Focal Statistic Modeling in order to Define Flood Interest in Hungary · 2018-01-08 · Focal Statistic Modeling in order to Define Flood Interest in Hungary Arif Rahmansyah DARANA1

Focal Statistic Modeling in order to Define Flood Interest in Hungary

Arif Rahmansyah DARANA1

1 Hydrogeological Engineering Program, Faculty of Earth Science and Engineering

University of Miskolc, Egyetemvaros, Miskolc, Hungary, 3515

*Corresponding author: [email protected]

Abstract

Based on the report of The European Environment Agency (2016) Hungary has approximately 18%

of the population, 1.8 million people, living on the flood-prone areas. That numbers refers to the one

of the most exposed country to suffer from flood in Central Europe. In facts, the losses could increase

up to fivefold by 2050. However, several techniques are existed to analyze the threat of the flood. One

of the techniques is remote sensing. That technique is one of the effective and efficient way to conduct

the analyses of the flood especially in big area study or in the low resolution, and it able to find the

flood interest. This study uses the ASTER image to generate the DEM, then the modeling of river and

geomorphological features are generated from it. The focal statistic modeling is run to locate the

interest of the flood prone area with drainage density and stream order as the approachming. The result

shows that from the middle to south east of Hungary, especially in Jász-Nagykun-Szolnok county,

Békés county and Bács-Kiskun county are the highest interest area for flood to occur, even the capital

city, Budapest, is one of the prone area. This method is still preliminary study; hence the advance

study is needed to make the better risk analyses.

Keywords: Flood, Hungary, Focal statistic Modeling, Remote Sensing, Natural Hazard

1. INTRODUCTION

Flood is one of destructive natural hazard. That event can cause abundant loss of life or property (Keller,

2011 P116). Flood is a disaster relating to hydro-meteorological event, thus it is included in the

hydrologic cycle. Flood is a big amount of runoff, which is a transport from land to ocean. Runoff water

will flow following the course determined by local topography and also it will infiltrate to the soils or

rocks. However due to the difference capability of earth materials the water just flows on surface to

transporting itself. In this case on the lowland area the runoff will be gather from the higher ground

surrounding (Keller,2011 P137).

Moreover, Individual flood waves emerge after prolonged high-intensity (up to a maximum of 260 mm

within 24 h) rainfall from local thunderstorms or following rapid temperature rises which cause

snowmelt over larger areas. Furthermore, the previous study (Dénes et al, 2009) also identified the

causes of destructive natural flood as:

1. Flood waves travelling down tributaries maybe often superimposed upon the flood of the main

river (Szlávik, 2001 in Dénes et al, 2009)..

2. Ice-jam floods (Kovács, 1987 in Dénes et al, 2009) are also common in this area, because the

permafrost blocks the water to infiltrate to subsurface.

3. Floods from impounding include the effect of a tributary joining a low-gradient trunk river

(Szlávik, 2001 in Dénes et al, 2009).

The 2nd Join Conference of Utsunomiya University and Universitas Padjadjaran, Nov.24,2017

122

Fig 1. Regional Map of Hungary with DEM (A) (Tamas et al, 2015), Geological Map of Hungary Surface

Lithology1 : 1,000,000 (EGDI, 2016).

Hungary is among the countries in Europe most severely endangered by floods. Fourteen rivers are

flowing in Hungary from the mountainous area of the Carpathian Basin which have high channel slope

gradients which serve as a source of considerable flood hazard (István, 2010). Hungary (Fig. 1) is placed

on the most lowland area with most of the lithology is floodplain sediment which contain clastic

sediment with wide range of particle sizes with muddy matrix (Table 1). The elevation difference in the

Great Hungarian Plain (GHP) is about 100 m, while its horizontal extent reaches 300 km. The elevation

differences of terrace levels within the Pleistocene–Holocene flood-plains are less than 20 m and the

width of these smooth plains is more than 10 km (Gábor and Gyula, 2008).

The report of floodlist (2013) the flood was occurred in June in Budapest while more than 1200 people

are forced to evacuate. Moreover, according to report of European Environment Agency about flooding

in Europe, the researchers have predicted the loses caused by flood will have increased fivefold, with

Hungary as one of the most exposed area due to approximately 1.8 of population live on flood plain

50 km

Nort

h Scale:

1 : 1,000,000

A

B

The 2nd Join Conference of Utsunomiya University and Universitas Padjadjaran, Nov.24,2017

123

area (Floodlist, 2016). Therefore, the study about flood is a must in order to reduce the risk, thus a flood

modeling to define the flood spot as a preliminary study is essential.

Table 1. Lithology Explanation of Geological Map of Hungary (EGDI,2016).

Legend Age Lithology

Holocene Epoch

Unsorted or poorly sorted, clastic sediment with a wide range of

particle sizes, including a muddy matrix.

Holocene Epoch Moisture-free basisorganic rich sediment.

Tortonian Age Mudstone

Aquitanian Age Diamictite

Carboniferous Period Granite

Miocene Epoch Tuff-breccia, agglomerate, or pyroclastic

Miocene Epoch Andesite-basalt

Cambrian Period Slate

Oligocene Epoch Impure dolomite

Triassic Period Limestone and dolomitic limestone.

2. METHODOLOGY

ASTER GDEM (2011) was used as a main data in this research. Through ArcMap 10.3 the modeling

was conducted to generate the flood interest map (Fig. 2).

Fig 2. Workflow Diagram of Applied Method.

ASTER GDEM (2011) was used as a main data in this research. Through ArcMap 10.3 the modeling

is conducted to generate the flood interest map (Fig. 2). The Aster images was exported to ArcMap

10.3. All of Images had to be processed by uniting all of them using “Mosaic to New Raster” tools then

The 2nd Join Conference of Utsunomiya University and Universitas Padjadjaran, Nov.24,2017

124

crop it with the study area border which in this case is Hungary country border by using “Clip Mask”

tools. The depression in the sink filled then, the flow direction and flow accumulation was generated

using hydrogeological tools. The conditional modeling of flow accumulation was run in order to

classified the stream Network. Next as the previous researcher (Ejikeme et al, 2015) conducted, the

focal statistic, was conducted and reclassified to different statement levels of flood.

The reclassified data were integrated with hill shade, the processed data from filled DEM, and also with

base map of Hungary (arivaIO, 2016 and openstreetmap.org, 2014). Hence the Flood Interest Map of

Flood in Hungary was created.

3. RESULT AND DISCUSSION

The Result is Shown on the Flood Interest Map of Hungary (Fig. 3). The flood interest is generated

from the flow direction which was continued by flow accumulation. It means the software generated

the place where the water could be accumulated, as a “stream network”, however it can be different

from river network, the focal statistic used it as an input. This modeling was classified using Jenks

natural break classification, because this statistical classification system determined the best

arrangement of values into different classes by minimizing each class’s average deviation from the class

mean, while maximizing each class’s deviation from the means of the other groups. In other words, the

method seeks to reduce the variance within classes and maximize the variance between classes (Jenks,

1967)

The focal statistic model is interpreted using drainage density and stream order from the morphometry

study of Purwanto (2013). The highest value of “stream network” means the highest stream order too.

Furthermore, the highest exposure of focal statistic modeling, signed by red color, the more streams are

close to each other which means higher drainage density.

Hence, the result of this model is defined into three categories of flood interest, there are

High Interest to flood

It has fast increasing flood water table, with low decreasing water table. The water flows

through the low resistant rocks, thus the transported sediment will be plentiful.

Moderate Interest to flood

It has moderate increasing and decreasing flood water table. The water flows through the

higher resistant rocks, thus the transported sediment is not too abundant

Low Interest to flood

It has fast increasing flood water table, with fast decreasing flood water table. The water flows

through hard rock, thus the transported sediment is just in little amount.

From the map (Fig 3), it can be seen that even in the heart of Budapest the high interest of flood is

placed on it, thus it will need extra treatment since it is a capital city of Hungary. Furthermore, the most

probable place to occur the flood is in Jász-Nagykun-Szolnok county, Békés county and Bács-Kiskun

county. Those two county are placed on very lowland area with or poorly sorted, clastic sediment with

a wide range of particle sizes, including a muddy matrix covering almost a whole area.

The 2nd Join Conference of Utsunomiya University and Universitas Padjadjaran, Nov.24,2017

125

Fig

3. F

loo

d I

nte

rest

Map

of

Hu

ng

ary

The 2nd Join Conference of Utsunomiya University and Universitas Padjadjaran, Nov.24,2017

126

4. CONCLUSION

All and all More than 60% of Hungary is lying on lowland and the major material is unsorted or poorly

sorted, clastic sediment with a wide range of particle sizes, including a muddy matrix. By interpreted

the model with the geomorphological aspect the interests of the flood are divided into high interest to

flood, moderate interest to flood, and low interest to flood. Furthermore, Jász-Nagykun-Szolnok county,

Békés county and Bács-Kiskun county has the most spot of highest interest to flood.

5. RECOMMENDATION FOR FUTURE STUDY

This Model is able to support the preliminary study of flood risk assessment in order to define the

probable spot of flood. However, further study is needed. For further study the high resolution DEM

will give better result, and the better quality data will give the better results, thus the area of the flood

interest can be defined, and more accurate flood interest can be generated too.

ACKNOWLEDGEMENT

Thanks to Prof. Dr. Szűcs Péter as the dean of Faculty of Earth Science and Engineering University

of Miskolc for providing and supporting this research with some references in English, and also for

the moral support.

REFERENCES

ArivaIO. 2016. Hungarian administrative borders for download as shape-files (SHP).

http://ariva.io/gis/ Accessed on 15th November 2017 at 17.00 CET.

Dénes Lóczy, Éva Kis, Ferenc Schweitzer. 2009. Local flood hazards assessed from channel

morphometry along the Tisza River in Hungary. Geomorphology: 113. Elsevier (P 200-209).

EGDI. 2016. EGDI 1:1 Million OneGeology pan-european Surface Geology, harvested from

INSPIRE conformant National WFS services on GeologicUnit.

http://egdi.brgm.fr/egdi/#baslay=baseMapEEA,baseMapGEUS&optlay=&extent=4391830,2041

210,4912000,2561930&layers=onegeoeuro_surface_lithology Accessed on 15th November 2017

at 19.00 CET.

Ejikeme, J.O, Igbokwe, J.I, Ojiako, J.C, Emengini, E.J, and Aweh, D.S. 2015. Modelling The Impact

of Flooding Using Geographic Information System and Remote Sensing. Volume 3 Issue 4.

International Journal of Technical Research and Applications (P 67-72).

Floodlist. 2013. Hungary Flood Update. http://floodlist.com/europe/hungary-flood-update. Accessed

on 14th October 2017 at 20.00 CET.

Floodlist. 2016. Hungary Expects Danube Floods. http://floodlist.com/europe/report-floods-europe-

increase-fivefold-2050. 14th October 2017 at 20.10 CET.

Gábor Timá, Gyula Gábris. 2008. Estimation of water conductivity of the natural flood channels on the

Tisza flood-plain, the Great Hungarian Plain. Geomorphology 98. Elsevier (250-261).

István Nagy, Ferenc Ligetvári and Ferenc Schweitzer. 2010. Tisza River Valley: future prospects.

Hungarian Geographical Bulletin 59 (4). (P361-370).

Jenks, George F. 1967. The Data Model Concept in Statistical Mapping, International Yearbook of

Cartography 7. (P 186–190).

NASA and METI. 2011. ASTER GDEM V2. https://earthexplorer.usgs.gov/. Accessed on 15th

November at 17.00.

Openstreetmap. 2014 Basemap of Hungary. http://www.openstreetmap.org/#map=8/47.182/19.506.

Accessed on 15th November 2017 at 19.00.

Purwanto, T.H. 2013. Ekstraksi Morfometri Daerah Aliran Sungai Dari Data Digital Surface Model

(Studi Kasus Das Opak). UGM. https://geo.ugm.ac.id/main/?p=2418 Accessed on 15th November

2017 at 21.00 (in Bahasa)

The 2nd Join Conference of Utsunomiya University and Universitas Padjadjaran, Nov.24,2017

127

TamasTamás BUDAI, Gábor, CSILLAG Zsolt KERCSMÁR, Ildikó SELMECZI, and Orsolya

SZTANÓ. 2015. Surface geology of Hungary. Explanatory notes to the Geological map of

Hungary (1:500 000). Geological and Geophysical Institute of Hungary

The 2nd Join Conference of Utsunomiya University and Universitas Padjadjaran, Nov.24,2017

128