IWRA Proceedings

< Return to abstract list

Hydrologic Partitioning And Vegetation Response In Selected Moist Zone Catchments Of Ethiopia: Analyzing Spatiotemporal Variability

World Water Congress 2015 Edinburgh Scotland
3. Management of water resources
Author(s): Fasil Worku
Guiling Wang

ETH Zurich1, University of Connecticut2

Keyword(s): Sub-theme 3: Hydrology,
Article: Oral:



Analyses relating spatiotemporal variability of hydrologic budget with that of catchment-vegetation is one of the avenues to explore vegetation response to climate variability and the resulting impact on water resources dynamics. Here the spatiotemporal variations of hydrologic budget and Normalized Difference Vegetation Index (NDVI) in six moist zone catchments of Ethiopia during 2000-2006 were analyzed and their relationship explored. It was found that the fraction of precipitation potentially available to catchment-vegetation (Wetting; W) ranged from 0.73 to 0.96, meaning up to 27% of precipitation was not available to vegetations. Horton Index (HI a.k.a. vegetation water use factor) ranged from 0.42 to 0.92, leaving up to 58% of wetting not consumed by vegetation in some catchments, and in others this unused fraction was as low as 8%. Although HI showed strong inter-catchment variation, it was relatively constant from year-to-year and can be considered a catchment characteristic. However, HI alone is not sufficient to indicate whether vegetation growth is limited by moisture availability, as NDVI of 0.77 was observed in a catchment with lowest HI and NDVI of 0.56 was observed in another catchment with highest HI. Our results demonstrate that catchments within the same climate zone exhibit variable hydrologic partitioning and vegetation response behavior.


1.1. Background

Hydrologic budget and vegetations are closely coupled. Hence, identifying hydrologic budget and vegetation link is one means of exploring vegetation response to climate variability and the resulting impact on catchments water resources (Brutsaert, 1988; Thompson et al., 2011; Zhou et al., 2003). Hence, improving our knowledge of hydrologic budget and vegetation coupling is essential to precisely predict the impacts of climate variability and to better manage catchments. However, understanding this coupling is challenging due to lack of data, complexity of catchment processes and lack of appropriate mechanism to scale information at patch level up to catchments (Sivapalan et al., 2011). In this paper, we quantify hydrologic budget using simple water balance theory and derive NDVI from satellite imagery in an attempt to explore hydro-ecologic coupling and examine their spatiotemporal variability in selected catchments found in the mostly moist climate regime of Ethiopia.

1.2. Objective

To quantify catchment-scale hydrologic budget and vegetation greenness index, explore their correlation, and analyze their spatiotemporal variability in catchments selected from mostly moist zone of Ethiopia

1.3. Importance of the study

This study is important in showing that issues like catchment management and the impact of climate variability on catchment-vegetation should be treated differently regardless of similarity in climate regime of catchments.

2. Methods/Materials

2.1. Study areas

The study was conducted in six catchments located in the moist climatic zone (i.e. annual humidity index >= 0.65) of Ethiopia. The catchments selection is primarily based on three parameters: land cover type, rainfall seasonality and availability of hydro-climatic observations

2.2. Data

The data used in this study include: stream flow, rainfall, max./min. temperature and Terra-Aqua combined land product (MCD43A4).

2.3. Research methods

2.3.1. Hydro-climatological data pre-processing

We spatially averaged the monthly climate data to create monthly and annual time series of precipitation and temperature for each catchment. Missing stream flow observations were filled and data quality assessment identified no significant suspect. We then preprocessed the stream flow and climate records to cover the 2000-2006 period to obtain data contemporaneous with NDVI at all study catchments.

2.3.2. Hydrograph separation

The aim of hydrograph separation is to distinguish two stream flow components (baseflow, B and surface runoff, S). The one parameter recursive digital filtering algorithm (Nathan and McMahon, 1990) was used in order to separate B and S.

2.3.3. Hydrologic partitioning

Hydrologic partitioning is a concept first defined by L'vovich (1979) as the two-stage partitioning of water at the land surface. At annual time scale, the incoming precipitation (P) is first divided between wetting (W) and surface runoff. Next, the wetting component is partitioned into vaporization (V) and baseflow. Here, vaporization is the sum of all evaporation and transpiration. According to Ponce and Shetty (1995), this concept enables a better water balance view than conventional methods.This concept is adopted to quantify water balance components at annual time scale.

2.3.4. Hydrologic and vegetation indices calculation

a. Horton index: It is calculated as the ratio of V to W (see Troch et al., 2009). This index is a function of the energy and water available to plants, and can be used to evaluate how climate interacts with the terrestrial vegetation (Brooks et al., 2011).

b. Humidity index: It is calculated as the ratio of P and annual Potential EvapoTranspiration, PET (see Hulme et al., 1992). The 1985 Hargreaves equation (Hargreaves and Samani, 1985) is used to calculate monthly PET. Monthly values were then summed to obtain annual PET.

c. NDVI: We used MODIS's MCD43A4 product to create catchment-average monthly and annual NDVI time series. The two bands used for NDVI calculation are Band 1 (620nm -- 670nm, red) and Band 2 (841nm -- 876nm, near infrared).

3. Results

Water balance components showed variability across catchments. This variability signifies in catchments receiving nearly equal average annual rainfall but showing more than double variability in their total flow (Table 1).

Table 1: Mean annual values (±standard deviations) of hydrologic budget over 2000-2006 in six catchments.

Strong variability in hydrologic and vegetation indices is also noticed (Table 2).

Table 2. Mean annual values (±standard deviations) of hydrologic and vegetation indices over 2000-2006 in six catchments

4. Discussion

The fraction of P contributing to either S or W was relatively constant. This suggests that P is partitioned between its major components in nearly similar ways in all catchments. Year-to-year similarity of HI indicates vegetation can utilize the largest possible proportion of wetting. Little inter-annual variability of NDVI in all catchments indicate rare natural and/or anthropogenic impact on vegetation.

5. Conclusion

We revealed that catchments found within the same climate regime exhibit variable hydrologic partitioning and vegetation response behavior. Therefore, for successful water resources development, issues like catchment management and the impact of climate variability on vegetation should be treated differently in different catchments regardless of similarity in climate regime. 1. Brooks, P. Troch, P. Durcik, M. Gallo, E. and Schlegel, M. (2011). Quantifying regional scale vegetation response to changes in precipitation: Not all rain is created equal. Wat. Res. Resr. 47, W00J08. DOI: 10.1029/2010WR009762.
2. Brutsaert, W. (1988). The parameterization of regional evaporation-some directions and strategies. Jour. of Hyd. 102, 409 – 426.
3. Hargreaves, G. and Samani, Z. (1985). Reference crop evapotranspiration from temperature. App. Eng'g. in Agri. 1, 96 – 99.
4. Hulme, M. Marsh, R. and Jones, P. (1992). Global changes in a humidity index between 1931 – 60 and 1961 – 90. Clim. Res. 2, 1 – 22.
5. LÂ’vovich, M. (1979). World Water Resources and Their Future: English translation, American Geophysical Union, Washington DC, USA.
6. Nathan, R. and McMahon, T. (1990). Evaluation of Automated Techniques for Baseflow and Recession Analyses. Wat. Res. Resr. 26, 1465 – 1473.
7. Ponce, V. and Shetty, A. (1995). A conceptual-model of catchment water-balance: 1. Formulation and calibration. Jour. of Hyd. 173, 27 – 40.
8. Sivapalan, M. Yaeger, M. Harman, C. Xu, X. and Troch, P. (2011). Functional model of water balance variability at the catchment scale: 1. Evidence of hydrologic similarity and space-time symmetry. Wat. Res. Resr. 47, W02522. DOI: 10.1029/ 2010WR009568.
9. Thompson, S. Harman, C. Konings, A. Sivapalan, M. Neal, A. and Troch, P. (2011). Comparative hydrology across AmeriFlux sites: The variable roles of climate, vegetation, and groundwater. Wat. Res. Resr.47, W00J07. DOI: 10.1029/2010WR009797.
10. Troch, P. Martinez, G. Pauwels, V. Durcik, M. Sivapalan, M. Harman, C. Brooks, P. Gupta, and H. Huxman, T. (2009). Climate and vegetation water use efficiency at catchment scales. Hyd. Proc. 23, 2409 – 2414.
11. Zhou, L. Kaufmann, R. Tian, Y. Myneni, R. and Tucker, C. (2003). Relation between inter-annual variations in satellite measures of northern forest greenness and climate between 1982 and 1999. Jour. of Geoph. Resr 108, 4004. DOI: 10.1029/2002JD002510.

IWRA Proceedings office@iwra.org - https://www.iwra.org/member/index.php