnon

31,598.07 109.86 109.86

Tikalaan, Picalin Rivers

Talakag, Bukidnon

7,527.20 60.57 16.98

Pigcotin, Bulaong Rivers

Talakag Buk; CDO;Bubong,LDS; Iligan,LDN

24,438.30 69.62 92.28

Munigi River Cagayan de Oro City 5,504.29 32.13 14.27

TOTALTOTAL 137,933.77 512.68 753.84



Chapter 3METHODOLOGY

The methods used in this study include research, field reconnaissance, watershed rainfall-‐runoff modeling, filed measurements, and flood modeling. The relationships of the different activities undertaken in the study can be seen in Figure 10. These major components of the methodology will be described in detail in the following sections.

Figure 10. Flowchart of overall project methodology.

Flood modeling including

calibration and validation

Surveys and measurements

Flood inundation maps from CC

rainfall scenarios

Conversion of PAGASA

Climate Change Rainfall Scenarios

Vulnerability Assessment

Flood Early warning system

Compilation of Topographic, Land Use and other Physical

Data

Online Hydromet Sensor Deployment and Installation

Data Conversion into GIS

Survey data Processing

Watershed runoff modeling, calibration/ validation

Evaluation of Land Use Impacts and Watershed Management Implications

Note: The boxes colored in grey are not part of the study but will be part of the overall framework in support of the disaster management activities for the study areas.

3.1 Research for existing data

Gathering of pertinent technical information and coordination with concerned agencies, including the National Mapping and Resource Information Authority (NAMRIA), were conducted. Reference data included locations and descriptions of horizontal controls (in WGS84 and PRS92 coordinate system) and vertical controls (elevation benchmarks) on or near the project area sites. These were gathered prior to the execution of the actual survey work.



Existing soil maps were obtained from the Bureau of Soils and Water Management. Elevation data (SRTM) used for the watershed modeling were obtained from the United State Geological Survey (USGS).

3.2 Sites Reconnaissance

Reconnaissance was done at least one day before the actual ground control survey for purposes of recovery of control point monuments on the ground. The most effective locations of control points (horizontal and vertical) were identified, as well as that of cross section lines, profile survey routes and tributaries for the river basins.

The reconnaissance activity also provided an opportunity to explore potential routes for doing the ground truth validation in relation to the generation of new land cover information that will be generated.

3.3 Watershed Rainfall-‐Runoff Modeling

A watershed study component, that considers the impact of climate change, was incorporated. Watershed runoff models were constructed by analyzing the relationship between land use, land cover, soil type, soil condition and watershed management conditions. These characteristics controlled the generation of peak flows values and arrival time, and total runoff in the considered rainfall–runoff models. Event flow due to different levels of rainfall conditions was also analyzed.

The model determined the effect of various land use and cover conditions on the runoff behavior of water at different stages and the severity of storm events. Results can be used to develop planning and policy interventions and overall flood management of the river basins.

The study utilized actual rainfall data sets collected from April to May, and June to July 2013 and hypothetical rainfall data sets based on the 24-‐hour Rainfall Intensity Duration Frequency (RIDF) curves provided by PAGASA. Return periods (also called “average recurrence intervals”) of 5, 25, 50 and 100 years were used to create hypothetical events with 12-‐hour duration and a maximum intensity on the 50% of the duration (i.e., on the 6th hour). In particular, changes in peak flow due to the predicted impact of projected rainfall were examined for the present scenario (2013), 2020 and 2050.

3.3.1 Image ClassiHication and Processing

Remotely-‐sensed imagery was classified to deduce and differentiate the land cover types found in the study areas. The different intensities of single electromagnetic waves, as reflected by objects present on the ground, are captured and stored as gray scale pixels in images. The degree of intensity is reflected in the brightness of specific features in the area of interest. In digital form, the pixels are given corresponding digital numbers, ranging from 0 to 255 (assuming an 8-‐bit quantization). In this study, LANDSAT 8 images were utilized. Red, Green and Blue wavelength band images were combined to create true color composite images, which were converted from raster to vector using Quantum GIS 1.8.0, a free and open source software.



3.3.2 Determination of Hydrological Parameters SCS-‐CN Determination

The Soil Conservation Services Curve Number (SCS-‐CN) method was developed by the Natural Resources Conservation Service (NRCS) of the United States Department of Agriculture (USDA). The CN model estimates precipitation excess as a function of cumulative precipitation, soil cover, land use, and antecedent moisture. The actual water retention of the land cover, watershed storage, actual direct runoff, total rainfall, and initial abstraction were taken into consideration. Table 2 shows the curve number values used as reference to determine watershed storage.

Table 2. Curve Number Values Adapted For Rainfall-‐Runoff Model.Land Cover AMC II Curve Number for Hydrologic Soil GroupAMC II Curve Number for Hydrologic Soil GroupAMC II Curve Number for Hydrologic Soil GroupAMC II Curve Number for Hydrologic Soil GroupLand Cover

A B C DBare Soil 77 86 91 94

Built-‐up Area 59 74 82 86Fallow Land 77 86 91 94Forestland 30 55 70 79Freshwater 98 98 98 98Grassland 39 61 74 80

Plantation/Shrubland 32 58 72 79*Values based from Schiariti lecture for Mercer County Soil Conservation District and Santillan (2008)*Values based from Schiariti lecture for Mercer County Soil Conservation District and Santillan (2008)*Values based from Schiariti lecture for Mercer County Soil Conservation District and Santillan (2008)*Values based from Schiariti lecture for Mercer County Soil Conservation District and Santillan (2008)*Values based from Schiariti lecture for Mercer County Soil Conservation District and Santillan (2008)

3.3.3 Land Cover Change Detection

Land cover conditions in the study area were highly dynamic, and change over time. To detect change, data on the chronological set of land cover in the area of interest were required, such as official land cover maps and satellite images. The rate of change in land cover rate was determined by obtaining the percentage change of specified land cover in an area and its change over the years. The following formula was used for this purpose:

(3.1)

In the event that questionable results are obtained, such as a great increase of forest land cover that realistically takes decades, the following thresholds for change were used as reference for arriving at the final land cover change.

Table 3. Land Cover Threshold for Change.Land Cover Type ThresholdThreshold

(-‐) (+)TOTAL Bare Soil -‐10.00% 10.00%Built-‐up Area 0.00% 10.00%Forestland -‐25.00% 5.00%Freshwater -‐2.00% 2.00%Grassland -‐10.00% 10.00%Plantation 0.00% 25.00%



3.3.4 Discharge Modeling using HEC-‐HMS

The Hydrologic Engineering Center-‐ Hydrologic Modeling System (HEC-‐HMS), is a free software developed by the United States Corps of Engineers-‐ Hydrologic Engineering Center (USACE-‐HMS). The HEC-‐HMS is constantly being updated and has been used extensively due to its well-‐documented performance and development.

The software is applicable to a wide range of geographic areas thus making it possible to solve a myriad of hydrologically-‐related problems. It is also well suited for urban watersheds when used to gauge the impact of land use changes and flooding. The program’s hydrographs can be utilized in studies on water availability, urban drainage, flow forecasting, future urbanization impact, flood damage reduction, floodplain regulation, and systems operation.

3.3.5 HEC HMS Rainfall-‐Runoff Hydrologic Model Components

The rainfall-‐runoff model was developed to simulate inflows to the floodplain being modeled. To properly simulate the basic hydrologic processes of runoff generation from rainfall, its transformation, and routing towards the outlet, the model is divided into three components. These are the (1) Infiltration Loss Model, (2) Direct Runoff Model, and (3) Channel Routing Model (Table 4).

The Infiltration Loss Model was employed to calculate the amount of rainfall that falls on the watershed by determining how much rainfall penetrates the surface and at what point water begins to run off. It is based on the Soil Conservation Services Curve Number (SCS-‐CN) method (USACE, 2000). Meanwhile, making use of the Unit Hydrograph, the Direct Runoff Model describes the situation prior to the infiltration of water into the watershed, and its position being just beneath the surface. Finally, the Channel Routing Model describes the runoff flow in the channels, and towards the main outlet of the rivers; it employs the Muskingum-‐Cunge method. Further details on the mathematical equations utilized in the models can be found in the HEC-‐HMS Technical Reference Manual.

Table 4. HEC-‐HMS models selected to constitute the three components of the rainfall-‐runoff model.

Component Model NameIn1iltration Loss US Soil Conservation Service–Curve Number (US SCS-‐CN)Direct runoff SCS Unit HydrographChannel routing Muskingum-‐Cunge Standard

The models presented in Table 4 were selected based on the following reasons:

1. They are well-‐established, well-‐documented and are readily available for use.2. There is seamless preparation of parameters and simulations due to the fact that HEC-‐HMS can combine the three models into a single system.

3. The GIS software can be utilized for the model preparation thus inputs required for the models can be provided within the time period allotted for the project. Inputs include the speci1ication of the 1low domain, boundary and initial conditions and parameter values.

4. The models are simple, which can be used despite minimal information about the watersheds. In addition, parameters can automatically be estimated or optimized, when the hydrologic data become available.



The Watershed Rainfall-‐Runoff Model was used to determine the effect of various land use and cover conditions on runoff behavior for the 5-‐, 25-‐, 50-‐ and 100-‐year rainfall return periods. In producing the outputs, the latest land cover data (2013) based on the satellite data analysis of the Cagayan de Oro, Iponan, Mandulog and Iligan watersheds was utilized. The predicted impact of projected changes in rainfall to peak flows were examined for the years 2013 (present), 2020, 2030 and 2050. The results for 2030 were utilized only for establishing the trend for the land cover changes but will not be employed in the flood model simulations. The HEC-‐HMS modeling system version 3.5 was used for this study. The flow chart of the watershed rainfall-‐runoff model development is shown in Figure 11.

Figure 11. Flow chart of Watershed Rainfall-‐Runoff Model development.

3.4 DEM Generation from LIDAR

Continuous fine scale elevation dataset is necessary to route the course of a river. Digital Elevation Models (DEM) were generated using aerial LiDAR surveys through the Disaster Risk and Exposure Assessment for Mitigation (DREAM) project. From 26-‐27 April 2013, two missions were launched using the Airborne LiDAR Terrain Mapper (ALTM™ Optech Inc.) Pegasus System. Data gathered were transferred to the data server and were processed. Orientation parameters were corrected while coordinates for each individual point cloud were computed and any misalignment was rectified using the LiDAR Mapping Suite (LMS™) (Optech Inc). Possible remaining misalignments between contiguous strips were checked, as well as the target density of the sites. Point clouds were classified into Ground, Low Vegetation, Medium Vegetation, High Vegetation and Buildings. These points were then rasterized to produce DEMs such as Digital Terrain Models (DTM) and Digital Surface Models (DSM) on a format readable by a Geographic Information Systems (GIS) software package. Minor errors were corrected manually to produce DEMs suitable for fine-‐scale flood modelling.

Ground truth points

Land cover data from NAMRIA and MSU-‐

Marawi

Post-‐1ield satellite image land cover classi1ication

Area calculation of each land cover type per subwatershed

Runoff and discharge output.

Soil type data SCS-‐CN determination

CN values plug-‐in to HEC-‐HMS



3.5 River Measurements

This section describes the methods used in obtaining direct measurements from the river needed for the flood modeling. Some key data about the geometry and the flow of the rivers were obtained through field survey.

3.5.1 Cross Section and ProHile Measurement

To tie the cross section and profile measurements in a uniform datum, control surveys were first conducted with the use of global positioning systems (GPS).

3.5.1.1 Horizontal and Vertical Control Survey

3.5.1.1.1 Establishment of Horizontal Control

Along the river, horizontal control points were established with the use of GPS receivers with accuracy standard of the third order, that is, a relative error of 1 in 10,000. Control points were obtained with the use GPS static surveys and were connected to NAMRIA reference stations with the same or higher order of accuracy.

Coordinates of the control points conformed with the Philippine Reference System of 1992 (PRS92) and were expressed in PRS92 and WGS84 and Philippine Plane Coordinate System/ Transverse Mercator (PPCS/TM) grid coordinate systems.

3.5.1.1.2 Establishment of Vertical Control

Vertical control points or elevation benchmarks were established with the use calibrated digital levels and used Third Order Vertical Control (Type 2) standards:

(3.2)

where D is the distance from the benchmark to the stations in kilometers. TGPS receivers were used when a known control point was far from the survey site. These control points were then referred to the mean sea level (MSL) referenced from the nearest NAMRIA benchmark. Established horizontal control points were also used as vertical control points if topography permits this. E stablished vertical control points were subjected to certification by NAMRIA.

3.5.1.2 River ProHile Survey

A combination of GPS-‐depth meter with vertical resolution of 10 cm or finer and horizontal resolution of 100 m or finer, were utilized in the river profile surveys. Ground control points were used for post-‐survey correction of survey tracks.

errorerror ≤12mm × D



3.5.1.2.1 Equipment Setup

The river profile surveys were done simultaneously with the cross section surveys. A separate team, the Profile Survey Team (PST) was deployed to conduct the profile measurements.

A HI-‐TARGET VF echosounder with Global Positioning System (GPS) mapping capability was utilized in the profile survey of the main tributary of the Cagayan de Oro, Iponan and Mandulog River Basins (Figure 12). For the Iligan River Basin, the OHMEX SonarMite echosounder with Global Positioning System (GPS) mapping capability was utilized. Both had a combination of GPS and depth meter with third order vertical and horizontal accuracies. The survey equipment was installed in a rubber boat or motorized banca.

The combined echo sounder-‐GPS setups recorded the location (longitude, latitude) and depth at half-‐meter horizontal distance (0.5 m) intervals with vertical (depth) resolution of 10 cm. A time stamp (time of recording) was also noted. The measured depth was between the water surface and the bottom (river bed) surface.

Figure 12. HI-‐TARGET VF echosounder with GPS setup in a rubber boat



Two approaches were employed to convert the measured depth to river bed elevation above/below the MSL. The first approach required the use of a time series of predicted tide levels. The depths measured by the echo sounder at a location (x, y) in the river during the time (t) was corrected for tidal effects and converted to river bed elevation using a linear relationship:

(3.3)

where RBE(x, y, t) is the elevation of the river bed at position (x, y) at time t. Tide(t) is the predicted tide depth (reckoned from Mean Lower Low Water, MLLW). Depth(x, y, t) is the depth recorded by the echo sounder at time t, and k is the height of the Mean Sea Level (MSL) above the MLLW datum. The value of k for Cagayan de Oro City (at the pier) was derived with the use of a Tidal Gauge Benchmark (TGBM) value. For Iligan City, meanwhile, a leveling survey was executed. These approaches were used to convert depths to RBE(x, y, t) for the downstream portions of the river basins connected to the sea, where the effect of tide is significant. Meanwhile, the second approach consisted of simultaneously measuring the water surface elevation while the profile survey progressed. The kinematic GPS technique using the Topcon GPS, for the Cagayan de Oro, Iponan and Mandulog River Basins, and Trimble GPS, for the Iligan River Basin, was employed. The processed data recorded by the Topcon and Trimble GPS (separate from the records by the echo sounder-‐GPS setup) produced datasets that contained a time stamp and its corresponding water surface elevation. The echo sounder records were then compared to the Topcon and Trimble GPS datasets to convert the measured depths to river bed surface elevation using the formula at time t:

where RBE(x, y, t) is the elevation of the river bed at location (x, y) at time t. WSE(x, y, t) is the recorded water surface elevation (above or below MSL) by the Topcon® or Trimble® GPS at time t, and Depth(t) is the depth recorded by the echo sounder at time t. The approach was used in the upstream portions of the river basins where the effects of tide are insignificant.

3.5.1.2.2 ProIile Survey Routes and Implementation

In all surveys, a handheld GPS was used for navigation. For the Cagayan de Oro, Iponan and Mandulog River Basins, the profile survey started upstream and then proceeded downstream towards the sea. Meanwhile, the profile survey for the Iligan river basin started downstream and then proceeded upstream until the last point of interest was reached.

The profile survey produced a number of elevation profile points which were utilized to derive a detailed bottom profile of the rivers. The survey routes followed a zigzag line; midpoints of each succeeding line were, at most ,100 meters apart. The profile points were used to derive additional cross sections of the river which were later utilized for the flood modeling. Additional bank points were also gathered to trace the outline of the river. This supplemented the crosssections that were surveyed by the Cross Section Team.

RBE x, y,t( ) = Tide t( )− Depth x, y,t( )− k

RBE x, y,t( ) =WSE t( )− Depth x, y,t( ) (3.4)



3.5.1.3 Cross Section Survey

A Cross Section Team (CST), composed of UPTCAGP Geodetic Engineers and locally hired field survey assistants, was formed to specifically perform the cross section surveys of the rivers.

Cross section surveys were done using calibrated instruments equipped with vertical distance measurement with at least one cm reading at 100 m intervals across the flood plain and 10 m intervals along the rivers to include low water channel, bank top and river top on both sides. Measurement points were referred from the established horizontal control points (in PRS92 and WGS84) and vertical control points (in MSL). The sectional view of main rivers and flood plains were derived from this.

The CST utilized the differential kinematic GPS survey technique in measuring the cross sections. Post-‐processing was done using the Topcon Tools ver. 7.5 software and Trimble Business Center ver. 2.4.

3.5.2 Data Processing

Data processing from ground control survey to cross section and profile survey were done using appropriate techniques and software. The GNSS observations were downloaded immediately after a day’s survey work to check for consistency and accuracy. Leveling works were downloaded and processed using a spreadsheet program. Plotting and processing of topographic data were done using a computer aided drafting and design (CADD) software.

Processing included the derivation of the stage-‐discharge curves or the rating curve for each river basin based on the discharge measurements. The rating curve was needed to convert the river stage to corresponding flow discharge.

In general, the output of the cross section and profile surveys are spreadsheet files containing the WGS84 geographic coordinates (Longitude, Latitude) and elevation above/below MSL of all the cross-‐section and profile points that were measured. These files were then imported in ArcView GIS 3.2 software (ESRI) and saved as vector files in shapefile format.

In addition, lateral inflow measurement datasets were also processed. The water cross sectional area A, was produced from the cross section data at the inflow stations. The area of each cross section segment was derived using the trapezoidal method. From the computed values of the cross section segment areas, A, were then derived. The velocities measured at each segment of the cross section were then used to calculate (using area-‐weighted averaging) inflow, Q. These Q values represent base flow conditions of the stream tributaries of the river in the flood modeling.

3.5.3 Hydrometry

Measurement of lateral inflows from the downstream tributaries of the rivers were conducted by measuring the height (nearest cm), width (nearest cm), and water velocity (meters per second) just before the river-‐stream junction. Measurement points were referred from the established horizontal control points (in PRS92) and vertical control points (in MSL).



Hydrometry was employed to collect accurate data for proper flood risk assessment. It is concerned with the measurement of flow discharges in the river basins. Data was collected over a period of two months in 2013: April to May and June to July, using several water level sensors and rain gauges. Rain gauges were used to measure precipitation over a certain period, while depth gauges and water level sensors were used to measure changes in water depth. Acoustic Doppler Current Profilers (ADCP), Velocity Meters and Flow Meters were used to measure current velocities.

Table 5. Sample Mandulog Bridge Hydrometry Dataset dated April 13, 2013Time Start

ReadingEnd

ReadingData Counts/sec Water

Velocity (ms)Water

Level (m)Discharge

m3s

9:00-‐9:10 AM 0 11655 11655 27.75 78 1.89 60.3653148

9:10-‐9:20 AM 11655 23339 11684 27.81904762 78 1.89 60.3653148

9:20-‐9:30 AM 23339 34922 11583 27.57857143 77 1.92 60.99237455

9:30-‐9:40 AM 34922 45381 10459 24.90238095 72 1.92 57.03183075

9:40-‐9:50 AM 45381 55825 10444 24.86666667 72 1.93 57.46849799

9:50-‐10:00AM 55825 64952 9127 21.73095238 62 1.95 50.23880017

3:00-‐3:10 PM 0 10406 10406 24.77619048 72 1.8 51.79182392

3:10-‐3:20 PM 10406 21559 11153 26.5547619 77 1.78 54.45449566

3:20-‐3:30 PM 21559 32402 10843 25.81666667 73 1.79 52.06842275

3:30-‐3:40 PM 32402 43253 10851 25.83571429 73 1.78 51.62569069

3:40-‐3:50 PM 43253 55257 12004 28.58095238 78 1.78 55.16169691

3:50-‐4:00 PM 55257 66165 10908 25.97142857 72 1.77 50.48182222

The sample dataset was visualized into its respective graph, which can be seen in Figure 13.

Figure 13. Flow Discharge from the Mandulog Bridge Hydrometry Dataset Dated April 13, 2013.

Discharge m3s



All ADCP data was collated and processed with the use of Microsoft Excel. The water velocity, water level and discharge values were determined from the set data. A sample data table from the Mandulog Bridge is found in Table 5.

Water level data was collected with the use of the flow meter placed in selected bridges along the river. Measurements were taken twice daily, six hours apart and for every set, a 10-‐minute interval was allotted for an hour. When the rain gauge-‐measured rainfall of two mm in April-‐May and 5 mm for June-‐July and consequently triggered two consecutive alerts, a six-‐hour reading with a 7-‐minute interval was done instead. Figure 14 show photographs of the flow meter set-‐up.

Figure 14. Photographs showing the the team setting up the Hlow meter used to collect water velocity

(a) tying up the 1low meter to the rope and counterweight; (b) lowering the 1low meter; (c) and (d) retrieving the 1low meter

(a) (b)

(c) (d)



The flow meter system initially involved only an assembly of a flow meter, pulley and counterweight. For more accurate results, a fin was later attached to the flow meter to prevent it from veering away from the direction of the flow. On the other hand, a counterweight was used to maintain the depth of the flow meter and to keep it from drifting during very high flow conditions (Figure 15).

Figure 15. Flow meter with Hin and counterweight

3.5.4 River bathymetry

Bathymetry data points were collected with the use of the HI-‐TARGET VF echosounder-‐GPS setup, OHMEX SonarMite echosounder-‐GPS setup, and using kinematic DGPS approach. The latter was utilized in the shallow portions of the rivers.

3.6 Spot Mapping of Flooded Areas

Field data on previous flooding events, particularly typhoon Pablo (December 2012) and typhoon Sendong (December 2011), were gathered to validate the flood model. Maps of actual flood extents were obtained from the City Planning and Development Office (CPDO). Features located within the flood zones and a few outside it were targeted for surveys. Survey questionnaires were used to gather data on the extent of flood in the area during Typhoons Sendong and Pablo. Local residents were asked the peak flood height and their time of occurrence as well. Flood marks in the homes were documented with the use of measuring tape while the exact location of the establishments/features were taken thru a handheld GPS. Figure 16 shows the spot map for Iligan.



Figure 16. Spot map of Hlooded areas in Iligan City during Typhoon Sendong

Source: Iligan City Planning and Development Office

3.7 Rainfall Statistics

PAGASA provided Rainfall Intensity Duration Frequency (RIDF) data, which is the amount of rainfall over return periods of 5, 25, 50 and 100 years. Original values in millimeters (mm) were converted to millimeters per hour (mm/hr) for data processing and modeling purposes. Data of shorter durations with peak levels of rainfall intensity were requested from PAGASA so as to see the result of intense rainfall during a short period of time. It was also done to assess the possibility of flooding from intense rainfall alone, especially in areas which do not directly surround the river basins.

3.8 Roughness from Land Cover Map

A land cover map was used to derive the roughness value of different areas of interest in the cities of Cagayan de Oro and Iligan. The map contains information on the biophysical cover of the land, all of which affect flooding. Manning’s coefficients n, or flow resistance coefficients, were utilized as the roughness values is dependent on land cover. A look-‐up table based on the HEC-‐RAS Hydraulic Reference Manual (Brunner, 2010b) was used. The values used are in Table 6, with unclassified ground features and other biophysical cover being assigned with a value of 0.01. The roughness values were assigned to each type of land cover for data processing purposes, enabling the duplication of the nature of water’s flow on different surfaces.



Table 6. Land Cover Roughness Values UtilizedLandcover type Roughness Value

Unclassi1ied 0.01Clouds 0.01Shadows 0.01Plantation (shrubland) 0.40Forest 0.80Seawater 0.01Fallowland 0.05Bare Soil 0.06Grassland 0.25Fresh Water 0.01Thin Clouds 0.01Built-‐up Area 0.01*Adapted from HEC-‐HMS Manual based on Barnes (1967) in Brunner (2010)*Adapted from HEC-‐HMS Manual based on Barnes (1967) in Brunner (2010)

3.9 Flood Inundation Modeling

3.9.1 Brief Model Description

The flood models for the four river basins were developed with the use of the Gerris Flow Solver, a free and continually updated programming language by S. Poppinet (2009). Its development is supported by the National Institute of Water and Atmospheric Research (NIWA) and Institut Jean le Rond d'Alembertthe. The 0.8.0 version was utilized for this study. Gerris Flow Solver is a programming language that was designed as a solution of partial differential equations which describe fluid flow. It has a number of features which include: solving time-‐dependent incompressible variable-‐density Euler, Stokes or Navier-‐Stokes equations, solving linear and non-‐linear shallow-‐water equations, adaptive mesh refinement, automatic mesh generation in complex geometries, unlimited number of advected/diffused passive tracers, dynamic load-‐balancing, parallel offline visualization, volume of fluid advection scheme for interfacial flows, accurate surface tension models and multiphase electrohydrodynamics.

The open source code is usable for 2D and 3D simulations. An advantage of using this is its ability to fully utilize the quadtree representation, which dynamically adapts to evolving flow features. Modeling was used to determine the maximum flood extent and inundation levels caused by rainfall events. These events were based on both actual and modeled rainfall data, the latter being based on the 24-‐hour RIDF curves provided by PAGASA. Other datasets utilized to develop the model were from the cross-‐section and profile surveys, and inflow measurements.



3.9.2 Pre-‐Requisite Data Files

The Gerris Flow Solver (GFS) system utilizes (i) Digital Elevation Model (DEM), along with the river channel position specified within it, (ii) surface friction parameter, Manning’s roughness coefficient n, to duplicate natural flow over various land surfaces, and (iii) hydrometric data, such as discharge and rainfall data, to simulate real and predicted scenario. Pre-‐requisite files were utilized to define the flood model of the river basins as shown in Table 7.

Digital Elevation Model files (dem.ascii) consisted of 2D raster array of ground elevations of the river basins and areas around them. Roughness files contained information on spatially variable floodplain friction or roughness (based on the Manning’s roughness). The flow files contained information on the discharge and other data on the flow of the river, given the evolution of time. The rain files defined the amount of rainfall over a period of time in the river basins.

Table 7. Input data and parameters used in the Hlood modeling using Gerris Flow Solver

Item name Description Value in the Project

DEM Digital Elevation Model file. Not readily read and must be converted to Gerris terrain database format through the xyz2kdt command and the awk scripting language (asc2xyz.awk)

dem.ascii

Roughness Name of file containing a grid of floodplain n values in ARC ascii raster format to allow spatially variable floodplain friction. This should have the same dimensions and resolution as the DEM file. Like the DEM file, this is converted as well.

n.ascii.

Flow File defining data on the evolution in time of the river 1low at a certain boundary. This is a Cartesian Grid Data where data is de1ined in a Cartesian grid and can be used with 1 to 4 space/time dimensions. For the project only 1 temporal dimension.

flow.cgd

Rain File de1ining data on the evolution in time of rainfall. This is also a Cartesian Grid Data type.

rain.cgd

3.9.3 Programming Code

Simulation files and classes of objects were incorporated in the programming code of the Gerris Flow Solver (GFS). In general, the same code was utilized for all simulations, with only a few minor changes, such as the name of files and the return period employed. In Table 8 are the simulation files and classes of objects utilized and the respective description of the simulation files.



http://gfs.sourceforge.net/wiki/index.php/Xyz2kdt

http://gfs.sourceforge.net/wiki/index.php/Xyz2kdt

http://www.wikipedia.org/wiki/awk

http://www.wikipedia.org/wiki/awk

Table 8. Code and Respective Descriptions.Code DeHinition

De1ine DRY 1e-‐3 “De1ine” is used to delineate text macros in parameter 1iles

DRY value for the water level parameter is employed rather than zero to distinguish "wet" from "dry" cells. A text macro is de1ined and used where the depth of dry cells is needed.

De1ine MAXLEVEL 10 MAXLEVEL 10 re1ines the initial mesh and creates a regular Cartesian grid with 210=1024 cells in each dimension. A higher level of re1inement produces better resolution. This increases computation time but produces a better visual output.

GfsRiver Utilized to solve the shallow water implementation Saint-‐Venant equations.

GfsBox The simulation domain is a single "GfsBox", which by default is a square centered on (0,0) and of size unity i.e. the coordinates of the simulation domain are within [-‐0.5:0.5] x [-‐0.5:0.5]

GfsGEdge Links the boxes created by the GfsBox

GModule De1ines objects used to perform cartographic projections within Gerris.

OutputSimulation Writes a description of the current state of the simulation that contains both standard simulation parameters, layout of the cell hierarchy and associated variable values. It serves as checkpoint for Gerris (for unexpected shutdown of the machine). Outputs here are used to determine the maximum velocity and water level at each pixel.

GfsEventScript Executes a shell-‐script at given intervals to save memory space. In this study, it creates a zip 1ile for all ouputs.

OutputPPM Writes a color image of the given scalar 1ield in Portable PixMap (PPM) format. It then turns it into an mpeg 1ile (animation of the event; 1 for water level and 1 for velocity).

OutputGRD Writes a raster (or gridded) dataset using the ESRI grid ASCII format.

Timeorder By default, GfsRiver uses a second-‐order time integration scheme. In this study, the timestep required for stability, with respect to maximum resolution used, is around 0.2 seconds.

Boundary De1ines boundary conditions on boundaries of a GfsBox.

BcDirichlet Imposes a Dirichlet boundary condition, i.e., the value of the variable on the boundary (entry of discharge).

BcSubcritical Imposes tidal boundary condition

BoundaryOut1low Allows "free" out1low (and in1low) on a given boundary.

3.9.4 Cases Supplied

The developed flood model was used further to generate flood inundation maps corresponding to four rainfall return periods (5-‐, 25-‐, 50-‐ and 100-‐year) extracted from the RIDF. The actual flooding extent of Typhoon Pablo was also generated. The impact of climate change was



http://www.wikipedia.org/wiki/ESRI_grid

http://www.wikipedia.org/wiki/ESRI_grid

http://gfs.sourceforge.net/wiki/index.php/GfsRiver

http://gfs.sourceforge.net/wiki/index.php/GfsRiver

http://gfs.sourceforge.net/wiki/index.php/GfsBox

http://gfs.sourceforge.net/wiki/index.php/GfsBox

http://www.wikipedia.org/wiki/Dirichlet_boundary_condition

http://www.wikipedia.org/wiki/Dirichlet_boundary_condition

considered, with projected rainfall applied on predicted land cover changes and respective rainfall runoff for years 2013, 2020 and 2050.

3.9.5 Model pre-‐processing

3.9.5.1 Creation of Geometric Data

Geometric representations of the Cagayan de Oro, Iponan, Mandulog and Iligan river basins were produced from data gathered in the field surveys and aerial LiDAR surveys. These include cross sections, river banks and centerline, and the flood plain boundaries. All the model pre-‐processing were done in ArcView GIS 3.3 and saved as vector files in shapefile format.

3.9.5.2 Model Parameterization

Land cover map information and river bed cross section data were utilized to assign flow resistance coefficients (Manning’s roughness coefficients, n) to river cross section segments (or the portion between cross section points). The transformation to Manning’s, n, land-‐cover classes were from a look-‐up table based on the HEC-‐RAS Hydraulic Reference Manual.

Table 9. Look-‐up table used to convert the land-‐cover map to Manning’s n values

Landcover type Roughness ValueUnclassi1ied 0.01Clouds 0.01Shadows 0.01Plantation (shrubland) 0.40Forest 0.80Seawater 0.01Fallowland 0.05Bare Soil 0.06Grassland 0.25Fresh Water 0.01Thin Clouds 0.01Built-‐up Area 0.01*Adapted from HEC-‐HMS Manual based on Barnes (1967) in Brunner 2010; Values are the same as those in Table 6.

3.9.5.3 Model Boundary Conditions

In GFS, boundary and initial conditions were set in order to: (1) describe the volume of water that flow to the rivers, at 5-‐, 25-‐, 50-‐ and 100-‐ year return periods, from their respective upstream watersheds and tributaries; and (2) reflect the river’s initial conditions prior to flood simulation.

The water runoff from the rainfall events were generated in a separate model with the use of the Hydrologic Engineering Center-‐Hydrologic Modeling System (HEC HMS). Several runoff hydrographs with the same rainfall return periods (5-‐, 25-‐, 50-‐ and 100-‐ year) were produced,



reflecting the change in landscape from 2013 to 2020 and 2050.

The locations of the GFS model boundary conditions (BC) points can be seen in Figures 17-‐20. Two BC points were set in each model for the four rivers, consisting of one open BC at the downstream portion and one upstream BC at the upstream portion. For the open BC points at the downstream portions, predicted tidal data published by NAMRIA was utilized. It is at this area that the effect of the tide greatly affects water surface elevation. Meanwhile, for the upstream BC points, the water discharge data was utilized. The lateral boundary of the flood models employed the use of the boundary outflow condition where it describes flow from the river to the flood plains. The flow chart of the flood modeling exerc ise and its application is shown in Figure 21.

Figure 17. Map showing the location of boundary condition points where time series of water surface elevation and incoming Hlow were assigned for every model simulation in the Cagayan de Oro River



Figure 18. Map showing the location of the boundary conditions points where time series of water surface elevation and incoming Hlow were assigned for every model simulation in the Iponan River.

Figure 19. Map showing the location of the boundary conditions points where time series of water surface elevation and incoming Hlow were assigned for every model simulation in the Mandulog River.



Figure 20. Map showing the location of the boundary conditions points where time series of water surface elevation and incoming Hlow were assigned for every model simulation in the Iligan River.

Figure 21. Flow chart of river Hlood model development.

Flood Model Geometric Data Preparation

(cross-‐sections, banks, river centerline, junctions, 1lood

plains)

Model Parameterization(land-‐cover and river bed roughness coef1icients)

Flood Model Simulation(per type of rainfall event: actual

and hypothetical)

Setting of Boundary Conditions

(initial conditions, water level/tide, upstream in1lows, lateral

in1low)

Post-‐processing of Simulation Results

(Maximum Flood Extent and Inundation Level)

Flood Maps(per type of rainfall event: actual

and hypothetical)

Gerris Flow Solver

Model Pre-‐processing using Arcview

Model Post-‐processing using Arcview



Chapter 4RESULTS OF FIELD SURVEY DATA PROCESSING AND ANALYSIS

4.1 Gathered Cross-‐section Survey Points

Figures 22 to 25 show the elevation data points gathered during the cross section surveys for the four river basins. The total number of points collected in Cagayan de Oro was 4,393; 3,035 in Iponan; 2,228 in Mandulog; and 1,181 in Iligan. The datasets were saved as Microsoft® Excel files and then exported into ArcView™ shapefiles. The files contain the elevation above/below MSL of each data point.

The data points consisted of points along cross section lines and points along the river banks. In Cagayan de Oro, there were 12 cross-‐section lines; 16 in Iponan; 15 in Mandulog; and 11 in Iligan. The total length of the cross section lines were 45.87 km, 21.5 km and 11.75 km, respectively.

Figure 22. Map showing the actual river proHile, cross-‐section, and bathymetry data gathered from the Hield survey in the Cagayan de Oro River.



Figure 23. Map showing the actual river proHile, cross-‐section, and bathymetry data gathered from the Hield survey in the Iponan River.

Figure 24. Map showing the actual river proHile, cross-‐section, and bathymetry data gathered from the Hield survey in the Mandulog River.



Figure 25. Map showing the actual proHile, cross-‐section, and bathymetry data gathered from the Hield survey in the Iligan River.



4.2 Generated DEM of River Basins

Figure 26 to Figure 29 show the generated digital evelevation models (DEM) of the Cagayan de Oro, Iponan, Mandulog and Iligan River Basins, based on the results of the aerial LiDAR surveys and its interpolation to the merged elevation datasets. In generating the DEM, elevation points were used in the interpolation of the river basins and processed for validation. The spatial interpolation was done in ArcGIS 3.3 software. Based on the validation data points, the DEM has an accepted elevation error of +/-‐0.2 meters and a total root mean square error (RMSE) of 0.05m. By the rule-‐of-‐thumb, the allowable RMSE must be less than one half of the DEM’s spatial resolution for it to be acceptable.

Figure 26. Map showing the DEMof the Cagayan de Oro River Basin and Flood Plains; and the flood model domain boundary



Figure 27. Map showing the DEM of the Iponan River Basin and Flood Plains; and the Hlood model domain boundary

Figure 28. Map showing the DEM of the Mandulog River Basin and Flood Plains; and the Hlood model domain boundary



Figure 29. Map showing the DEMof the Iligan River Basin and Flood Plains; and the Hlood model domain boundary

4.3 Gathered River ProHile Survey Points

The river bed data points from the profile survey are shown in Figures 30 to 33. The total length of points collected for Cagayan de Oro was 34 km; 34.73 km for Iponan; 24.70 km for Mandulog; and 32.08 km for Iligan. These vector datasets were converted to ArcView shapefiles and projected in UTM Zone 51 WGS 84 datum. The shapefiles contain the elevation above/below MSL of each data point.



Figure 30. Bed elevation proHile of Cagayan de Oro River. The coordinate of the Hirst point is (940436.917 N, 682364.126 E)

Figure 31. Bed elevation proHile of Iponan River. The coordinate of the Hirst point is (942107.489 N, 677567.738 E)



Figure 32. Bed elevation proHile of Mandulog River. The coordinate of the Hirst point is (912657.717 N, 637105.214 E)

Figure 33. Bed elevation proHile Iligan River. The proHile data was collected from downstream to upstream.

4.4 Merged Elevation Data Points from Field Surveys and Other Sources

Detailed DEM of Cagayan de Oro, Iponan, Manduog and Iligan Rivers and their respective flood Plains were generated using the elevation data points collected from the field surveys and were merged with the aerial LiDAR survey data sets.

4.5 River Bed Characteristics

Through the field surveys, basic geometric characteristics of the four river basins were obtained. These include the river bed elevation profile and the location of the narrowest and widest portion of the rivers along the path of the field surveys.

Figures 30-‐33 show the bed elevation profile of the Cagayan de Oro, Iponan, Mandulog and Iligan Rivers, respectively. The plot points run from downstream to upstream. Known structures along the river were marked in the profile plot for better reference. Figures 34-‐37 show the location of the narrowest and widest portions of the Cagayan de Oro, Iponan, Mandulog and Iligan Rivers, respectively. Table 10 shows the summary of these characteristics.



Table 10. Summary of river characteristics based on the results of the Hield surveys.

River

Average Bed

Elevation from Mean Sea Level,

m.

Average Bed

(Bottom) Slope(m/m)

Minimum/ Narrowest River

Width, m.

Maximum/ Widest River

Width, m.

Location of Narrowest Portion

Location of Widest Portion

Cagayan de Oro 5.63 -‐0.006 62.62 689.97 Brgy.

Macasandig

At the outlet, boundary of

Brgy. Macabalan and Brgy. Bonbon

Iponan 4.37 0.008 29.68 157.28Boundary of Brgy. Iponan and Brgy. Patag

Brgy. Baikingon

Mandulog -‐0.52 -‐0.0004 58.13 334.73 Brgy. Hinaplanon

At the outlet, Brgy. Santa Filomena and Brgy. Santo Rosario

Iligan -‐0.30 -‐0.0009 22.68 258.19 Brgy. Ubaldo Laya

Boundary between Brgy. Palao and Brgy. Mahayhay

It can be seen from the bed elevation plots that the rivers exhibit varied bed and bank topography, from shallow to deep. The Cagayan de Oro bed elevation graphs show great differences in elevation but as the length of the river reaches 7,000 m, the elevation gradually increases, with a few depressions in between.

The Iponan bed elevation graphs show a gradual increase in elevation as it heads upstream. Much like the Cagayan de Oro graphs, it starts with plot points of little elevation differences but as its distance reaches 4,250 m, there is a noticeable increase in elevation.

The Mandulog graph is observed to be irregular, with the presence of ridges and troughs. The bed increases in elevation, though there are a number of points with drastic elevation drops and rises.

The Iligan graph generally shows an increase in elevation, with a few drops in between.



Figure 34. Map showing the narrowest and widest portions of the Cagayan de Oro River.



Figure 35. Map showing the narrowest and widest portions of the Iponan River.



Figure 36. Map showing the narrowest and widest portions of Mandulog River.



Figure 37. Map showing the narrowest and widest portions of the Iligan River.



Chapter 5RESULTS OF RIVER BASIN MODELING

5.1 Modeling domain

The computational domain of the flood models are shown in Figures 38 to 41.

Figure 38. Map showing the Hlood model domain (enclosed in red line) in the Cagayan de Oro River Basin while the subbasin divides are shown in violet line.



Figure 39. Map showing the Hlood model domain (enclosed in red line) in the Iponan River Basin while the subbasin divides are shown in violet line.

Figure 40. Map showing the Hlood model domain (enclosed in red line) in the Mandulog River Basin while the subbasin divides are shown in violet line.



Figure 41. Map showing the Hlood model domain (enclosed in red line) in the Iligan River Basin while the subbasin divides are shown in violet line.

5.2 HEC HMS Model Preparation

Figures 42 to 45 show the watersheds and streams of the Cagayan de Oro, Iponan, Mandulog and Iligan River Basins that were subjected for rainfall-‐-‐-‐runoff modeling.

Hydrologic elements, such as watersheds, reaches and junctions were physically represented in the HEC-‐HMS, to simulate the runoff processes. A lumped parameter was employed. The direct runoff, using the SCS-‐CN Model, was then computed for each watershed, routed and then translated toward each watershed outlet. The latter end utilized the SCS Unit Hydrograph Model. Computed direct runoff hydrographs, with the use of the Muskingum-‐Cunge Method, were then routed through channels toward the main outlet of the watersheds.

Three model components were utilized, namely the basin model, meteorological model and a control specification sets. The Basin Model describes the physical representation of the watersheds and river systems into hydrological elements. Each element is configured with their respective methods for the proper simulation of hydrologic processes. Meanwhile, the Meteorological Model consists of the time series data of rainfall utilized for the simulations. The control specification sets were employed to determine the time step and the duration of the simulation.



Figure 42. The Cagayan de Oro River Basin model generated thru HEC-‐HMS



Figure 43. The Iponan River Basin model generated thru HEC-‐HMS

Figure 44. The Mandulog River Basin model generated thru HEC-‐HMS



Figure 45. The Iligan River Basin model generated thru HEC-‐HMS

5.2.1 Model Pre-‐processing

The Aquaveo™ WMS® (v. 8.1) software was utilized in the preparation of data in the basin model. It is a proprietary watershed modeling software, capable of providing a variety of modeling data formats that include HMS-‐compatible ones.

Datasets required in the basin model preparation were the (i) DEM, (ii) river survey data, (iii) l and cover map, (iv) soil map (from the 2010 Bureau of Soils and Water Management), and (v) river Manning’s roughness index. These data delineated the watershed boundaries (sub-‐basins) and generated the reach elements of the model. The results were then processed in HEC-‐HMS. The Cagayan de Oro River Basin model consists of 56 sub-‐watersheds (or sub-‐basins), 54 reaches and 55 junctions. Its main outlet is at 901C. The Iponan River Basin model consists of 26 watersheds, 25 reaches and 26 junctions. Its main outlet is at 62C. The Mandulog River Basin model consists of 26 watersheds, 24 reaches and 33 junctions. Its main outlet is at 143C. The Iligan River Basin model consists of 31 watersheds, 29 reaches and 30 junctions. Its main outlet is at 75C.



5.2.2 Land Cover Change Projections

Considering the impact of climate change and land cover changes in the watersheds, the event flow at different levels of rainfall conditions was analyzed in the study. Surveys in the areas of interests and other means of data gathering were done to produce the 2013 land cover. The rate of change of one land cover type to another was then determined from primary data, including the NAMRIA land cover data (2003), MSU-‐Marawi forestry department data (2004 & 2006 from the EcoGov Project), classified Archived LANDSAT images and the Bureau of Agricultural Statistics (BAS) crop data (for forest/plantation/grassland change determination).

Thresholds, based on previous studies and BAS data, were applied for the derived change projections in 2020, 2030 and 2050. Each land cover type was given a (-‐) and (+) threshold, depending on the likeliness of being converted to other types and the rate at which this happens. These thresholds also ensure that unrealistic rates of change can be checked and limited (i.e. a calculated 1% yearly rate in built-‐up areas, if left unchecked, would mean covering the half of the entire subwatershed with structures in fifty years’ time, which is too fast considering the expansion and reduction of other land cover types).

Table 11. Land Cover Threshold for Change.Land Cover Type ThresholdThreshold

(-‐) (+)TOTAL Bare Soil -‐10.00% 10.00%Built-‐up Area 0.00% 10.00%Forestland -‐25.00% 5.00%Freshwater -‐2.00% 2.00%Grassland -‐10.00% 10.00%Plantation 0.00% 25.00%

Figure 46. Land cover distribution in the Cagayn de Oro River Basin



Figure 47. Trends in agricultural area planted area for Misamis Oriental and Bukidnon (a) and (b), palay (c) and (d) corn, (e) and (f) coconut; ( g) and (h) coffee; (i) and (j) banana and (m) and (n) trends in pineapple

(a) (b)

(c) (d)

(e) (f)

(g) (h)



(i) (j)

(k) (l)

5.2.3 Model Parameterization

Several parameters were set in the Watershed Rainfall-‐Runoff Model. The Curve Number (CN) parameter of the SCS-‐CN infiltration loss component, and the Watershed lag time of the SCS Unit Hydrograph component, were both computed with the use of the land-‐cover and soil maps. According to NRCS, the lag time, or the amount of time between the centroid of rainfall mass and the peak flow of the produced hydrograph, can be estimated by taking 60% of the time concentration. Meanwhile, the river roughness coefficient parameter, utilized in the Muskingum-‐Cunge Model, was taken from the look-‐up table in the Brunner (2010b).

The CN parameter has several factors, namely, the land cover (Figures 48 to 51), soil type and texture (Figures 52 to 55) and the hydrologic condition. The CNs are assigned to hydrologic soil-‐cover complexes, a combination of hydrologic soil group (soil), land cover and treatment class (cover) to indicate their specific runoff potential. As the CN value increases, the runoff potential also increases.

Before assigning the CNs, the antecedent moisture condition (AMC) of the watersheds (Table 12) were first considered. The AMC utilized in this study is defined as the total rainfall during the growing season, i.e., the first five days after the rainfall event of interest. It was indicated with an index of a minimum value of one and a maximum value of three. The AMC I represents a condition that is unusually dry; AMC II represents an intermediate condition, i.e., a “normal” condition for the watersheds; and AMC III represents a condition that is basically wet. The computation of CN for AMC II was derived using tables developed by the NRCS (National Resources Conservation Services, 1972). The CNs for AMC I and AMC III,



were derived from the equations by Chow, et. Al. (1988), provided that the CN for AMC II is known:

where the CN(I), CN(II), and CN(III) refer to CN values under AMC I, II, and III, respectively.

Table 12. Classification of antecedent moisture conditions (AMC) for the runoff curve number method

AMC Group Total 5-‐day antecedent rainfall (inches)AMC GroupDormant season Growing season

III III

Less than 0.5 Less than 1.40.5 to 1.1 1.4 to 2.1Over 1.1 Over 2.1

Source: Chow, et. al., 1988Source: Chow, et. al., 1988

Table 13 shows the different CN(II) values adapted from the US Natural Resources Conservation Service, formerly known as the SCS (NRCS, 1986). If the watershed area has more than one land cover class and soil type, an area-‐weighted averaging approach was used instead of assigning a CN(II) value. The area-‐weighted CN(II) maps for the river basins are shown for each watershed from Figures 56 to 59.

Table 13. Curve Number (II)values adapted for the rainfall-‐-‐-‐runoff model (Source: NRCS, 1986).

Land-‐use/Land-‐cover AMC II Curve Number for Hydrologic Soil GroupAMC II Curve Number for Hydrologic Soil GroupAMC II Curve Number for Hydrologic Soil GroupAMC II Curve Number for Hydrologic Soil GroupLand-‐use/Land-‐coverA

(sandy)B

(silty)C

(sandy clayey)D

(clayey)Bare Soil 77 86 91 94Built-‐up Area 59 74 82 86Fallow Land 77 86 91 94Forestland 30 55 70 79Freshwater 98 98 98 98Grassland 39 61 74 80Plantation/Shrubland 32 58 72 79

4.2 CN(II)CN(I)= 10 -‐ 0.058 CN(II)

23 CN(II)CN(III)= 10 + 0.13 CN(II)



Figure 48. Land cover map of the Cagayan de Oro River Basin used for the estimation of the CB and watershed lag parameters of the rainfall-‐runoff model.

Figure 49. Land cover map of the Iponan River Basin used for the estimation of the CB and watershed lag parameters of the rainfall-‐runoff model.



Figure 50. Land cover map of the Mandulog River Basin used for the estimation of the CB and watershed lag parameters of the rainfall-‐runoff model.

Figure 51. Land cover map of the Iligan River Basin used for the estimation of the CB and watershed lag parameters of the rainfall-‐runoff model.



Figure 52. Soil map of the Cagayan de Oro River Basin Used for the Estimation of the CN Parameter

Source: Digital soil map of the Philippines published in 2004 by the Bureau of Soils and Water Management, Department of Agriculture

Figure 53. Soil map of the Iponan River Basin Used for the Estimation of the CN Parameter




Figure 54. Soil map of the Mandulog River Basin Used for the Estimation of the CN Parameter


Figure 55. Soil map of the Iligan River Basin Used for the Estimation of the CN Parameter




Figure 56. Map showing the weighted CN values assigned to each watershed in the Cagayan de Oro River Basin(also called Sub-‐basin in HEC-‐HMS)

Figure 57. Map showing the weighted CN values assigned to each watershed in the Iponan River Basin(also called Sub-‐basin in HEC-‐HMS)



Figure 58. Map showing the weighted CN values assigned to each watershed in the Mandulog River Basin(also called Sub-‐basin in HEC-‐HMS)

Figure 59. Map showing the weighted CN values assigned to each watershed in the Iligan River Basin(also called Sub-‐basin in HEC-‐HMS)



5.3 Actual Rainfall Events

Actual rainfall events were used in calibrating t h e f l o o d models of the river basins. For the Cagayan de Oro River Basin, actual rainfall events were taken during typhoon Pablo on 4 December 2012. Peak discharge was 372.9025mm/hr at 4:30 PM. For the Iponan River Basin, actual rainfall events were taken from 15 June 2013 at 7:00 PM to 16 June 2013 at 6:00 AM. Peak discharge was at 12.192 mm/hr. For the Mandulog River Basin, this was taken from December 3, 2012 at 10:00 PM to 4 December 2012 at 10:00 PM. Peak discharge was 333.80mm/hr at 3:30 PM. For the Iligan River Basin, this was taken from 26 December 2012 at 4:00 PM to 29 December 2012 at 11:00 AM and 27 December 2012 at 5:00 PM to 29 December 2012 at 11:00 AM. These rainfall events are plotted in Figures 60 to 63.

Figure 60. Rainfall and hydrograph event recorded at Pelaez Bridge. These rainfall events were used to generate runoff hydrographs to calibrate the Cagayan de Oro River Basin Hlood model

Figure 61. Six-‐hourly rainfall event at San Simon Bridge. This rainfall event was used to generate runoff hydrographs to calibrate the Iponan River Basin Hlood model



Figure 62. Six-‐hourly rainfall event at Mandulog 2 Bridge. This rainfall event was used to generate runoff hydrographs to calibrate the Mandulog River Basin Hlood model

Figure 63. Six-‐hourly rainfall event recorded at Mandulog Bridge. This rainfall event was used to generate runoff hydrographs to calibrate the Iligan River Basin Hlood model



5.4 Hypothetical Rainfall Events

The hypothetical extreme rainfall events were developed in the flood model in order to predict the maximum flood extent and inundation levels. Utilizing the HEC-‐HMS, they were generated from the RIDF curves of PAGASA in Cagayan de Oro and Lumbia, and the Mindanao State University (MSU). The RIDF curves have a duration of 24 hours in the month of December. The time specified was chosen under the assumption that the rainfall events occurred during this period. It also realistically situates the flooding and rainfall conditions in Cagayan de Oro, Iponan, Mandulog and Iligan, since it was also during the month of December that the Sendong and “low pressure area” rains preceding Pablo events occurred. The data interval was at 10 minutes, and the maximum rainfall depths were on the 12th hour, located at the middle of the hydrograph.

The hypothetical rainfall events for 2020 and 2050 are shown in Tables 14, 15 and 16. The rainfall depth is distributed at 10-‐minute intervals in 24-‐hour duration, with the peak manifesting at the middle (the 12th hour). When all the values of the rainfall depth per interval are added up, this produces the total rainfall depth per return period.

Table 14. Rainfall Intensity Frequency Duration (RIDF) data generated by PAGASA for Cagayan de Oro.

RAINFALL (mm)RAINFALL (mm)RAINFALL (mm)RAINFALL (mm)RAINFALL (mm)RAINFALL (mm)RAINFALL (mm)RAINFALL (mm)RAINFALL (mm)RAINFALL (mm)RAINFALL (mm)RAINFALL (mm)

CagayanCagayanCagayanCagayanCagayanCagayanCagayanCagayanCagayanCagayanCagayanCagayan

MonthMEANMEAN

BIAS

Projected Change (%)Projected Change (%)

BIAS CORRECTED PROJECTED CHANGE (mm)BIAS CORRECTED PROJECTED CHANGE (mm)BIAS CORRECTED PROJECTED CHANGE (mm)BIAS CORRECTED PROJECTED CHANGE (mm)BIAS CORRECTED PROJECTED CHANGE (mm)BIAS CORRECTED PROJECTED CHANGE (mm)

MonthMEANMEAN

BIAS

Projected Change (%)Projected Change (%) RR (mm/day)RR (mm/day)RR (mm/day) RR Total (mm)RR Total (mm)RR Total (mm)Month

Observed 1971-‐2000

Model 1971-‐2000

BIAS2020 2050 1971-‐2000 2020 2050 Obs

1971-‐20002020 2050

Jan 3 3 1 19.7 -‐8.7 3 3.6 2.8 93.0 111.6 86.8

Feb 2.1 2 1.1 -‐17 -‐28.9 2.1 1.7 1.4 58.8 47.6 39.2

Mar 1.7 3.4 0.5 -‐10.8 -‐33.6 1.7 1.5 1.1 52.7 46.5 34.1

Apr 1.5 6 0.3 -‐26.2 -‐38.1 1.5 1.1 0.9 45.0 33.0 27.0

May 2.7 7.2 0.4 -‐15.5 -‐30.9 2.7 2.3 1.9 83.7 71.3 58.9

Jun 6.9 7.4 0.9 -‐11.5 -‐26.3 6.9 6.1 5.1 207.0 183.0 153.0

July 6.8 5.9 1.2 -‐24.3 -‐33.9 6.8 5.2 4.5 210.8 161.2 139.5

Aug 6.2 5.5 1.1 -‐17.9 -‐35.7 6.2 5.1 4.0 192.2 158.1 124.0

Sept 6.7 4.8 1.4 -‐10.7 -‐18.7 6.7 6.0 5.4 201.0 180.0 162.0

Oct 6.1 5.8 1.1 10.1 -‐15 6.1 6.7 5.2 189.1 207.7 161.2

Nov 4.4 7.1 0.6 -‐14 -‐29.9 4.4 3.8 3.1 132.0 114.0 93.0

Dec 3.2 6.4 0.5 -‐2.2 -‐26.3 3.2 3.2 2.4 99.2 99.2 74.4



Table 15. Rainfall-‐-‐-‐Intensity Frequency Duration (RIDF) data generated by PAGASA for Lumbia


LumbiaLumbiaLumbiaLumbiaLumbiaLumbiaLumbiaLumbiaLumbiaLumbiaLumbiaLumbia

MonthMEANMEAN

BIAS



MonthMEANMEAN

BIAS



Model 1971-‐2000

BIAS2020 2050 1971-‐2000 2020 2050

Obs 1971-‐2000 2020 2050

Jan 2.7 3.1 0.9 17.2 2.7 2.7 3.1 2.7 83.7 96.1 83.7

Feb 2.3 2.5 0.9 -‐18.5 -‐19.7 2.3 1.9 1.8 64.4 53.2 50.4

Mar 1.4 3.2 0.4 -‐5.4 -‐19.9 1.4 1.4 1.2 43.4 43.4 37.2

Apr 1.9 4.3 0.4 -‐16.9 -‐27.3 1.9 1.6 1.4 57.0 48.0 42.0

May 3.5 3.6 1 4.5 17.2 3.5 3.7 4.2 108.5 114.7 130.2

Jun 6.9 3 2.3 14 16.5 6.9 7.9 8.1 207.0 237.0 243.0

July 7.9 2.7 2.9 -‐6.8 -‐12.7 7.9 7.3 6.9 244.9 226.3 213.9

Aug 6.6 2.9 2.3 -‐0.1 -‐4.9 6.6 6.6 6.3 204.6 204.6 195.3

Sept 6.8 2.7 2.5 0.3 2.3 6.8 6.8 7 204.0 204.0 210.0

Oct 6.1 3 2 42.5 16.6 6.1 8.6 7.1 189.1 266.6 220.1

Nov 4.6 4 1.2 -‐13.6 -‐2.3 4.6 4 4.5 138.0 120.0 135.0

Dec 3.4 4.4 0.8 18.1 10.2 3.4 4 3.8 105.4 124.0 117.8

Table 16. Rainfall-‐-‐-‐Intensity Frequency Duration (RIDF) data generated by PAGASA for MSU.


MSUMSUMSUMSUMSUMSUMSUMSUMSUMSUMSUMSU

MonthMEANMEAN

BIAS



MonthMEANMEAN

BIAS



Model 1971-‐2000

BIAS2020 2050 1971-‐2000 2020 2050

Obs 1971-‐2000 2020 2050

Jan 3.8 2.8 1.3 13.3 -‐0.9 3.8 4.3 3.7 117.8 133.3 114.7

Feb 4.1 2.8 1.5 -‐1.5 -‐15.5 4.1 4.1 3.4 114.8 114.8 95.2

Mar 3 3.9 0.8 -‐3.4 -‐10.6 3 2.9 2.7 93.0 89.9 83.7

Apr 3.1 5.3 0.6 -‐15.8 -‐7.8 3.1 2.6 2.9 93.0 78.0 87.0

May 6 5.7 1 0.9 -‐0.5 6 6 5.9 186.0 186.0 182.9

Jun 8.2 5 1.6 6.6 -‐1 8.2 8.7 8.1 246.0 261.0 243.0

July 7 4.3 1.6 -‐15 -‐17.1 7 5.9 5.8 217.0 182.9 179.8

Aug 6.1 4.4 1.4 -‐13.7 -‐5.4 6.1 5.2 5.7 189.1 161.2 176.7

Sept 6.9 4 1.7 -‐4.3 5.3 6.9 6.6 7.3 207.0 198.0 219.0

Oct 6.1 4.4 1.4 11.7 1.4 6.1 6.8 6.1 189.1 210.8 189.1

Nov 5 4.2 1.2 -‐5.9 -‐11.9 5 4.7 4.4 150.0 141.0 132.0

Dec 4.2 3.9 1.1 8.3 1.5 4.2 4.5 4.2 130.2 139.5 130.2



5.5 Rainfall-‐Runoff Simulations using HEC-‐HMS

The rainfall-‐runoff simulations were able to produce outflow hydrographs (discharge). These are shown in Figures 65 to 112. Figures 65 to 112 depict the volume of water passing along the Cagayan de Oro, Iponan, Mandulog and Iligan rivers, with respect to the land cover in 2013 and subsequent projected changes in 2020 and 2050. The outflow hydrographs provide information on the expected amount of water flow from the rivers, given the same series of rainfall events (5, 25, 50 and 100 year return periods).

The total outflow volume and peak outflow for the 4 rivers are shown in Tables 17 to 32. The former is the amount of water which passes through the river during the simulation while the latter is maximum rate of flow during the simulation period.

Two types of simulations were done. First, those based on actual rainfall events, and second, those based on the hypothetical rainfall events. The former was determined to have an AMC of III (“wet condition”) while the latter had an AMC II (“normal condition”). This classification was based on the rainfall data provided by PAGASA. It is assumed that for each simulation, rainfall was distributed across all the watersheds within the domains of the models.

The interface of the HEC-‐HMS Model can be seen in Figure 64. It consists of a main menu and several smaller windows utilized for model set-‐up, parameters, construction and editing of simulation scenarios, visualization of results such as graphs and tables, and exporting of the results.

Figure 64. Interface of the Cagayan de Oro River Basin HEC HMS Rainfall-‐Runoff Model developed in this project

There were a total of four HEC-‐HMS simulations produced for each river basin to derive the

Graphical Results Window for Junctions (shows runoff hydrographs at river

junctions)

HEC HMS Model Table of Contents (model components and parameters, simulation scenarios, and model results)

Basin Model Schematic Window( contains the arrangement of model elements such as sub-basins, reaches, junctions and outlet)

HEC HMS Main Menu



runoff hydrographs. Four were produced for each of the 2013, 2020 and 2050 land covers. For each land cover, one simulation was done for the 5-‐year rainfall event and the next three for the 25-‐, 50-‐ and 100-‐year rainfall events. The simulation time step employed was 10 minutes. 5.5.1 Simulated Runoff from Cagayan de Oro River

5.5.1 Simulated Runoff from Cagayan de Oro River

The estimated discharge from the Cagayan de Oro watershed for the different rainfall return periods (5, 25, 50-‐ and 100-‐year) of 2013 and projected scenarios of 2020 and 2050 are found in Figures 65 to 76. In all of the modeled scenarios, the maximum discharge is found within 9 hours after the peak rainfall. The peak outflow rates increasingly vary from 5-‐year (1,598 CMS) to 100-‐year (5,585.10) return periods (See Table 17 to Table 20). The peak outflows however vary from 2013, which gradually increased to 2020 before lowering back to 2050.

Interestingly, in the 100-‐year RP, the model results exhibited a double peak for the 100-‐year return period in Figure 68. This is due to the presence of the large subcatchment in the eastern side of the watershed which is responsible for the first lower peak. The second larger peak comes from the Kabula side of the catchment.

Figure 65. Cagayan de Oro watershed outHlow hydrographs for the 5-‐Year Rain return period in 2013 land cover conditions.

Figure 66. Cagayan de Oro Watershed simulated outHlow hydrographs for the 25-‐Year rain return period with 2013 land cover conditions.



Figure 67. Cagayan de Oro Watershed OutHlow Hydrographs Generated from the Runoff Hydrographs Produced by the HEC HMS Model for the 50-‐ Year Rainfall Event in 2013 Land Cover.


Figure 69. Cagayan de Oro Watershed OutHlow Hydrograph for the 5-‐ Year Rainfall Event in 2020 land cover and rainfall pattern from climate change projection.



Figure 70. Cagayan de Oro Watershed OutHlow Hydrograph for the 25-‐ Year Rainfall Event in 2020 land cover and rainfall pattern from climate change projection.











Table 17. Simulated Outflow Volume and Peak Outflow Rate at Cagayan De Oro River for the 5-‐ Year Rainfall Event.

Land Cover Year Total outflow volume during the period (mm) Peak outflow rate (cms)

2013 154,557.40 1,598.32020 206,056.20 2,207.82050 186,164.10 1,974.9

Table 18. Simulated Outflow Volume and Peak Outflow Rate at Cagayan De Oro River for the 25-‐ Year Rainfall Event


2013 270,989.30 2,965.92020 347,249.40 3,870.32050 316,811.40 3,515.4



2013 318,914.20 3,534.302020 407,682.30 4,578.302050 373,080.20 4,174.10





2013 684,862.30 5,585.102020 896,689.50 7,015.402050 834,487.00 6,591.70



5.5.2 Simulated Runoff from Iponan River

The estimated discharge from the Iponan watershed for the different rainfall return periods (5-‐, 25-‐, 50-‐ and 100-‐year) of 2013 land cover and projected scenarios land cover scerarios for 2020 and 2050 are found in Figures 77 to 88. The maximum discharge is found within 12 to 16 hours after the peak rainfall. The more frequent return periods peaks much slower than the more intense and less frequent rainfall events.

The peak outflow rates increasingly vary from 5-‐year (592 cms) to 100-‐year (1,589) return periods (Tables 21 to 24). The peak outflows maintain the same trend as that of Cagayan de Oro which however increases from 2013 to 2020 before lowering back to 2050.

Figure 77. Iponan Watershed OutHlow Hydrographs Generated from the Runoff Hydrographs Produced by the HEC HMS Model for the 5-‐ Year Rainfall Event in 2013 Land Cover.

Figure 78. Iponan watershed outHlow hydrographs for the 25-‐year rainfall event in 2013 land cover.



















Table 21. Simulated outflow volume and peak outflow rate at Iponan River for the 5-‐year rainfall event.


2013 50,490.90 592.402020 76,193.70 866.902050 67,298.50 760.10

Table 22. Simulated outflow volume and peak outflow Rate at Iponan River for the 25-‐year rainfall event.


2013 88,325.70 1,122.202020 122,498.40 1,530.702050 110,417.80 1,373.70

Table 23. Simulated Outflow Volume and peak outflow rate at Iponan River for the 50-‐ Year Rainfall Event


2013 103,637.20 1,349.102020 141,908.50 1,818.802050 128,602.50 1,641.90

Table 24. Simulated outflow volume and peak outflow rate at Iponan River for the 100-‐year rainfall event


2013 119.892.00 1,589.702020 161,373.10 2,109.902050 146,815.50 1,915.10



5.5.3 Simulated Runoff from Mandulog River

The estimated discharge from the Iponan watershed for the different rainfall return periods (5-‐, 25-‐, 50-‐ and 100-‐year) of 2013 and projected scenarios of 2020 and 2050 are found in Figures 89 to Figure 100. The maximum discharge peaks immediately (one hour or less) after the peak rainfall. The more frequent return periods peaks much slower than the more intense and less frequent rainfall events.

The peak outflow rate increases vary from 5-‐year (415 cms) to 100-‐year (1,586 cms) return periods. The peak outflows experience the same trend as that of Cagayan de Oro which however increases from 2013 to 2020 before lowering back to 2050.

Figure 89. Mandulog Watershed OutHlow Hydrographs Generated from the Runoff Hydrographs Produced by the HEC HMS Model for the 5-‐ Year Rainfall Event in 2013 Land Cover.
























Table 25. Simulated Outflow Volume and Peak Outflow Rate at Mandulog River for the 5-‐ Year Rainfall Event.


2013 57,285.90 415.102020 74,823.50 611.502050 68,926.30 548.60

Table 26. Simulated Outflow Volume and Peak Outflow Rate at Mandulog River for the 25-‐ Year Rainfall Event


2013 111,210.90 1,020.602020 139,384.90 1,366.502050 130,036.40 1,269.10



2013 133,904.40 1,295.402020 167,518.70 1,705.702050 156,693.50 1,592.80



2013 158,208.70 1,586.702020 196,021.30 2,050.802050 183,705.50 1,922.10



5.5.4 Simulated Runoff from Iligan River

The estimated discharge from the Iponan watershed for the different rainfall return periods (5, 25, 50-‐ and 100-‐year) of 2013 and projected scenarios of 2020 and 2050 are found in Figures 101 to 112. The maximum discharge peaks within 8 hours after the peak rainfall. The more frequent return periods peaks much slower than the more intense and less frequent rainfall events. For the 100-‐year return period, the peak discharges occurs within 7 hours after the peak rainfall in the watershed.

The peak outflow rates increasingly vary from 5-‐year (592 CMS) to 100-‐year (1589) return periods (See Table 29 to Table 32). The peak outflow has the same trend as that of Cagayan de Oro which however increases from 2013 d to 2020 before lowering back to 2050.

Figure 101. Iligan Watershed OutHlow Hydrographs Generated from the Runoff Hydrographs Produced by the HEC HMS Model for the 5-‐ Year Rainfall Event in 2013 Land Cover.




















Table 29. Simulated Outflow Volume and Peak Outflow Rate at Iligan River for the 5-‐ Year Rainfall Event


2013 19,776.30 384.90

2020 26,786.10 531.60

2050 25,397.70 491.40



2013 33,737.80 704.20

2020 43,410.40 919.90

2050 41,134.70 854.10



2013 39,345.80 840.60

2020 50,387.60 1,086.40

2050 47,750.20 1,009.20



Table 32. Simulated Outflow Volume and Peak Outflow Rate at Iligan for the 100-‐ Year Rainfall Event


2013 45,295.00 981.30

2020 57,403.80 1,255.00

2050 54,370.80 1,167.00



Chapter 6RESULTS OF FLOOD MODEL SIMULATIONS

The flood model simulations were able to produce flood inundation maps based on maximum flood depth. These are shown in Figures 113 to 161. The inundation levels are generalized in 5 levels to relate to common height references in public. Maximum flood water levels lower than 0.2 m are intentionally not colored. 0.2 to 0.5m levels relate to foot to knee heights. 1m refer to level to waist level heights while 2m levels reach the ceiling of first floor of a house or a building. 5m heights will cover two storeys.

Simulations were configured for the 5-‐, 25-‐, 50-‐ and 100-‐ year events. Each event was done for the 2013, 2020 and 2050 land covers. Outflow hydrographs (discharge) from the HEC-‐HMS model and baseflow measurements were utilized to generate the simulations.

These flood inundation maps will be utilized in the assessment of flood hazards in the cities of Cagayan de Oro and Iligan.

6.1 Simulated Flood Maps and Analysis

The GFS simulated the water level, at every time step, for all the models. The results determine which areas in the flood domain are affected by flood. The flood maps, reflecting the maximum flood extent and inundation levels (flood depths), were then generated. For the 2013 land cover, the cumulative rainfall for all the flood plains is 129.2 mm during the 5-‐ year rainfall event. It is at 189.9 mm, 214.8 mm and 239. 7 mm for the 25-‐, 50-‐ and 100-‐year rainfall events, respectively. In the 2020 land cover, the cumulative rainfall for all the flood plains is 151 mm during the 5-‐year rainfall event. It is at 223.3 mm, 252.7 mm and 282 mm for the 25-‐, 50-‐ and 100-‐ year rainfall events, respectively. Lastly, for the 2050 land cover, the cumulative rainfall for all the flood plains is 144.4 mm during the 5-‐year rainfall event. It is at 212.1 mm, 240 mm and 267.97 mm for the 25-‐, 50-‐ and 100-‐year rainfall events, respectively.

The cumulative rainfall for all the flood plains increases as the year of the rainfall event increases. Meanwhile, comparing the results from the different land covers, cumulative rainfall increases from the results of the 2013 land cover to 2020. This, however, decreases in the 2050 land cover results.

6.1.1 Cagayan de Oro Simulated Flooding

6.1.1.1 Cagayan de Oro River Flooding: 2013 Land Cover Condition

The Cagayan de Oro flood inundation maps simulated for the rainfall events in the 2013 land cover scenario can be seen in Figures 113, 115, 117 and 119. Over 44% of the Cagayan de Oro area will experience flooding during the rainfall events. Barangays Macasandig, Balulang and Indahag will experience the greatest extent of flooding for all the flood depths, while, Barangays Macasandig, Kauswagan and Balulang will experience the greatest extent for flood depths of 2.00 m and greater.

Figures 114, 116, 118 and 120 summarizes the extent of flooding for barangays for each



depth in the 5-‐, 25-‐, 50-‐ and 100 year rainfall events, respectively. Barangays 1 to 40 are not shown in the figures due to the minor effect of flooding in the areas.

For the 5-‐year rainfall event, 77% of the total affected area will experience a flood depth of less than 0.20 meters. Meanwhile, 10% will experience floods between 0.20 to 0.50 meters high. Areas that will experience a flood depth of 0.51 to 1.00, 1.01 to 2.00, 2.01 to 5.00 and greater than 5.00 meters amount to only 5%, 4%, 3% and 2% of the total affected area, respectively. Barangay Macasandig will experience the greatest extent of flooding, contributing to over 38% of the affected area for flood depths of 2.00 meters and above. This is followed by Barangays Kausawagan and Balulang, amounting to 9% each.

For the 25 year rainfall event, 68% of the total affected area will experience a flood depth of less than 0.20 meters. Over 12% will experience floods between 0.20 to 0.50 meters high. Areas that will experience a flood depth of 0.51 to 1.00 and 2.01 to 5.00 meters will each amount to 6% each. Flood depths of 1.01 to 2.00 and greater than 5.00 meters will affect 5% and 3% of the total affected area, respectively. Barangay Macasandig will still experience the greatest extent of flooding, contributing to over 36% of the affected area for flood depths of 2.00 meters and above. This is followed by Barangays Kauswagan and Balulang, amounting 10% and 9%, respectively.

For the 50 year rainfall event, 65% of the total affected area will experience a flood depth of less than 0.20 meters. Meanwhile, 12% will experience floods between 0.20 to 0.50 meters high. Areas that will experience a flood depth of 0.51 to 1.00, 1.01 to 2.00, 2.01 to 5.00 and greater than 5.00 meters amount to only 6%, 5%, 7% and 3% of the total affected area, respectively. Barangay Macasandig will contribute over 35% of the affected area for flood depths of 2.00 meters and above. This is followed by Barangays Kauswagan and Balulang, amounting to 10% each.

For the 100 year rainfall event, 59% of the total affected area will experience a flood depth of less than 0.20 meters. Over 12% will experience floods between 0.20 to 0.50 meters high. Flood depths of 0.51 to 1.00 and 2.01 to 5.00 meters will affect 8% and 9% of the total affected area, respectively. Areas that will experience a flood depth of 1.01 to 2.00 and greater than 5.00 meters will each amount to 6% each. Barangay Macasandig will contribute over 31% of the affected area for flood depths of 2.00 meters and above. This is followed by Barangays Kauswagan and Balulang, amounting 12% and 10%, respectively.



Figure 113. Flood Map of the Cagayan de Oro Hlood plain showing the maximum Hlood extent and depths resulting from the 5 year rainfall event for the 2013 land cover overlain over the hillshaded topography. The

roads, streets and the barangay names are also superimposed in the inundation map.



Figure 114. Estimated Extent of Flooding in Barangays in Cagayan de Oro for the 5-‐Year Rainfall Event, 2013 Land Cover



Figure 115. Flood Map of the Cagayan de Oro Hlood plain showing the maximum Hlood extent and depths resulting from the 25 year rainfall event for the 2013 land cover



Figure 116. Estimated Extent of Flooding in Barangays in Cagayan de Oro for the 25 Year Rainfall Event, 2013 Land Cover



Figure 117. Flood Map of the Cagayan de Oro Hlood plain showing the maximum Hlood extent and depths resulting from the 50 year rainfall event for the 2013 land cover condition.






Figure 119. Flood Map of the Cagayan de Oro Hlood plain showing the maximum Hlood extent and depths resulting from the 100 year rainfall event for the 2013 land cover condition.





Figures 121 to 125 show the simulated flood maps for the 2020 land cover scenario. Peak discharge is at 2211.2, 3873, 4578.3 and 7015.4 m3 for the 5, 25, 50 and 100 year rainfall events, respectively. While the lag period between the peak of rainfall and discharge is at 9 hours and 50 minutes for the 5 year rainfall event, and 9 hours and 30 minutes for both the 25 and 50 year rainfall events, it is at 24 hours and 10 minutes for the 100 year rainfall event.

The 100 year return period results show that Barangays 6, 7, 10, 13, 15, 17, 18, 22, 24 and 35 will be fully flooded with depths of 2.00 meters and greater. Sixty-‐nine percent of Barangay 6 will experience floods with depths that are greater than 5.00m.



Figure 121. Flood inundation map of the Cagayan de Oro Hlood plain showing the maximum Hlood extent and depths resulting from the 5 year rainfall event for the 2020 land cover.













Figures 125 to 128 show the simulated flood maps for the 2050 land cover scenario. Peak discharge is at 1976.8, 3515.4, 4174.1 and 6591.7 m3 for the 5, 25, 50 and 100 year rainfall events, respectively. The lag period between the peak of rainfall and discharge is the same with the results from the 2020 land cover.

The 100 year return period results show almost the same with the results in the 2020 land cover. Barangays 6, 7, 10, 13, 15, 17, 18, 24 and 35will be fully flooded with depths of 2.00 meters and greater. Meanwhile, 99% of Barangay 22 will experience flooding. Sixty-‐six percent of Barangay 6 will experience floods with depths that are greater than 5.00m.















6.1.2 Iponan Simulated Flooding

6.1.2.1 Iponan River Flooding: 2013 Land Cover Condition

The Iponan flood maps simulated for the rainfall events in the 2013 land cover scenario can be seen in Figures 129, 131, 133 and 135. About 10110 hectares in the Iponan area will experience flooding during the rainfall events, a majority of which with a depth of less than 2.00 meters. Of the total affected area, 3185 hectares will experience depths of more than 2.00 meters. Barangays Malanang, Patag and Canito-‐An will experience the greatest extent of flooding for all the flood depths. Meanwhile, Barangays Canito-‐An, Pagatpag and San Simon will experience the greatest extent of flooding for flood depths of 2.00 meters and greater.

Figures 130, 132, 134 and 136 summarize the extent of flooding for barangays for each depth in the 5, 25, 50 and 100 year rainfall events, respectively.

For the 5 year rainfall event, 68% of the total affected area will experience a flood depth of less than 0.20 meters. Over 9% will experience floods between 0.20 to 0.50 meters high. Flood depths of 0.51 to 1.00, 1.01 to 2.00, 2.01 to 5.00 and greater than 5.00 meters will affect 8%, 7%, 6% and 1% of the total affected area, respectively. When taking all flood depths into consideration, Barangay Malanang, Patag and Canito-‐An will most extensively experience flooding. However, when taking into consideration flood depths of 2.00 meters and above, Barangay Canito-‐An will experience flood most extensively, contributing over 15% of the affected area. This is followed by Barangays Pagatpat and San Simon, amounting 10% and 9%, respectively.

For the 25 year rainfall event, 64% of the total affected area will experience a flood depth of less than 0.20 meters. Floods between 0.20 to 0.50, 0.51 to 1.00 and 1.01 to 2.00 meters high will each be experienced by 9%. Flood depths of 2.01 to 5.00 and greater than 5.00 meters will affect 7% and 2% of the total affected area, respectively. Barangay Canito-‐An will experience flood most extensively for depths of 2.00 meters and greater, contributing over 15% of the affected area. This is followed by Barangays Pagatpat and San Simon, amounting 10% and 9%, respectively.For the 50 year rainfall event, 63% of the total affected area will experience a flood depth of less than 0.20 meters. Floods between 0.20 to 0.50 and 0.51 to 1.00 meters high will each be experienced by 9%. Flood depths of 1.01 to 2.00, 2.01 to 5.00 and greater than 5.00 meters will affect 10%, 7% and 2% of the total affected area, respectively. Barangay Canito-‐An will experience flood most extensively for depths of 2.00 meters and greater, contributing over 15% of the affected area. This is followed by Barangays Pagatpat and San Simon, amounting 11% and 9%, respectively.

For the 100 year rainfall event, 62% of the total affected area will experience a flood depth of less than 0.20 meters. Floods between 0.20 to 0.50 and 2.01 to 5.00 meters high will each be experienced by 8%. Flood depths of 0.51 to 1.00, 1.01 to 2.00, and greater than 5.00 meters will affect 9%, 11% and 3% of the total affected area, respectively. Barangay Canito-‐An will experience flood with depths of 2.00 meters or greater most extensively, contributing over 15% of the affected area. This is followed by Barangays Pagatpat and San Simon, amounting 11% and 9%, respectively.



Figure 129. Flood Map of the Iponan Hlood plain showing the maximum Hlood extent and depths resulting from the 5 year rainfall event for the 2013 land cover condition.



Figure 130. Estimated Extent of Flooding of Iponan River for the 5 Year Rainfall Event, 2013 Land Cover






















Figures 137 to 140 show the simulated 1lood maps for the 2020 land cover scenario. Peak discharge is at 866.9, 1530.7, 1818.8 and 2109.9 m3 for the 5, 25, 50 and 100 year rainfall events, respectively. The lag period between the peak of rainfall and discharge is at 15 hours and 30 minutes for the 5 year rainfall event and 13 hours and 30 minutes for the 25 year rainfall event. It is at 13 hours for the 50 year rainfall event and 12hours and 30 minutes for the 100 year rainfall event.

The 100 year return period results show that Barangay Barrra will be fully 1looded while 98% of Barangay Igpit will experience 1loods with depths of 2.00 meters and greater. Sixteen percent of Barangay Taglimao and 14% of Barangay Pagalungan will experience 1loods with depths that are greater than 5.00m.
















Figures 141 to 144 show the simulated 1lood maps for the 2050 land cover scenario. Peak discharge is at 760.1, 1373.7, 1641.9 and 1915.1 m3 for the 5, 25, 50 and 100 year rainfall events, respectively. The lag period between the peak of rainfall and discharge is at 16 hours for the 5 year rainfall event and 14 hours for the 25 year rainfall event. It is at 13 hours and 20 minutes for the 50 year rainfall event and 12hours and 50 minutes for the 100 year rainfall event.

The 100 year return period results show lower 1lood extent than those in the 2020 land cover. Ninety-‐nine percent of Barangay Barrra’s area will experience 1loods with depths of 2.00 meters and greater while it will be 97% of Barangay Igpit. Fifteen percent of Barangay Taglimao and 13% of Barangay Pagalungan will experience 1loods with depths that are greater than 5.00m.






Figure 142. Flood Map of the Iponan Hlood plain showing the maximum Hlood extent and depths resulting from the 25 year rainfall event for the 2050 land cover condition









6.1.3 Mandulog River Simulated Flooding

6.1.3.1 Mandulog River Flooding: 2013 Land Cover Condition

The Mandulog flood maps simulated for the rainfall events in the 2013 land cover scenario can be seen in Figures 145, 147, 149 and 151. Thirteen percent of the Mandulog area will experience flooding during the 5 year rainfall event. Barangays Hinaplanon, San Miguel and Santiago will have the largest flooded areas of 52%, 35% and 35%, respectively. For the 25 year rainfall event, the flooded areas will increase to about 20% of Mandulog. The largest areas that will experience flooding are Barangays Hinaplanon, Santiago and Santo Rosario. They will be flooded by around 74%, 71% and 85%, respectively. For the 50 year rainfall event, around 21% will experience floods. Barangays Hinaplanon, Santiago and Santo Rosario will still have the largest flooded areas of 77%, 73% and 87%, respectively. For the 100 year rainfall event, about 22% will be flooded. The same barangays will experience the most flooding, reaching over 79%, 75% and 88% of their areas.


For the 5 year rainfall event, the peak discharge is 415.1 m3. There will be a 1 hour and 10 minute lag period between it and the peak of rainfall. Over 72% of the total affected area will experience a flood depth of less than 0.20 meters. Over 12% will experience floods between 0.20 to 0.50 meters high. Flood depths of 0.51 to 1.00, 1.01 to 2.00, 2.01 to 5 and greater than 5 meters will affect 8%, 4%, 3% and 1% of the total affected area, respectively. Barangay Santa Filomena will experience the greatest extent of flooding for flood depths of 2.00 meters and greater, contributing 35% of the affected area for flood depths of 2.00 meters and above. This is followed by Barangays Hinaplanon and San Roque, amounting 19% and 15%, respectively.

For the 25 year rainfall event, the peak discharge is 1020.6 m3. There will be a 1 hour and 10 minute lag period between it and the peak of rainfall. Over 57% of the total affected area will experience a flood depth of less than 0.20 meters. Floods depths between 0.20 to 0.50, 1.01 to 2.00 and 2.01 to 5 meters high will each affect 11%. Meanwhile, 7% and 3% of the total affected area will experience flood depths of 0.51 to 1.00 and greater than 5 meters, respectively. Barangay Santa Filomena will experience the greatest extent of flooding for flood depths of 2.00 meters and greater, contributing 32% of the affected area for flood depths of 2.00 meters and above. This is followed by Barangays Hinaplanon and San Roque, amounting to 18% each.

For the 50 year rainfall event, the peak discharge is 1295.4 m3. There will be a 1 hour lag period between it and the peak of rainfall. Over 55% of the total affected area will experience a flood depth of less than 0.20 meters. Floods depths between 0.20 to 0.50 and 1.01 to 2.00 meters high will each affect 11%. Meanwhile, 8%, 12% and 3% of the total affected area will experience flood depths of 0.51 to 1.00, 2.01 to 5 and greater than 5 meters, respectively. Barangay Santa Filomena will experience the greatest extent of flooding for flood depths of 2.00 meters and greater, contributing 31% of the affected area for flood depths of 2.00 meters and above. This is followed by Barangays Hinaplanon and San Roque, amounting to 18% and 17%, respectively.



For the 100 year rainfall event, the peak discharge is 1586.7 m3. There will be a 1 hour lag period between it and the peak of rainfall. Over 54% of the total affected area will experience a flood depth of less than 0.20 meters. Floods depths between 0.20 to 0.50 and 1.01 to 2.00 meters high will each affect 11%. Meanwhile, 9%, 13% and 4% of the total affected area will experience flood depths of 0.51 to 1.00, 2.01 to 5 and greater than 5 meters, respectively. Barangay Santa Filomena will experience the greatest extent of flooding for flood depths of 2.00 meters and greater, contributing 30% of the affected area for flood depths of 2.00 meters and above. This is followed by Barangays Hinaplanon and San Roque, amounting to 18% and 17%, respectively.



Figure 145. Flood Map of the Mandulog Hlood plain showing the maximum Hlood extent and depths resulting from the 5 year rainfall event for the 2013 land cover condition

Figure 146. Estimated Extent of Flooding in Barangays in Mandulog for the 5 Year Rainfall Event, 2013 Land Cover
















Figures 153 to 156 show the simulated flood maps for the 2020 land cover scenario. Peak discharge is at 611.5, 1366.5, 1705.7 and 2050.8 m3 for the 5, 25, 50 and 100 year rainfall events, respectively. While the lag period between the peak of rainfall and discharge is at 1 hour and 10 minutes for the 5 year rainfall event, it is at 1 hour for the 25, 50 and 100 year rainfall events.

The 100 year return period results show that 92% of Barangay Santo Rosaio and 85% of Barangay Hinaplanon will experience floods with depths of 2.00 meters and greater. However, flood depths greater than 5.00m will be experienced by 17% of Barangay San Roque and 5% of Barangay Santa Filomena.










Figures 157 to 160 show the simulated flood maps for the 2050 land cover scenario. Peak discharge is at 548.6, 1269.1, 1592.8 and 1922.1m3 for the 5, 25, 50 and 100 year rainfall events, respectively. While the lag period between the peak of rainfall and discharge is at 1 hour and 10 minutes for the 5 year rainfall event, it is at 1 hour for the 25, 50 and 100 year rainfall events.

The 100 year return period results show slightly lower flood extent than those in the 2020 land cover. Ninety-‐two percent of Barangay Santo Rosario and 84% of Barangay Hinaplanon will experience floods with depths of 2.00 meters and greater. Flood depths greater than 5.00m will still be experienced by Barangays San Roque and Santa Filomena, amounting to 16% and 5%, respectively.











6.1.4 ligan River: Simulated Hlooding

6.1.4.1 Iligan River 2013 Land Cover Scenario

The Iligan flood maps simulated for the rainfall events in the 2013 land cover scenario can be seen in Figures 161, 163, 165 and 167. Over 27% of the Iligan area will experience flooding during the rainfall events. Barangays Tubod, Villa Verde and Tambacan will experience the greatest extent of flooding for all the flood depths. However, when considering flood depths of 2.00 meters and greater, Barangays Tubod, Ubaldo Laya and Villaverde will be the ones to experience the greatest extent.


For the 5 year rainfall event, 81% of the total affected area will experience a flood depth of less than 0.20 meters. Floods depths between 0.20 to 0.50 meters high will affect 11%. Meanwhile, 5%, 2%, 2% and 0.1% of the total affected area will experience flood depths of 0.51 to 1.00, 1.01 to 2.00, 2.01 to 5 and greater than 5 meters, respectively. Barangay Ubaldo Laya will contribute to 17% of the affected area for flood depths of 2.00 meters and above. This is followed by Barangays Tubod and Villa Verde, amounting 16% and 12%, respectively.

For the 25 year rainfall event, 72% of the total affected area will experience a flood depth of less than 0.20 meters. Floods depths between 0.20 to 0.50 meters high will affect 14%. Meanwhile, 7%, 4%, 2% and 0.5% of the total affected area will experience flood depths of 0.51 to 1.00, 1.01 to 2.00, 2.01 to 5 and greater than 5 meters, respectively. Barangay Tubod will contribute to 16% of the affected area for flood depths of 2.00 meters and above. This is followed by Barangays Ubaldo Laya and Villa Verde, both amounting to 15%.

For the 50-‐year rainfall event, 69% of the total affected area will experience a flood depth of less than 0.20 meters. Floods depths between 0.20 to 0.50 meters high will affect 15%. Meanwhile, 8%, 6%, 3% and 1% of the total affected area will experience flood depths of 0.51 to 1.00, 1.01 to 2.00, 2.01 to 5 and greater than 5 meters, respectively. Barangay Villa Verde will experience the greatest extent of flooding, contributing to 16% of the affected area for flood depths of 2.00 meters and above. This is followed by Barangays Tubod and Ubaldo Laya, amounting 15% and 14%, respectively.

For the 100 year rainfall event, 66% of the total affected area will experience a flood depth of less than 0.20 meters. Floods depths between 0.20 to 0.50 meters high will affect 15%. Meanwhile, 8%, 7%, 3% and 1% of the total affected area will experience flood depths of 0.51 to 1.00, 1.01 to 2.00, 2.01 to 5 and greater than 5 meters, respectively. Barangay Villa Verde will experience the greatest extent of flooding, contributing to 16% of the affected area for flood depths of 2.00 meters and above. This is followed by Barangays Tubod and Ubaldo Laya, amounting 15% and 14%, respectively.



Figure 161. Flood Map of the Iligan Hlood plain showing the maximum Hlood extent and depths resulting from the 5 year rainfall event for the 2013 land cover condition



Figure 162. Estimated Extent of Flooding in Barangays in Iligan for the 5 Year Rainfall Event, 2013 Land Cover






















Figures 169 to 172 show the simulated 1lood maps for the 2050 land cover scenario. Peak discharge is at 531.6, 919.9, 1086.4 and 1255 m3 for the 5, 25, 50 and 100 year rainfall events, respectively. The lag period between the peak of rainfall and discharge is 7 hours and 40 minutes for the 5 year rainfall event and 6 hours and 50 minutes for the 25 year rainfall event. Meanwhile, it is 6 hours and 30 minutes for the 50 year rainfall event and 6 hours and 20 minutes for the 100 year rainfall event.

The 100 year return period results show that 75% of Barangay Mahayhay’s area will experience 1loods with depths of 2.00 meters and greater while it will be 71% of Barangay Ubaldo Laya. Six percent of former and 4% of the latter will experience 1loods with depths that are greater than 5.00m.
















Figures 173 to 176 show the simulated 1lood maps for the 2050 land cover scenario. Peak discharge is at 491.4, 854.1, 1009.2 and 1167 m3 for the 5, 25, 50 and 100 year rainfall events, respectively. While the lag period between the peak of rainfall and discharge is 7 hours and 50 minutes for the 5 year rainfall event and 6 hours and 50 minutes for the 25 year rainfall event. Meanwhile, it is 6 hours and 40 minutes for the 50 year rainfall event and 6 hours and 30 minutes for the 100 year rainfall event.

The 100 year return period results show lower 1lood extent than those in the 2020 land cover. Seventy-‐three percent of Barangay Mahayhay’s area will experience 1loods with depths of 2.00 meters and greater while it will be 70% of Barangay Ubaldo Laya. Five percent of former and 3% of the latter will experience 1loods with depths that are greater than 5.00m.















6.2 Combined Flood Hazard Map

Combined maps were prepared for the Cagayan de Oro and Iligan Cities. The Cagayan de Oro flood inundation map combines the flood simulation results of the Cagayan de Oro and Iponan river systems while the Mandulog and Iligan River systems for the City of Iligan.

6.2.1 Cagayan de Oro City (Cagayan and Iponan Rivers)

Figure 177. Flood Map of the Cagayan de Oro and Iponan Hlood plains showing the maximum Hlood extent and depths resulting from the 5 year rainfall event for the 2013 land cover condition



















6.2.2 Iligan City(Iligan and Mandulog River)

Figure 186. Flood Map of the Iligan and Mandulog Hlood plains showing the maximum Hlood extent and depths resulting from the 5 year rainfall event for the 2013 land cover condition






Figure 188 Flood Map of the Iligan and Mandulog Hlood plains showing the maximum Hlood extent and depths resulting from the 100 year rainfall event for the 2013 land cover condition





















6.3 Comparison of Flood Depths

A comparison of flood depths subjected to various rainfall return periods for the present (2013) condition are shown from Figure 195 to Figure 200. Of interest in these graphs are the trends in the distribution of inundation with decreasing recurrence intervals (i.e. longer return periods). In general, it can be seen that areas with shallow inundations experiencing worsened flooding.

The graphs basically show increasing area inundated with higher flood depths from 5-‐ to 100-‐year rain return period for ranges 2 to 5 and >5m for Cagayan de Oro while areas with shallower floods depths generally decrease with depth. For Iponan, the trend is similar except that the increasing inundation with return period starts to occur at 1 to 2m level.

The flooded areas in Mandulog River follow a similar trend as Cagayan de Oro River where increase in inundated areas starts to occur at 2 to 5m. For Iligan River however, the increasing trend starts even with 0.2 to 0.5m level. This indicates smooth terrain in the flooded areas inundated by the river.

Peculiar to Cagayan de Oro River is its pronounced secondary peak of flooding in the 2 to 5m levels. This is due to prominence of a deep gorge incising the River immediately downstream of the watersheds which will be flooded when a strong rainfall occurs.

Figure 195. Distribution of Hlood depths for various return periods for Cagayan de Oro River for present (2013).



Figure 196. Distribution of Hlood depths for various return periods for Iponan River for present (2013).

Figure 197. Distribution of Hlood depths from combined effects of Cagayan de Oro River and Iponan River for various return periods for Cagayan de Oro City in present condition (2013).



Figure 198. Distribution of Hlood depths for various return periods for Mandulog River for present (2013).

Figure 199. Distribution of Hlood depths for various return periods for Iligan River for present (2013).



Figure 200. Distribution of Hlood depths from combined effects of Mandulog River and Iligan River for various return periods for Iligan City in present condition (2013).

The series of figures below shows the comparison of flood depths through the different time scenarios at 5 (Figures 201 & 204), 25 (Figures 202 & 205) and 100-‐ (Figures 203 & 206) year rainfall return period. The areas flooded with less than 0.2m inundation have been removed for clarity.

In general, areas affected by 5-‐year return period, regardless of scenario considered remain unchanged expect for the deeper interval (2 to 5m). This portion typically characterizes the river bank portion of the flood plain. For the 25-‐ and 100-‐year rainfall return periods, the area affected by deeper floods increase with longer return period.

Figure 201. Comparison of of Hlood depth distribution from combined effects of Cagayan de Oro and Iponan Rivers for 5-‐year rainfall return period for 2013, 2020 and 2050

Figure 202. Comparison of Hlood depth distribution from the combined effects of Cagayan de Oro and Iponan



Rivers for 25-‐year rainfall return periods for Cagayan de Oro City for present condition (2013) and from future scenario (2020 and 2050).

Figure 203. Comparison of Hlood depth distribution from combined effects of Cagayan de Oro and Iponan Rivers for 100-‐year rainfall return periods for Cagayan de Oro City for 2013, 2020, 2050



The trend is similar for Mandulog River, though the signs less pronounced. The 5-‐year RRP follows a decreasing area in flood depth for all scenarios. The 25-‐year and 100-‐year RRP effectively shows the transition of formerly shallow flooded areas into a higher flood level in the future scenario. The small number of areas inundated with 0.5 to 1m flood levels eventually gets redistributed to the 1 to 2m levels.

Figure 204. Comparison of Hlood depth distribution from combined effects of Mandulog and Iligan Rivers for 5-‐year rainfall return periods for Iligan City for 2013, 2020, and 2050

Figure 205. Comparison of Hlood depth distribution from combined effects of Mandulog and Iligan Rivers for 25-‐year rainfall return periods for Iligan City for 2013, 2020, and 2050



Figure 206. Comparison of of Hlood depth distribution from combined effects of Mandulog and Iligan Rivers for 100-‐year rainfall return periods for Iligan City for 2013, 2020, 2050



6.5 Results of Flood Inundation Height Validation

The simulated flood models are validated using the most recent flood event, the tropical storm Sendong (international name: “Washi”) which passed through Cagayan de Oro about midnight of December 17, 2011. The recorded total one-‐day total rainfall for the period was 180mm. Observed flood heights were taken from surveys of residents which experienced the flooding. These observed flood heights were compared with the maximum flood heights that were culled out of the simulations for 100-‐year return period.

Correlation plots are shown in Figures 207 -‐ 210. Correspondence for Cagayan de Oro, Iponan and Mandulog Rivers appear to be in acceptable agreement with that of the

observed flood height values where the coefficient coefficients ( ) range from 0.32 (Iligan) to 0.63 (Iponan). Biases (+0.56, +0.90, +0.29, and +0.56) were all positive in all four locations indicating underestimation of the flood heights compared to the model values.

In the case of Cagayan de Oro River, for the most part, flood levels tend to have been underestimated, but where overestimates occurred, these were located in floods no lower than 1m in height difference. In Iponan, more overestimates are found but the average differences were only 0.03m. For Mandulog River, the average estimation error is 0.77m most of which are due to overestimation. The worst results were found in Iligan where flood heights were mostly underestimated.

The flood inundation errors analyzed should not be interpreted as fallibility of the model results especially since there were a few cases where true negatives exist. True negatives refer to areas where flooding occurred but appeared in the model result to be absent or too shallow (less 0.2m) to be mapped. On the other hand, there were far fewer false positives (where flooding resulted in simulation model but was not in reality) which while indicating unrealistic results will keep occupants safe by at least “erring in the side of caution”. Thus, the generated flood inundation maps appear to be reliable indicators of flood extent and heights.



Figure 207. Correlation of Hlood heights during Sendong event for Cagayan de Oro River (n=37).

Figure 208. Correlation of Hlood heights during Sendong event for Iponan River (n=35).



Figure 209. Correlation of Hlood heights during Sendong event for Mandulog River (n=143).

Figure 210. Correlation of Hlood heights during Sendong event for Iligan River (n=35).



6.5 Simulated Velocity Maps

Velocity maps were also simulated by the Gerris Flow Solver to illustrate the speed and direction of flood water during the occurrence of rainfall events. Flow velocities are shown as vectors in the maps, with their lengths increasing as the speed increases. The results from the simulations can be utilized for flood emergency response plans and flood deference measure. These velocity maps are shown in Figures 211 to 258.

6.5.1 Cagayan de Oro

The velocity maps for the 2013 land cover of Cagayan de Oro can be seen in Figures 211 to 214. In all of the rainfall events, Barangay 6 experiences the greatest velocities of water due to the fact that it lies directly on the river’s path. In the 5 year rainfall event, the flow of water is fastest in Barangay 6, where the depth of water is greater than 2.00 meters. A greater amount of areas in Barangays 7, 10, 13, 15 , 17 and Consolacion will experience faster velocities of water in the 25, 50 and 100 year rainfall events.

Figures 215 to 218 show the velocity maps for the 2020 land cover while Figures 219 to 222 show those of the 2050 land cover. Barangays 6, 7, 10 and 13 still experience the fastest water during the 5 year rainfall event of both land covers. The water velocity and depth drastically increase during the 25, 50 and 100 year rainfall events, affecting a greater extent of Barangays 15, 17 and Consolacion.

Water velocity increases with depth of the flood. It also increases from the results in the 2013 land cover scenario to 2020. Its extent, however, decreases from the 2020 land cover scenario to 2050.



Figure 211. Simulated flood inundation and velocity map of Cagayan de Oro River for a 5-year rainfall return period under 2013 land cover conditions.




































6.5.2 Iponan Simulated Flooding

The velocity maps for the 2013 land cover of Iponan can be seen in Figures 223 to 226. In all of the rainfall events, Barangays Canito-‐An and Patag experience the greatest velocities of water due to the fact that it lies on the river’s path. In the 5 year rainfall event, the flow of water is fastest in Barangays Canito-‐An and Patag, where the depth of water is greater than 2.00 meters. A greater amount of areas in Barangays Barra, Igpit, Bulua, Taboc and Baikingon will experience faster velocities of water in the 25, 50 and 100 year rainfall events.

Figures 227 to 230 show the velocity maps for the 2020 land cover while Figures 231 to 234 show those of the 2050 land cover. Barangay Canito-‐An and Patag still experience the fastest water during the 5 year rainfall event of both land covers. The water velocity and depth drastically increase during the 25, 50 and 100 year rainfall events, affecting a greater extent of the northern part of the Iponan flood plain.




Figure 223. Simulated flood inundation and velocity map of Iponan River for a 5-year rainfall return period under 2013 land cover conditions.

































Figure 234. Simulated flood inundation and velocity map of Iponan River for a 100-year rainfall return period under 2050 land cover conditions



6.5.3 Mandulog Velocity Maps

The velocity maps for the 2013 land cover of Mandulog can be seen in Figures 235 to 238. In all of the rainfall events, Barangay San Roque experiences the greatest velocities of water due to the fact that it lies on the river’s path. In the 5 year rainfall event, the flow of water is fastest in Barangays San Roque and Santa Filomena, where the depth of water is greater than 2.00 meters. A greater amount of areas in Barangays Santa Filomena and Hinaplanon will experience faster velocities of water in the 25, 50 and 100 year rainfall events. Barangay Santo Rosario will also experience faster velocities of water by the 50 and 100 year rainfall events.

Figures 239 to 242 show the velocity maps for the 2020 land cover while Figures 243 to 246 show those of the 2050 land cover. Barangay San Roque and Santa Filomena still experience the fastest water during the 5 year rainfall event of both land covers. The water velocity and depth drastically increase during the 25, 50 and 100 year rainfall events, affecting a greater extent of the southern part of Santa Filomena , and the upper parts of Santo Rosario and Hinaplanon.


Figure 235. Simulated flood inundation and velocity map of Mandulog River for a 5-year rainfall return period under 2013 land cover conditions.


























6.5.4 Iligan Velocity Maps

The velocity maps for the 2013 land cover of Iligan can be seen in Figures 247 to 250. In all of the rainfall events, areas near the river’s path experience the greatest velocities. Barangay Mahayhay experiences the fastest flow of water. A greater amount of areas in Barangays Ubaldo Laya, Villa Verde and Palao will experience faster velocities of water in the 25, 50 and 100 year rainfall events.

Figures 251 to 254 show the velocity maps for the 2020 land cover while Figures 255 to 258 show those of the 2050 land cover. Barangays Mahayhay, Ubaldo Laya, Villa Verde and Palao will still experience the fastest water during the 5 year rainfall event of both land covers. The water velocity and depth drastically increase during the 25, 50 and 100 year rainfall events, affecting a greater extent of the same areas.




Figure 247. Simulated flood inundation and velocity map of Iligan River for a 5-year rainfall return period under 2013 land cover conditions.




































Chapter 7DISCUSSIONS

7.1 Enhancements introduced in the Flood Hazard Maps

The flood modelling exercise resulted in differentiation in flood susceptibility levels in both inundation level and velocity flow. As can be gleaned from the summary statistics, the area inundated by different flood depth levels vary according to the scenario generated from various rainfall return periods (5-‐year, 10-‐year, 25-‐year, 50-‐year and 100-‐year) for 24-‐hour duration. These have been generalized to five level types (less than 0.1m, >0.1 to 0.2m, >0.2 to 0.5m, 0.5m to 1.0m, >1.0 to 2.0, >2.0 to 5.0m and >5.0) are clearly indicated in the maps and graphs. This is a stark departure from the current flood hazard maps only indicate different degrees of susceptibility.

The fine-‐scale flood hazard maps can also be used as a guide in improving local drainage. Areas that are often flooded due to heavy rainfall events which are not within the reach of river can now be identified. Ordinary flood models show the expansion or contraction of river carriage according to discharge from the watershed and not those caused by surface flooding. Because 2D models were used in this study, it was able to show areas flooded not only by overflowing rivers, but also from storms assuming an overwhelmed drainage system. Provision for urban storm drainage is the responsibility of the city government. Faced with budget constraints and other priorities, urban drainages such as culverts, designed to drain 10-‐year storms. The city government can use the flood hazard maps with lower frequency storms to design and implement drainage systems with higher capacity.

Probabilities of occurrence can now be inferred based on the rainfall scenarios simulated. The generation of probability-‐based hazards will be useful in the different facets of flood mitigation ranging from preparation to long-‐term mitigation of its effects. Based on probabilities, planners can recommend plans and identify flood zones that will be off-‐limits to settlements but may be used for select agricultural land uses or for recreation. Property developers and investors can then follow the zoning regulations by developing real estate in areas in areas supine to flood or may opt to develop in flood-‐prone areas provided adequate flood defenses are properly implemented. Engineers can present designs of intervention measures such as dikes, levees and dams or to fortify bridge and roads that can withstand floods according to the level of protection they need or can afford. Those living in high-‐hazard areas may choose to stay with compulsory retroffiting such as stilting or non-‐wall ground levels of their buildings or avail of risk transfer commodities such as insurance.

The different maps depicting the flood inundation and velocity determine the courses of action that can be used by decision-‐makers, planners and engineers can take to prepare and respond to flooding disaster. spatial planning/zoning and infrastructure. Ultimately, the spectrum of decisions in flood management can range from which areas to protect, what to adapt to and what to give up.

Based on the projected changes in rainfall pattern from the simulation by PAGASA, the flooding scenarios in the future were also generated, particularly for 2020 and for 2050. The effect of changes in rainfall amount within 24-‐hour period was incorporated in the runoff models. The results of the models that the water levels in the rivers will increase by 2020. However, in 2050, when the 24-‐hour duration nears the present (but still higher



than) the present levels.

The rainfall dumped by typhoon Sendong levels was recorded at 165mm in 24 hours (Talakag). Frequency analysis against historical rainfall data place Sendong between a 35-‐year (log-‐normal distribution) to 75-‐year (exponential distribution) return period. With the projected increase in daily rainfall as a result of the climate change projection, the Sendong event is considered to now occur more frequently in the area. Plans and designs for storm frequencies attuned Sendong-‐type events should now be a norm.

The modelling exercises for the four watersheds also resulted in the determination of the possible lead time for local disaster managers to take action provided rainfall amount in the upper part of the catchment is known. Table 17 to Table 32 indicate peak runoff and the lag time between the peak rainfall and runoff for the four rivers (Cagayan de Oro, Iponan, Iligan and Mandulog) . These can be used as guide in evacuating citizens to safe areas. Specific rainfall amounts translate to approximate flooding scenarios, the local disaster managers can prioritize whom, where and how to evacuate since the flood hazard maps indicate the rainfall.

7.2 Factors Aggravating the Flooding Problem

During the course of the flooding study, various aspects of the Cagayan de Oro watershed pointed out the factors that may influence flooding and provide keys to proper approach to manage the flooding issue. The different aspects and insights are discussed below. Although the floodings are mainly triggered by intense and prolonged rainfall through a fixed catchment size and geometry, the degree of flooding may be modified by the conditions in the watershed. The land cover changes in the watershed land cover conditions in the watershed lead to worse flooding either directly or indirectly.

7.2.1 Changes in land use/land cover conditions

Changing the land cover from forest, idle or grassland to agricultural plantations was observed from the BAS statistics (See Figure 47) directly increases runoff because of the watersheds decrease in the ability to retain water. The ability to retain water by the watershed depends on the quality and quantity of vegetative cover and the type of soil. The water-‐retention of vegetation is possible through interception of rainfall by the canopies. The thicker and higher the vegetation type, the larger the volume retained which flows down the canopy at a later time. Direct raindrops or flowing water below the vegetation flow through the soil as infiltration until such time that soil is saturated with water. On the other hand, compaction of bare soil surface material by direct exposure to sunlight decreases ability of water to directly penetrate canopy and hence induces immediate overland flow. The presence of vegetative cover therefore causes delay in the flow of water through the underlying soil cover and hence reduce and prolong the arrival of peak runoff.

Shifting the vegetative cover from forests to agriculture significantly decreases interception capacity and accelerates water infiltration by soil. By converting idle grasslands to tilled agricultural lands also reduce interception to a lesser extent and accelerates infiltration rates due to tillage. There is almost zero interception and no infiltration on impermeable built surfaces such as concrete or asphalt.



7.2.2 Sedimentation and Flooding

Changes in vegetative cover may also aggravate flooding by way of erosional and sedimentation processes. Eroded materials from the soil surface within the watershed are transported and eventually deposited along the channels. It was shown in the Chapter of this report that the land use in the upper part of the catchment has been actively changing possibly due to shifting to a) intensive agricultural land uses; b) land slides and bank erosion which remove topsoil and deposit along the channels; and c) mass movements and debris flows from slope failure resulting from other induced/man-‐made or natural causes. The soils eroded from agricultural areas are transported in the streams and river tributaries (Figure 259-‐260) especially during the rainy season when heavy rainfall events actively remove on rain impact and eventually deposit and remobilize the sediments from overland flow into the channels.

One of the immediately identified solutions to the flooding problem in the PCTP sites is dredging which entails the physical removal of sediments along the river channel to increase its flow capacity. Removal of sediments on strategic parts of the river aims either widen and/or deepen the river or make the channel flow more efficiently. Before any dredging activity is pursued, it is important to understand the characteristics of sedimentation in the upper part of the Cagayan de Oro watershed.

Figure 259. Photo of Cagayan de Oro River merging (foreground) tributary taken from Bubunawan station, Bubunawan. Bukidnon upstream.

Left tributary shows turbid waters coming from an area dominated by agricultural and uses, particularly pineapple (upper left) planted in an extensive plateau in nearby town of Bukidnon. This is in contrast to the right side which show relatively pristine waters (left).



Figure 260. View of San Simon Bridge along Iponan River.

Notice the turbid waters 1lowing in the river (photo taken 09 June 2013). A water level and rain gauge (top) is installed on the bridge.

As the channels deepen and widen downstream, the stream power of the river waters allow sediments to settle along the channels as wash loads or bed loads. The deposition of sediments in main channels, in turn, makes it more shallow and narrower, thereby reducing the overall capacity of the river to flush out water and also clog some of the waterways.

To counter the decreasing river capacity due to sedimentation, dredging is one of the immediate interventions proposed. But since there will be continuous supply of sediments one rainy season after the other, dredging activities must also be done seasonally in order to remove them. Dredging is not a one-‐time solution -‐ it entails much resource in terms of heavy equipment to remove, logistics transport the materials, and proper planning to strategically place dredge materials.

Although the amount of soils that has been removed is not quantified in this study, several inferences may be devised to estimate the sediment load. Based on the DENR-‐EMB Region 10 regular monitoring from 2011 to 2012, the sediment concentration in terms of Total Suspended Solids (TSS), range from <1 to 206mg/L. The River Basin Control office estimates the annual runoff from Cagayan de Oro River at 3,883 Million Cubic Meters (MCM) per year. Using the TSS ranges, the sediment load could be as much as to 800,000 tonnes per



year1. At bulk density of about 2.7g/CM this weight will translate to 296.296 MCM of sediment volume. Assuming that that a massive dredging activity is undertaken similar to the magnitude of the Pasig River dredging project, where the daily dredging capacity is 14,000 CM/day, it will take more than 100 years to dredge. This Pasig River Project entailed Php4.5 Billion. The dredging materials out of the main river channel will only be effective on the shorter term.

Figure 261. Paseo de Oro high-‐end shopping and hotel complex in front of Cagayan de Oro River.

Note the very turbid water gently running through the river.

7.2.3 Urban Development Aspects of Flooding

Development of urban spaces should carefully consider the magnitude of flooding that may occur based on the different scenarios depicted in the flood hazard maps. Unplanned urban development does not only increase risks to those dwelling on them, they also worsen the flooding condition about their surroundings. Although structural measures such as dikes Figure 261 are often proposed to control the flood, they may also be proven inadequate and neglected through time. Since concentration of people will put them in higher level of risk than necessary, real estate development should there be veered away from the rivers as much as possible (Figure 262).



1 In the latest MGB sediment analysis performed in Iponan River (MGB 2014), the sediment concentration range from 31 to 872 mg/L.

Figure 262. Sky view of Paseo del Rio right beside Cagayan de Oro River

Sitting on what ought to be a natural 1loodplain, the building-‐complex, though claimed to be adaptive to 1lood-‐prone conditions by being supported on stilts, will have fencs impede incoming 1lood waters and eventually defeat the purpose of having pier piles in the approach Cagayanon Bridge from the rotunda (foreground, right). Notice the concrete fences built on the left side of the complex. (Photo courtesy of CdoDev.com).

The flood hazards are also useful in diagnosing vulnerabilities of the transportation network to flooding. Roads that have elevations are below the highest flood levels will be cut off during flooding events. Flooding in roads will not only affect those living nearby, but also those located upstream in the watersheds because it will cut off the supply of goods and service to those in the floodplain. Therefore, economic activities even by those outside the flood-‐prone areas may be also be affected.



Chapter 8CONCLUDING REMARKS

8.1 Summary

Flood modeling and hazard mapping has been undertaken for four major river systems found in the Cagayan de Oro and Iligan, namely Cagayan de Oro River, Iponan River, Mandulog River and Iligan River. The flood model employed a 2-‐dimensional numerical simulation of shallow water flow. The inflows used in the flood models were based on flows estimated from each of the equivalent watersheds for scenarios of 5-‐, 25-‐, 50 and 100-‐year return periods. The flood model was routed over a high-‐resolution topographic data from recent fine-‐scale LIDAR surveys. The flood models and maps were validated using flood levels observed by local folks. The flood model results were processed to extract the maximum inundation depth and used as a basis for producing the flood hazard map. Estimates of flooding extent and inundation level were generated. Results show the flood maps against local observations indicate valid and reasonable levels of accuracy with minimal cases of spurious results (false positive or true negative results).

8.2 Recommendations

The problems of flooding in cities of Cagayan de Oro and Iligan should not be dealt with in isolation. Those concerned LGUS may not be able to address the river flooding issues alone since the watersheds where most of the floodwaters come from are found in other municipalities, cities and provinces.The source of excessive waters does not originate within these areas alone. The rivers of Cagayan de Oro and Iponan watersheds not only transverse Cagayan de Oro City but also the municipalities of Talakag, Baungon, Libona and Pangantucan found the province of Bukidnon.

In the case of Iligan City, the Mandulog and Iligan watersheds also include parts of Kapai, Bubong Tagoloan II, Pantar and Baloi. Clearly, the solution of the flooding woes, it is imperative to involve those beyond one’s administrative area of responsibility. A basin-‐wide approach to planning the land use and development strategies must therefore be adapted.

Much is also at stake for LGUs located within these watersheds. Many of the watershed residents depend on the cities of Cagayan de Oro and Iligan for trade of their agricultural produces, services and supply of their basic needs and commodities. If the flow of these goods and services are interrupted by increasing flooding events and magnitudes, they would also be economically affected. The symbiotic relationship between watershed and flood plain communities must be therefore be recognized and addressed when attempts to address the flooding issue.

The increased and accelerated watershed runoff is also symptoms of alarming rates environmental degradation taking place slowly in the upland areas. Increased runoff does not only cause sedimentation, shallowed rivers and consequently flooding in the lowland areas but also leads to badlands.

Intense agricultural activities, deforestation and unregulated mining are possible culprits to the increased erosion rates. Continuously cultivated upland areas are susceptible to erosion.



The eroded topsoils supply sediments to the streams and unto the rivers. Sediment-‐laden rivers decrease their depth and reduce flow efficiency. Since the rivers become shallow, excess runoff could easily overtop river channel and therefore exacerbate flooding.

On the flood plain areas, urban development along riverbanks should be discouraged because it impedes flow of water through the floodplain and aggravates flooding around the river. Large urban development projects fronting or located just beside the river can constrict river flows. The constriction in turn makes it difficult for water to flush out from the upper adjacent part of the floodplain or across the river making these areas more prone to higher and prolonged levels of inundation. In general flood plains and river banks are better left in their natural state to perform its function of serving as impoundment or carries of excess waters overflowing from the channels during flooding episodes. During the dry season, they may be enjoyed for recreational uses as parks and open fields. To some extent, agricultural uses may be allowed but may be planted with produces of greater flood resistance like mango orchards or coconut trees.

Therefore, aside from examining the climate change, anthropogenic events and activities must be taken fully into account in understanding the complex hydrologic processes taking places in the basins being studied. An integrated analysis of climatic and land cover conditions in future studies is recommended.

8.3 Concluding Remarks

The flood hazard mapping exercise was able to define in greater detail the flooding susceptibility of the four river systems. The newly generated spatial information serves as a valuable input for land use planning and zoning since it identifies decision zones where settlement and development activities must be regulated.

The flood scenarios also guides investment decisions such as what flood and drainage control facilities must be put in place, and how will they be designed to withstand future impacts of climate change. It will also guide priorities for watershed management by influencing policies on managed production while and protecting the upland areas. The flood inundation model which simulates the velocity of water flow will also guide the integrated flood monitoring and early warning system that government intends to put in place. By being able to pinpoint areas that are of highest risk to future flood events, local governments will be more equipped in prioritizing their programs and projects that will address the vulnerability of communities therein, come up with better preparedness programs, and ensure the safety of everyone.



REFERENCES

Mines and Geosciences Bureau (2014), Decreasing sediments in Iponan River, accessed: 24 March 2014, URL: http://www.mgb10.com/mgb10/wp-‐content/uploads/2014/03/Decreasing-‐Sediments-‐in-‐Iponan-‐River.pdf

Paringit, E. C. (2012), High-‐resolution digital elevation dataset derived from airborne lidar for flood hazard assessment and mapping applications. Proceedings of the 33rd Asian Conference on Remote Sensing (ACRS), Phuket, Thailand (in DVD).

Popinet, Stéphane (2003), Gerris: a tree-‐based adaptive solver for the incompressible Euler equations in complex geometries, Journal of Computational Physics, 190(2), pp. 572-‐600.

Popinet, S. (2012) Adaptive modelling of long-‐distance wave propagation and fine-‐scale flooding during the Tohoku tsunami, Natural Hazards and Earth System Sciences (12)1213–1227.

Hydrologic Engineering Center (2010) HEC-‐RAS (Version 4.1) River Analysis System, User’s Manual, U.S. Army Corps of Engineers, Davis, CA.

Kelman, Ilan, Spence, Robin (2004), An overview of flood actions on buildings, Engineering Geology 73, 297–309)



http://www.mgb10.com/mgb10/wp-content/uploads/2014/03/Decreasing-Sediments-in-Iponan-River.pdf

http://www.mgb10.com/mgb10/wp-content/uploads/2014/03/Decreasing-Sediments-in-Iponan-River.pdf



Technical Report: River basin and Flood Modeling and Flood ...

Documents