Jacobs Journal of Environmental Sciences

Estimation of Spatially Distributed Groundwater Recharge in Modjo River Catchment, Awash Basin, Central Ethiopia

*Negash Bedaso
Department Of Agricultural Sciences, Sinana Agricultural Research Center, Ethiopia

*Corresponding Author:
Negash Bedaso
Department Of Agricultural Sciences, Sinana Agricultural Research Center, Ethiopia
Email:baatii2004@gmail.com

Published on: 2019-07-05

Abstract

Sustainable groundwater management requires knowledge of recharge. Recharge is an important factor in evaluating groundwater resources but difficult to quantify. Hence estimation of groundwater recharge requires modeling of the interaction between all the important processes in the hydrological cycle. In this study the long term seasonal and annual groundwater recharge of Modjo river catchment (2,202 km2) was estimated and recharge map were developed through a grid based physically distributed model, WetSpass. Long term average hydro-meteorological data and spatial pattern of watershed physical grid maps were used as main inputs for the model. All input maps for the model were prepared using ArcGIS 10.2 spatial analysis tool. Soil, land use and runoff coefficient parameters in data base files, season independent gridded base map of topography, slope, and soil were used in the model; whereas precipitation, potential evapotranspiration, temperature, wind speed, groundwater depth and land use map were prepared and employed by the model, in ASCII grid format of 120m cell size with 647 numbers of row and 425 numbers of columns for both winter and summer seasons. From the result, it is found that the long-term temporal and spatial average annual rainfall of 933 mm was distributed as: surface runoff of 164 mm (17.6%), evapotranspiration of 686 mm (73.5%), and recharge of 83 mm (8.9%). Thus an average of 183Mm3 of groundwater will be recharged per year or 5,802 liter/second from the catchment area. Flood control dams (artificial recharge) practice was recommended in this study area to harvest the excess water (simulated annual surface runoff of 361Mm3) which is helpful in one way to reduce soil erosion and in the other way to enhance more recharge to groundwater.

Keywords

Groundwater, Flood control dams, Modjo river, Evapotranspiration, Reduce soil erosion

Introduction

Groundwater is a critical source of fresh water throughout the world. It has become an important and dependable source of water supplies in all climatic regions including both urban and rural areas of developed and developing countries. Comprehensive statistics on groundwater abstraction and use are not available, but it is estimated that more than 1.5 billion people worldwide rely on groundwater for potable water. Other than water stored in ice caps and glaciers, groundwater accounts for approximately 97% of fresh water on Earth. As the world population continues to grow, more people will come to rely on groundwater sources, particularly in arid and semiarid areas. Long-term availability of groundwater supplies for burgeoning populations can be ensured only if effective management schemes are developed and put into practice. Hence, knowledge of groundwater resource potential is important for its management and sustainable use, because the optimal exploitation of the groundwater requires previous knowledge on the aquifers potential.

Groundwater potential is directly dependent on recharge. For efficient and sustainable management of the groundwater resource, understanding and quantification of groundwater recharge have paramount importance [6]. Recharge can be defined as the entry into the saturated zone of water made available at the water table surface. It is the process by which water percolates down the soil and reaches the water table either by natural or artificial methods to replenish the aquifer with water from the land surface. Generally, quantification of natural rates of groundwater recharge (i.e. the rates at which aquifer waters are replenished) is imperative for efficient groundwater resource management. Even though recharge is an important factor in evaluating groundwater resource, it is found difficult to quantify [7].Therefore, groundwater models have been used as tools for investigating groundwater system dynamics [8] .Hence, WetSpass was built as a physically based methodology for estimation of the long-term average, spatially varying, water balance components: surface runoff, actual evapotranspiration and groundwater recharge [9] . It is an acronym for water and energy transfer between soil, plants and atmosphere under quasi-steady state that was built upon the foundations of the time dependent spatially distributed water balance model (Batelaan and Smedt, 2001 and 2007). In this study geographical information system GIS and WetSpass model, were used for groundwater recharge estimation.

Studies on groundwater potential of Ethiopia show erroneous results of 2.5 BCM by WAPCOS, to 185 BCM by Ayenew and Alemayehu, [10]. Best guesses in this respect ranges between 12-30 BCM or even more if all aquifers in the lowlands are assessed [11]. This ambiguity is an indication of how much detailed study and survey is needed to estimate the countries resources with a better precision; because such uncertainty can have a hindering effect on the countries pursuit to utilize its water resources potential to the limit. To formulate technically-sound groundwater resources management policies, decision makers always ask questions like: How long can an aquifer maintain the current rate of groundwater abstraction? What is the safety yield that the aquifer can sustain the continuous abstraction? What is the capture zone of a water supply well field? The analysis of estimating groundwater recharge will assist scientific community, policy makers, donors, non-governmental organizations and other development practitioners to deliver right policy and programs in required areas on time. As a result, this study provides valuable data and baseline information in formulating technically sound groundwater resources management policies for the response of declining groundwater tables in the study area. Therefore, this study was carried out with general objective of estimating groundwater recharge in the Modjo river catchment. Hence, the study had focused on developing groundwater recharge map and estimating the average recharge amount of Modjo river catchment.

Material and Methods

Description of Study Area

This study was conducted at Modjo river catchment, which is located at upper Awash Basin. The sub-basin covers about 2201.98 km2 area and is bounded within 8° 75’N-9° 05’N latitude and 38° 56’E -39° 17’E longitude. The average elevation in the catchment ranges between (1591 to 3060) meter above mean sea level. The area has a bi-modal rainfall with a short rainy season from March to May and with a long rainy season from June to September. The mean annual temperature of the sub-basin is 19.91° C and the mean annual rainfall is 933 mm. The relief of the study area is generally flatland with an undulation of some ridges and mountains like eastern part of Yerer and the catchment generally shows an eastward decrease in elevation above mean sea level. The water units found in the study area are Modjo and Gale Wemecha rivers. As to the geological set up,the area belongs to the Quaternary rocks of Pleistocene and Holocene which is 70% quaternary volcanic rock and 30% unconsolidated sediment. According to (WSP, 2006) those rocks are highly fractured rocks and resulted a favorable situation for groundwater recharge and occurrence and are very important hydro-geological formation that are used as good source of groundwater in Ethiopia.

Figure 1: Location map of Modjo Sub-basin.

Data and software used

ArcGIS10.2 were used for all data processing, data management and data preparation of the model required, and WetSpass model to estimate the long-term average spatially varying water balance components of the area. Twenty eight years (1987-2014) climatic data were collected for all station from Ethiopian National Meteorological Services, and the reference evapotranspiration (ETo) was computed using FAO-Penman-Monteith equation. Forty-one static water level data for groundwater depth estimation in the sub-basin was collected from MoWIE and finally all the data were processed and grid map of each data was prepared using ArcGIS 10.2. The watershed delineation was carried out with Ethiopian digital elevation model obtained from SRTM (Shuttle radar topographic mission), topographic map and slope map of the study area were also developed from this DEM (digital elevation model). In addition Enhanced Thematic Mapper Plus (ETM+) satellite imagery was used to develop land use classification of the study area; and the soil map of the study area were prepared from soil map data-base of the Northeastern African.

Climatic data

The study area is located within the main Ethiopian rift and mostly affected by the southerly and easterly Indian Ocean air currents, as a result the air currents supply rain with bimodal characteristics. There are two main seasons in the study area, namely summer and winter also locally called as Kiremt, Bega respectively. The main rainfall season is summer season ends only for four months from June to September. During this period the region receives more than 70% of the total annual rainfall. Twenty eight years (1987-2014) climatic data was used for six metrological stations selected for this study (see Table 1) and the spatial areal rainfall distribution of each station is computed with Thiessen polygon using ArcGIS 10.2.

Table 1: Rain gauge stations in Modjo river catchment

Debrazeyit, Modjo, Koka and Hombole are class one metrological stations which include rainfall, max and min temperature, relative humidity, wind speed and sunshine hour whereas stations Chefedonsa and Ejere are class two metrological stations which contain rainfall, max, min temperature and wind speed. Other parameters like relative humidity and sunshine hour were filled for missed period of those two stations by normal ratio-method which is recommended to estimate missing data in the sub-basin where annual rainfall among stations differ by more than 10% [12] for this study the variation among station shows 12%. The homogeneity of annual rainfall data for each station was tested using XLSTAT 2017 software by means of SNHT test (Standard Normal Homogeneity Test) and double mass curve method was used to check and correct consistency of recording data.

WetSpass application

The total water balance for a raster cell (Figure 2) is split into independent water balances for the vegetated, bare-soil, open-water and impervious parts of each cell. This allows one to account for the non-uniformity of the land-use per cell, which is dependent on the resolution of the raster cell. The processes in each part of a cell are set in a cascading way. This means that an order of occurrence of the processes, after the precipitation event, is assumed. Defining such an order is a prerequisite for the seasonal timescale with which the processes will be quantified. The quantity determined for each process is consequently limited by a number of constraints.

Figure 2: Schematization and integration of data for a hypothetical raster cell in the WetSpass water balance model, after Batelaan and Smedt (2001).

Water balance calculation

The water balance components of vegetated, baresoil, open-water, and impervious surfaces are used to calculate the total water balance of a raster cell as follows:

???????????????????????????????? = ???????????????????? +???????????????? +????????????????+???????????????? 

???????????????????????????? = ???????????????? +???????????????? +????????????????+???????????????? 

???????????????????????????? = ???????????????? +???????????????? +????????????????+????????????

where ETraster, Sraster, Rraster are the total Evapotranspiration, Surface runoff, and Groundwater recharge of a raster cell respectively, each having a vegetated, bare-soil, open-water and impervious area component denoted by av, as, ao, and ai, respectively. Precipitation is taken as the starting point for the computation of the water balance of each of the above mentioned components of a raster cell, the rest of the processes (interception, runoff, evapotranspiration, and recharge) follow in an orderly manner.

WetSpass adjustment

WetSpass is originally developed for conditions in the temperate regions; there should be some adjustment to use it for the case of Ethiopia. Summer and winter land-use, soil and runoff coefficient parameters are the four parameters tables used by WetSpass and be adjusted according to different area. They are connected to the model as attribute tables. The land-use attribute table includes parameters such as land-use type, rooting depth and leaf area index and vegetation height. The soil parameter table contains soil parameters such as textural soil class, and plant available water contents. The runoff coefficient attribute tables contain parameters for runoff classes of various land-uses, slope and runoff coefficient. These attributes tables allow for easy definition of new land-use and soil type as well as changes to each parameter value.

The original land use parameter tables for the model was developed based on land use types and characteristics from temperate regions, Europe in particular. Thus an attempt has been made, in this study, to modify the land use parameter table as the case in Ethiopia is different than what someone can find in Europe. Basic modification has been done on the land use parameters such as leaf area index, crop height, interception and percentage. In this regard, the vegetative area, bare land area, impervious area, and open water area proportions of each land use class in Modjo sub-basin has been defined based on the knowledge about the natural characteristics of the different land use types in the sub-basin; some of the seasonal land uses parameter values were readjusted as it has been used by [13] for Ethiopia case. In addition, the parameters tables were adjusted by try and error through WetSpass watershed simulation repeatedly; this was done depending on the estimated groundwater recharge from well data and adjusting WetSpass recharge simulation value to the known ground truth spatial groundwater recharge obtained from the well data.

Model Input Data Preparation

WetSpass is a steady state models and therefore, needs long-term average climatic data (precipitation, reference evapotranspiration, temperature and wind speed), and catchment configuration data (groundwater depth, slope, elevation, and land use), soil data and boundary conditions (extent of area to be modeled). All input data must be in ASCII grid file format. Thus as a first step, a mask map was prepared from the delineated catchment based on the DEM and Modjo sub-basin with a total area of 2202 km² was delineated and this mask map has been used to develop all other grid-map inputs for the model. 

Result and discussion

Spatial Data for Model Input

Soil map

The sub-basin is covered with four different soil textural classes, 54% of silty clay, 34% of silty loam, 8% of clay and 4% of loam soil. The soil map of sub-basin used in watershed simulation is shown in Figure 3 below.

Figure 3: Dominant soil textural class of Modjo sub-basin

Topography

Altitude in the basin increases from south to north and from west to east. The lowest point in the basin is located in the western edge and the highest in the north. The mean elevation of the basin is 2030 m with a standard deviation of about 273 m. This considerably large standard deviation explains the fact that the topography is rugged. Figure 4 below is the topographic grid map of study area used in the watershed simulation.

Figure 4: Topographic map of Modjo sub-basin

Slope

The slope ranges from 0 to 52%, with 3% mean and standard deviation of 4%. Most of the agricultural area lies within the slope of 3-10%.

Figure 5: Slope map of Modjo sub-basin.

Land Use Classification

The land use classification of the study area were developed from Landsat image and the identity of each class is determined by a combination of experience and ground truth (i.e. visiting the study area and observing the actual cover types) and the land use/cover type obtained accuracy estimation were justified by error (confusion) matrix. Accordingly, in Modjo sub-basin agricultural land being the dominant which comprises 49.3% of the total sub-basin area, grassland 16.5%, Tree & Shrub 15% and Water body 7%, Settlement 11% were the prominent types of land use type.

Figure 6: Summer land use map of Modjo sub basin.

Table 2: Soil parameter table for Modjo sub-basin.

Table 3: Summer land-use parameter table for Modjo sub-basin.

Figure 7: Winter land use map of Modjo sub-basin (source: own analysis).

Adjusted land use parameter table

The land use parameter table was modified and land use parameter tables for Modjo sub-basin summer and winter seasons was developed. The modified summer and winter land use parameter tables have been used to run WetSpass for the sub basin modeling processes. The highlighted portion on Table 3 and 4 below indicates the amended parameter table values for the study area.

Table 4: Winter land-use parameter table for Modjo sub-basin.

Climatic Data for Model Input

Areal rainfall distribution

To compute spatial areal rainfall, Theissen polygon method was used in ArcGIS environments and the gauge weights developed for sub-catchments are presented in Table 5.

Table 5: Theissen gauge weights for Modjo sub-basin.

Figure 8: Theissen polygon developed for Modjo sub-basin.

Precipitation

The mean annual precipitation value calculated for the sub-basin is 933mm and its grid map used in the model for watershed simulation is shown on Figure 9 below.

Figure 9: Average annual precipitation of Modjo sub-basin

Temperature

The study area has an average temperature of 19.91° C with a minimum temperature of 11.6° C and a maximum temperature of 29.2° C. Maximum temperature values were obtained in the months of May and minimum temperature was recorded in the month of August and it has generally been observed that the average annual temperature decreases with an increase in altitude.

Figure 10: Average annual temperature of Modjo sub-basin.

Evapotranspiration

Often a value for the potential evapotranspiration is calculated at a nearby climate station on a reference surface, conventionally short grass. This value is called the reference evapotranspiration (ETo) which is equal to PET. For this simulation PET grid map shown on Figure 11 is used as input for the model.

Figure 11: Average annual PET of Modjo sub-basin.

Wind speed

The mean monthly wind speed of the area measured at two meters above the ground varies from 1.6 to 2.6 m/s. with maximum values observed in the months between February to May and minimum values in the months of August and September.

Figure 12: Average annual wind speed of Modjo sub-basin

Groundwater depth

In this study, an average value of groundwater level was used for all station from data collected and assumed as the groundwater depth for both summer and winter seasons.

Figure 13: Average groundwater depth of Modjo sub-basin.

Water Balance Analysis

Water balance represents the hydrological gains and losses of a given system. For this study WetSpass modling was applied and total water balance of a raster cell computations were calculated depending on the following equation in the model and all water balance were carried out from the model output.

???????? =???? − ????????− ????????????− ????????− 

Where(Rv)recharge,(P)precipitation,(Sv) surface runoff (ETv) Evapotranspiration,(Es)evaporation from bare soil and(I) Interception of a raster cell.

Evapotranspiration

The WetSpass model calculates the total actual evapotranspiration as a sum of the evaporation of water intercepted by vegetation, the transpiration of the vegetative cover and the evaporation from the bare soil between the vegetation. The simulated average minimum and maximum annual evapotranspiration of the catchment are 359 mm and 952 mm respectively, with 686 mm mean and standard deviation of 141 mm distribution.

Figure 14: Simulated annual evapotranspiration with the WetSpass model for Modjo sub basin.

The average evapotranspiration accounts more than 73.5% of the total annual rainfall. This shows that evapotranspiration is the main processes by which water is lost in the catchment. This is attributed to the high rates of radiation and the persistence of strong dry Westerly winds coming from the awash depression. The evapotranspiration is largely determined by the solar radiation, which is fairly constant between years. As a result evapotranspiration varies little from year to year, especially in the dry season. About 77% of the total annual evapotranspiration is lost during summer season while the rest 23% is released in the winter season. This variation occurs due to difference in precipitation within the two seasons. According to [14] the average annual actual evapotranspiration of Modjo river catchment is about 650.33 and 789.47 mm with Thornthwaite and Turc method respectively. [15], reports similar results for the nearest watershed Welanchiti area, annual actual evapotranspiration, from Turc method give value of 697 mm. Thus 952 mm maximum and 686 mm average annual evapotranspiration, simulated by WetSpass for Modjo river catchment, is reasonable. WetSpass has also simulated constituents of evapotranspiration, i.e. interception, soil evaporation and transpiration. Interception is the part of the rainfall that is intercepted by the earth’s surface and subsequently evaporated. It can take place by vegetal cover and depression storage in puddles and in land formation such as rills and furrows. Interception can amount up to 15-50% of precipitation, which is significant part of the water balance. For this sub basin 260 mm and 15 mm interception was simulated by WetSpass as maximum and minimum respectively and the mean was 177 mm which accounts about 25% of the mean evapo- transpiration. WetSpass also simulate 447 mm and 52 mm maximum and minimum transpiration with 290 mm mean value, soil evaporation of 375 mm maximum and 12 mm minimum with mean of 220 mm was simulated for the sub-basin. See the map of three constitute of evapotranspiration on Figure (15-17).

Figure 15: Simulated annual transpiration with WetSpass model for Modjo sub basin.

Figure 16: Simulated annual soil evaporation with WetSpass model for Modjo sub basin.

Figure 17: Simulated annual interception with WetSpass model for Modjo sub basin.

The relation between transpiration and soil evaporation depends strongly on the density of the plant cover, expressed by the leaf area index (LAI). In general, soil evaporation decreases rapidly with increasing LAI while the opposite is true for transpiration [16]. Since WetSpass determines the total evapotranspiration as the sum of the evaporation from soil, intercepted water and transpiration, it is obvious that the evapotranspiration varies spatially according to the land use class and soil types. Since the catchment is located in one of the sub-humid regions of Ethiopia, soil evaporation is less important than transpiration. The simulated result also supports this fact that, as can be understood; evapotranspiration in the catchment is mostly in the form of transpiration and hence it is less dependent on the soil type.

Surface runoff

According to [17] , surface runoff is dependent on the availability of vegetation, soil type and slope of the sub-basin. Hence the surface runoff of Modjo river catchment varies spatially with topography and other catchment characteristics. Rugged topography and silty clay, silty loam soil dominated the lands and gave the largest amount of runoff in the catchment. This is due to a lower concentration time of overland flow for rugged topographic surface and the lesser infiltration capacity of the soil type.

Figure 18: Simulated annual surface runoff with the WetSpass model for Modjo sub-basin.

The simulated annual runoff varies from 83 mm to a maximum of 341 mm with a mean and standard deviation of 164 mm and 66.89 mm respectively. This accounts about 17.6% of the total annual precipitation. Considering the (2202 km2 ) area of Modjo river catchment, average annual surface runoff will be 361Mm3 . The rainfall exceeds the infiltration capacity of the soil during the wet season, this leads to high surface runoff. Equivalently about 81% of the surface runoff occurs during the wet months (June to September) while the remaining 19% occurs during the dry months (October to May) from the catchment applied runoff coefficient method and revealed that the annual surface runoff of Modjo river catchment of area of 2202 km2 is 130 mm, also the research conducted in the nearest watershed Welanchiti area by Getachaw (2007) reveals the annual surface water out flow, estimated using runoff coefficient methods from an area of 780 km2 resulted 129.5 Mm3 . Thus 83 mm minimum and 341 mm maximum annual surface runoff, simulated by WetSpass for Modjo river catchment, is reasonable.

Groundwater recharge

Recharge is promoted by natural vegetation cover, flat topography, permeable soils, a deep water table and the absence of confining beds. The WetSpass model determines the long term average spatially distributed recharge as a spatial variable dependent on the soil texture, land use, slope and meteorological conditions.

Figure 19: Simulated annual groundwater recharge with the WetSpass model for Modjo sub-basin.

Figure 20: Simulated summer groundwater recharge with WetSpass model for Modjo sub-basin

Figure 21: Simulated winter groundwater recharge with the WetSpass model for Modjo sub basin.

The resulting groundwater recharge from WetSpass for the present land use ranges from about 274 mm/yr to zero, with an average value of 83 mm/yr, which is about 8.9% of the mean annual precipitation, and standard deviation of about 32 mm. About 70% of the annual ground water recharge of the sub-basin occurs during the wet season (summer), and the remaining 30% in dry season (winter) specially (April and May).Thus an average of 183 Mm3 of ground water will be recharged per year for total catchment area. For the study conducted on groundwater potential of Ada’a Becho plain, annual groundwater recharge contributed from Modjo River to Ada’a Becho plain is 85 mm and 153 Mm3, which is almost comparable with WetSpass result. Similarly (WAPCOS, 1990) report 7% of annual groundwater recharge for Awash basin for the mean annual rainfall of 850 mm. Therefore the result simulated by WetSpass for the Modjo river catchment is within the previous study range and was found reasonable.

Seasonal Result Analysis

The potential evapotranspiration in the catchment is larger in winter (66.5%) than in summer (33.5%). This is apparently associated with longer sunshine hour and larger wind speed values in winter than in summer. However, the simulated actual evapotranspiration in winter is lower for most of the land use types than for summer. This is due to the lower rate of soil evaporation, transpiration and interception in winter due to insufficient rainfall to meet the potential demands. During summer the evapotranspiration is so excessive that it causes negative recharge in some parts of the catchment, in winter since the rainfall is not excessive to create too much runoff and also due to the fact that the soil is dry the groundwater recharge has found to be larger than summer. However, the total recharge in the catchment is larger in summer due to the fact that high rainfall in this season.

Table 6: Mean seasonal and annual WetSpass hydrologic output for Modjo river catchment.

Water Balance Analysis by Land Use and Soil Type

Since WetSpass determines the total evapotranspiration as the sum of the evaporation from soil, intercepted water and transpiration, it is automatic that the evapotranspiration varies spatially according to the land use class and soil types. Water bodies (936 mm) and trees and shrub (756 mm) land use is the land cover types where the simulated evapotranspiration is the highest followed by grassland (675 mm) due to high evaporation from open water surface and transpiration rates from vegetation respectively, while highland areas of the cropland and settlement with lower evaporation and transpiration, because large areas of this land use is found to be hilly stone covered impervious surfaces and relatively low temperature areas. The WetSpass model uses the runoff coefficient method for the estimation of surface runoff, whereas this parameter is a function of vegetation type, soil texture and slope. Similarly, the highest runoff rate occurs from the agricultural lands (290 mm) and settlement (233 mm) due to disturbed soil and impervious surfaces in these land-use classes respectively. Plain areas, more silty loam and silty clay soil covers, and areas with vegetation cover have relatively lower runoff rates in the catchment.

Groundwater recharge is generally found to be much higher in non-vegetated land-uses than in vegetated land-uses and is greater in areas of annual crops and grasses than in areas of trees and shrubs. In this simulation, from WetSpass water balance of the catchment high recharge occurs from grassland on the plain areas (155 mm). This landuse unit has lower runoff and evapotranspiration compared to the cultivated land and tree and shrubs; secondly the groundwater recharge was high on agricultural land (143 mm), because this land unit is bare and the soil is dry at the beginning of the rainy season having low transpiration; and drying up of soil favors more water infiltration contributing to recharge.

Table 7: Average water balance in land use difference.

The evapotranspiration seems not to show a clear pattern with soil texture in general, it is less dependent on the soil type in the sub-basin. The highest runoff is generated from silty loam and clay soil located in the high slope and high rainfall area. This is due to high rainfall and rugged topographic nature of the area. The runoff rate decreases as the soil gets lighter loam and silty clay; however, the runoff difference in this simulation is more affected by rainfall amount and topographic effect than soil type. High values of groundwater recharge are observed in the grassland and cropland with silty loam, silty clay and loam soils. This is due to relatively good permeability of the soils and gentle slope. Areas where clay soil is dominant yield the lowest rate of recharge (36 mm). However, the variation of recharge with land use type is more pronounced than its variation with soil type. In general, the annual results analysis reveals that evapotranspiration is the most important hydrologic process in the basin, by which most of the precipitation is lost, than surface runoff and recharge.

Table 8: Average water balance in soil type difference.

Figure 22: WetSpass simulation for average runoff, actual evapotranspiration, and recharge for different slop classes in Modjo sub basin.

Conclusion and Recommendations

In this study the long term seasonal groundwater recharge of Modjo river catchment (2,202 km2 ) was estimated and the recharge zone is mapped through use of a grid based physically distributed model, WetSpass. The model applies up to date physical and empirical relationships of the sub basin for its efficiently running processes. Obviously, long term average hydro meteorological data and spatial patterns of watershed physical grid maps are used as main inputs for the model. Seventeen model parameter variables are used as an input for the WetSpass model in ASCII grid map and dbase file formats. Season independent gridded base maps of soil, slope, and topography; and soil, land use and runoff coefficient parameters in dbase files are used in the model. Precipitation, potential evapotranspiration, temperature, wind speed and groundwater depth and land use map are also prepared and employed by the model, in ASCII grid format for both winter and summer seasons. GIS and remote sensing techniques have been applied to develop the land use map of the sub-basin and ArcGIS 10.2 to prepare the data per model required.

Finally annual and seasonal values and ASCII grid map of runoff, Evapotranspiration, interception, transpiration, soil evaporation and finally recharge of the sub basin are obtained as model output results. From the results of this study, the following recommendations are made:

• The water balance results obtained from this modeling can be used as base for future groundwater resources development and improvement of the catchment in particular, for soil and water conservation work in general.

• The largest amount of evapotranspiration simulated for the catchment, relative to the groundwater recharge and the surface runoff, indicates that much effort is needed to change the environmental conditions of the catchment by applying some soil and water conservation practices.

• An average 5,802 l/s recharge is calculated for the catchment, for further groundwater resource development plans of sustainable use, the abstraction rate both for irrigation and domestic water supply from the sub-basin should be maintained under consideration of this rate.

• Well drilling requires huge amount of investment cost, however it is common to have problem in exploitation of groundwater in that unsuccessful rate of well production encounter, hence, using WetSpass output groundwater recharge map with consideration of geologic properties of the aquifer is recommended to reduce this problem in site selection for water abstraction point.

• The simulated annual surface runoff is 361Mm3 . Therefore, to harvest this excess water, it could be advantageous to practice flood control dams (artificial recharge). This is helpful in one way to reduce soil erosion and in the other way to enhance more recharge to groundwater.

• This study depends only on the hydro metrological data and spatial pattern of the sub-basin to understand spatial and temporal groundwater recharge, further study should be investigated including aquifer property for more understanding of groundwater dynamics in the catchment for better groundwater resource development and management.

Acknowledgement

The authors would like to thank Ethiopian National Meteorological Agency for providing the necessary data. We also acknowledge USGS Earth Resources Observation portal for providing ASTER DEM data and Prof. O. Batelaan for the provision of WetSpass program. This research work was funded by Oromia Agricultural Research Institute.

References

1. Todd D.K: Groundwater hydrology: Wiley; 2005.

2. Clarke R, Lawrence A and Foster S: Groundwater: A threatened resource. Nairobi, Kenya: United Nations Environment Programme Environment Library No. 15;1996.

3. Shiklomanov I. A and Rodda J. C: World Water Resources at the Beginning of the Twenty-frst Century. Cambridge: Cambridge University Press; 2003.

4. D. N. Lerner, A. S. Isaar and I Simmers: Groundwater recharges and water resources management. In Groundwater Recharge, A Guide to Understanding and Estimating Natural Recharge. International Contributions to Hydrogeology Hanover: Verlag Heinz Heise; 1990: 3-22.

5. Benjamin N.N, Jacques M and Jean S.R, et al: Water Quality Data Analysis, Plotting, and Modeling. User’s Manual (Ed.).Waterloo Hydrogeologic, Inc, World Applied Sciences Journal; 2007.

6. Obuobie Emmanuel, B. Diekkrüger and B. Reichert: Estimation of groundwater recharge in the context of future climate change in the White Volta River Basin, West Africa. Dissertation; 2008.

7. Alley W. M, Healy R. W and LaBaugh J. W, et al. Flow and storage in groundwater systems. Science 2002; 296(5575).

8. Anderson MP and Woessner W W: Applied Groundwater Modeling:Acad Press, San Diego, USA;1992.

9. Batelaan and F.De Smedt: WetSpass: a flexible GIS based distributed recharge methodology for regional groundwater modelling”, Impact of Human Activity on Groundwater Dynamics: IAHS Publ; 2001:269.

10. Semu Moges: Agricultural Use of Ground Water in Ethiopia: Assessment of Potential and Analysis of economics, Policies, Constraints and Opportunities; 2012.

11. Ministry of Water Resources. Ethiopia: Strategic Framework for Managed Groundwater Development; 2011.

12. Dingman S.L: Physical Hydrology.2nd edition:Prentice-Hall; 2002.

13. Tesfamichael G, De Smedt F and Miruts H et al: Large scale geological mapping of the Geba basin, northern Ethiopia. Tigray Livelihood VLIR – Mekelle University IUC Program; 2010.

14. Behailu Berehanu: Impact of Industries and Urbanization on water Resource in Modjo River Catchment. Unpublished MSc Thesis. Addis Ababa University; 2007.

15. Getachew Abrha: Groundwater Potential Assessment of Welenchiti Area. Unpublished MSc Thesis. Addis Ababa university; 2007.

16. Merta M. Analysis of land use change in the Eastern Erzgebirge regarding both nature protection and flood prevention. Soil and Water Research 2008; 3(1).

17. Batelaan.O and F.De Smedt. GIS-based recharge estimation by coupling surface-subsurface water balances, Journal of Hydrology 2007; 337(3-4).

18. Hoboken NJ and Yazew E: Development and management of irrigated lands in Tigray, Ethiopia. PhD thesis; 2005.

19. WAPCOS , Preliminary Water Resources Development Master Plan for Ethiopia. Final Report. Prepared for EVDSA. Addis Ababa; 1990.

20. WSP. Country Case-Study by the Cost-Effective Boreholes Flagship of the Rural Water Supply Network Final Report June; 2006.