Monitoring Small Reservoirs in the Volta Basin of Ghana

A Partner of

Monitoring Small Reservoirs in the Volta Basin of Ghana

Frank Annor, Nick van de Giesen, Tom Bogaard & Dirk Eilander([email protected])

CPWF Scientific Workshop , 17-19 September 2013, Ouagadougou, Burkina Faso

Andes • Ganges • Limpopo • Mekong • Nile • Volta

Content1. Main message & Background2. Methodology3. Preliminary results & Discussions4. Outlook

MAIN MESSAGE & BACKGROUND

Andes • Ganges • Limpopo • Mekong • Nile • Volta

Main message

© pubs.usgs.gov

12mm/day

9mm/day

1.4mm/dayend of June

Feb. - March

A Partner of

5

The Volta River System

≈400,000 Km2

UER = 8,570 km2

RF 500 (N)-1100 (S)mm/yr

Andes • Ganges • Limpopo • Mekong • Nile • Volta

How can we bridge the poverty gap?

Andes • Ganges • Limpopo • Mekong • Nile • Volta

Small Reservoirs – multiple usesFarming Rain-fed – not enough (1970s -1980s

Drought!)

Andes • Ganges • Limpopo • Mekong • Nile • Volta

What are the issues with SR?

Lack of adequate storage facilities (design & capacity problems!) & productive use of existing facilities

Large (?) Evaporation losses

Sedimentation not really a problem – 3400t/year

Andes • Ganges • Limpopo • Mekong • Nile • Volta

Monitoring network

German

yBenin

Burkina F

aso

Ghana

Ivory

Coast

MaliTogo

Nigeria

South

Africa

0

10,000

20,000

30,000

40,000

50,000

60,000

70,000

1,244

18,78024,927

11,359

23,033

62,010

6,309

27,993

7,052

Area per WMO Station in km²

Data: Deutscher Wetterdienst (2008)

Monitoring network - TAHMO

METHODS

A Partner of

Methodology

Satellite & GPS

Ground Measurements

Reservoir Masks

BEAMILWIS, ERDAS WS + ECS Isotopes

variables

Evap – EBM(heat fluxes, Ef) Evap - ECS Evap - WB

Satellite Imagery

Retrieval of

variables

A Partner of

Monitoring - SatelliteEvaporation

SN Satellite DataSpatial

resolution

Geo-biophysical Variables to be

retrieved DatesData

availability Source

1MERIS 260m

LST & NDVI, Emissivity, Albedo

Mar-11 to Mar 2012 √ ESA

2AATSR 1km

LST & NDVI, Emissivity

Mar-11 to Mar 2012 √ ESA

3Landsat 7 ETM+ (8) 30m, 60m

LAI, LST, NDVI, Emissivity, Albedo Mar-07 to Mar 2012; 2013 √X ESA/USGS

4ASTER 15, 30,90mLAI, LST, EVI, Emissivity, Albedo Mar-11 to Mar 2012 √ USGS (NASA)

5Sentinel - 2 20m (?)LAI, LST, NDVI, Emissivity ? X ESA

Water Mask and Water Monitoring

1Landsat 30m Mar-92 to Mar 2012 √ ESA/USGS2Radarsat-2 < 7m Nov-12 to Present √ CSA3ENVISAT (ASAR) 30m Mar-07 to Mar 2012 √ ESA4ALOS 7m ? √ JAXA5SPOT 10m ? √ ESA

Andes • Ganges • Limpopo • Mekong • Nile • Volta

Issues - RS Evaporation estimatesL - Avg. RMSE of 50Wm-2 & 15 – 30% REPhysical and Analytical methods are not better than empirical and statistical approachesEmpirical methods are limited by in-situ training dataDiurnal variation of evaporation vrs instantaneous measurements with RS (Bowen Ration & evaporative fraction)Temporal & Spatial Scaling issues (extrapolation, gap filling & data fusion) + cloudsSurface energy models (SEBAL, SEBS etc.)

Kalma et al. (2008) & Farah et al. (2004)

Andes • Ganges • Limpopo • Mekong • Nile • Volta

Base Map & Radar monitoring

Landsat SPOT

ASAR, ALOS & Radarsat-2

Radar Monitoring

Optical – base map

Growing MLC - automation

NIR & visible wavelengths Intensity - Roughness & Polarisation

Fieldwork

1-3 3-5 5-7 7-9 9-11 11-13

13-15

15-17

17-19

>1902468

1012

Area [ha]

FieldworkSR154

SR120

RADARSAT-2 SAR imagerySR154, Pauli decomposition

RGB Composition:

R: Single BounceG: Volume Scattering B: Double BounceRGB composition RGB comp [3x3 window]

Single Bounce

A

Double bounce

C

Volume Scattering

B

Andes • Ganges • Limpopo • Mekong • Nile • Volta

One temporal prior for previous and two for subsequent classifications. Subsequent is dependent on rain / dry season.

Based on 3 pix moving average

Growing way of filtering unclassified pixels/ removing water and land ‘islands’

SR Delineation

WOIS

20

Storage - Bathymetry

0

1

2

3

4

5

6

0

10,000

20,000

30,000

40,000

50,000

Dep

th (m

)

Are

a (m

2 )

Volume (m3)

Area Depth

Volume = 0.00857 x Area1.4367 m3

Andes • Ganges • Limpopo • Mekong • Nile • Volta

Time Series of SR – Mass balance

21

Small Reservoir Storage (S)

Seepage

EvaporationRainfall

IrrigationRunoff

≈ 0 ?

Onset of the dry season - Nov

PRELIMINARY RESULTS & DISCUSSIONS

Andes • Ganges • Limpopo • Mekong • Nile • Volta

Time series of SR …1

Andes • Ganges • Limpopo • Mekong • Nile • Volta

Time series of SR …2

26 Reservoirs

Weather Station @ Binaba (Telemetry)

Measurements (ECS + WS + L) Instrumentation

41230 41238 41246 41254 41262 41270 41278 41286 41294 41302 41310 41318 41326 41334 41342-160

-140

-120

-100

-80

-60

-40

-20

0

20

MW_Upstream MW_MidstreamMW_Downstream Lake

Water Level Fluctuations

Andes • Ganges • Limpopo • Mekong • Nile • Volta

Water Balance of SR

27

Isotopes

Evaporation

11/11/2

012

11/18/2

012

11/25/2

012

12/2/2

012

12/9/2

012

12/16/2

012

12/23/2

012

12/30/2

012

1/6/2

013

1/13/2

013

1/20/2

013

1/27/2

013

2/3/2

013

2/10/2

013

2/17/2

013

2/24/2

013

3/3/2

013

3/10/2

013

3/17/2

013

3/24/2

013

3/31/2

013

4/7/2

013

4/14/2

013

4/21/2

013

4/28/2

013

5/5/2

013

5/12/2

013

5/19/2

013

5/26/2

013

6/2/2

013

6/9/2

013

6/16/2

013

6/23/2

013

6/30/2

0130

5

10

15

20

25

30

35

40

0

1

2

3

4Evaporation Temperature (atm) Wind Speed

Date

Evap

orati

on (

mm

/day

)Te

mpe

ratu

re (o

C)

Win

d Sp

eed

(m/s

)

Avg - 9mm/dayMin - 1.4mm/dayMax - 12mm/day

OUTLOOK

Andes • Ganges • Limpopo • Mekong • Nile • Volta

Outlook

Evaporation estimates (RS)Comparison with Flux measurementsScale effects & boundary conditions of evaporation from SRDesign of reservoirs for optimum storage in dry season?

Andes • Ganges • Limpopo • Mekong • Nile • Volta

Thank you!

www.smallreservoirs.org