Keywords

1 Introduction

Global economic development and human welfare are often limited by the availability and quality of water. The rapid growth of the world population and economic development are increasing the global water demand (Kundzewicz et al. 2008). Agriculture, food production and water are inseparably linked (Watts et al. 2015). Water use in agriculture accounts for 70% of the global total water use (Hatfield 2014) and has a significant impact on the water balance components. Agricultural land use affects the hydrologic cycle in terms of the partitioning of rainfall between evapotranspiration, runoff, and groundwater recharge (Watts et al. 2015). The quality of surface water and groundwater has generally declined in recent decades mainly due to an increase of agricultural and industrial activities (Parris 2011). Moreover, climate change studies show that surface water and groundwater resources are expected to decrease significantly in the future in most dry subtropical regions (Pachauri et al. 2014). This could cause water stress and intensive competition for water among sectors (Björklund 2001). The effect of climate change is expected to be more severe on the social-ecological systems in developing countries since they have the lowest capacity to adapt (Dile et al. 2018). Besides this, the adverse effects of climate change on freshwater systems aggravate the impacts of other stresses, such as population growth, changing economic activities, land-use change, and urbanization (Kundzewicz et al. 2008). Thus, providing sufficient quantity and acceptable quality of water to the world’s human population is one of the preeminent challenges of the twenty-first century (Tarhule 2017). As a result, availability and sustainable management of water and sanitation for all are part of the United Nations’ sustainable development goals (Brookes and Carey 2015).

Ethiopia is likely to be one of the countries with the largest increase in population between 2019 and 2050 (UN 2019). The agriculture sector plays a central role in its economy. About 85% of all employment relies on it (FAO, 2014). The sector is dominated by small-scale farms that are characterized by rain-fed mixed farming. Crop production accounts for about 60% of the agricultural outputs (Gebre-Selasie and Bekele 2012). The crop productivity varies with the availability of water and water use in agriculture. Although Ethiopia is perceived as the water tower of Eastern Africa, temporal variability (seasonality) and uneven spatial distribution of water resources remain a primary challenge. Availability of water is highly dependent on the seasonality and inter-annual variability of rainfall and streamflow. Most of the rivers have their peak flows during the rainy months (June–September) and cause flooding on areas of their surroundings. On the Streamflow volumes are considerably low during dry months (October–May) (Berhanu et al. 2014). The temporal variability of rainfall and streamflow extremes are linked to low frequency climate processes centred over the mid-latitudes of the Pacific basin (Taye et al. 2015) causing widespread, devastating droughts and floods that occur every 3–5 years (World Bank 2006). Crop failure or a decrease in agricultural production, and livestock perishing are the major consequences of drought that frequently occur in the country. Hence, the overall national GDP is frequently affected by the quantity and timing of rainfall. To overcome this widespread problem, understanding the effect of agricultural crops on the hydrologic cycle is important (World Bank 2006). Therefore, a better understanding of the historical, present, and future hydrologic situation is essential to meet the water resources management challenges.

While Ethiopia has considerable annual renewable freshwater potential, its agricultural crop production still depends on seasonal rainfall, and access for safe drinking water is still low. Because of siltation problems in rivers and reservoirs, resulting in substantially higher treatment and maintenance costs of the water schemes, groundwater is the primary source of water for urban and rural water supply (World Bank 2006). However, excessive groundwater abstraction has led to groundwater depletion and pollution and thus challenges water resources management in the country. These developments also have negative effects on the streamflow of groundwater fed rivers, the health of the ecosystems, and the depths of local groundwater tables (De Graaf et al. 2014). The complete drying up of Haramaya Lake in Eastern Ethiopia since 2005 is an example for the consequences of decreasing groundwater levels due to over-pumping of groundwater for agriculture and household use (Abebe et al. 2014).

The Upper Blue Nile Basin which originates from Lake Tana covers an area of 199,812 km2, i.e. 20% of Ethiopia (Dile et al. 2016). Past studies in the Upper Blue Nile Basin (e.g., Elshamy et al. 2009; Setegn et al. 2011; Woldesenbet et al. 2017) have reported that water resources in the basin are not being managed adequately due to the hydrological variability, climate change, land use changes, rapid population growth, soil erosion, and deforestation. Additionally, studies of hydrological and climate change came to different and contradictory conclusions. For instance, Tessema et al. (2010) reported that the mean Bnnual streamflow at the Lake Tana outlet (Abay) was significantly increasing during 1964–2003. On the other hand, Tekleab et al. (2013) found a decrease in the mean annual streamflows of Gilgelabay and Ribb sub-catchments (inflows) of the Lake Tana Basin during 1973–2003 and 1973–2005. Similarly, climate change studies of the Upper Blue Nile basin have shown different results of future rainfall at basin or sub-basin scale (Taye et al. 2015). Elshamy et al. (2009) reported that changes in total annual projected rainfall for the late twenty-first century in the Upper Blue Nile basin vary between −15% and + 14%, while Setegn et al. (2010) found no significant change. Kim et al. (2009) analyzed the changes in projected rainfall and temperature for six GCMs and the ensemble mean of the GCMs showed an increase of 11% of the mean annual rainfall for the 2050s. Another climate change impact study carried out by Beyene et al. (2010) in the Nile River basin, based on 11 GCMs, showed an increase in rainfall for the early twenty-first century (2010–2039) and a decrease during the mid (2040–2069) and late (2070–2099) century. Taye et al. (2011) also investigated the impact of climate change on hydrological extremes of the Lake Tana Basin. In this study, both decreasing and increasing rainfall and streamflow are expected. In general, past climate change studies on the Upper Blue Nile Basin and its sub-basins showed contradictory results of the projected rainfall changes, because most of the studies did not use an ensemble approach and they the climate data such as rainfall and temperature were from different climate model.

The Lake Tana Basin (Fig. 24.1) supports the livelihood of more than 3 million people, and it is well known for its water resource potential. Being the sub-basin of Upper Blue Nile basin, Lake Tana Basin is dominated by intensive agriculture with a strong impact on the hydrologic regime and ecological condition (Setegn et al. 2011). Moreover, Lake Tana Basin became a focus area of many scientific studies because of its national and international importance. These include water balance analyses of different sub-catchments including the lake (Derib 2013; Tegegne et al. 2013; Dessie et al. 2015), hydrological modeling with emphasis on surface water (e.g., Dessie et al. 2014; Worqlul et al. 2015; Polanco et al. 2017), hydrometeorological trend analyses (e.g., Gebrehiwot et al. 2014; Tigabu et al. 2020a), climate change impact studies (e.g., Koch and Cherie 2013; Teshome 2016), land use/cover change impact on hydrologic responses (e.g., Woldesenbet et al. 2017), implications of water harvesting intensification on upstream–downstream ecosystem services and water availability (e.g., Dile et al. 2016), and groundwater and hydrogeology (e.g., Kebede et al. 2005; Tigabu et al. 2020b).

Fig. 24.1
A set of 3 maps depicts the location of Ethiopia, land use, and major soil classes. The many soil types, such as chromic and more, as well as the different land uses, such as marsh, cultivated land, and more, can be seen.

Topography, soil map, land use and cover map, and location of the Lake Tana Basin with reference to Ethiopia

Full size image

2 Hydrological Modeling

Hydrological models are relevant tools to improve land and water management practices (Palanco et al. 2017). The use of physically based hydrologic models has been increasing over time because of their capabilities to incorporate physical processes of the system. The Soil and Water Assessment Tool (SWAT) model (Arnold et al. 1998; Arnold and Fohrer 2005) is one of the most widely applied river basin-scale hydrologic model. Because of its adaptability for both, large- and small-scale catchments and continuous improvement, researchers in different fields have identified SWAT as one of the most elaborate, dependable, and computationally efficient models (Neitsch et al. 2011; Palanco et al. 2017). The model is capable of simulating spatially distributed water balance components based on hydrological response units (HRUs).

Likewise, application of the SWAT model in Ethiopian watersheds is expanding over time. Among others, the following are a few case studies that applied SWAT in Ehiopia. Setegn et al. (2008, 2011) used SWAT2005 to test its applicability in the Lake Tana Basin and applied to evaluate the past and future water balance situation of the basin. In this study, the authors concluded that the SWAT model was capable of modeling the hydrological processes of the Lake Tana basin. Their water balance results showed that actual ET accounts for more than 60% of the basin rainfall, and the groundwater contribution to the streamflow varies between 40 and 60%. Dile et al. (2013) used the SWAT model to study the effects of climate change on hydrological responses of the Gilgelabay catchment. According to this study, the monthly mean volume of runoff for the catchment tends to increase significantly (up to 135%) for the mid and long-term in the twenty-first century. Dile et al. (2016) used SWAT2012 for their study to investigate the implications of intensive water harvesting on upstream–downstream ecosystem services in the Lake Tana Basin. In this study, they found that intensification of water harvesting in the basin would increase low flows and decrease peak flows. They also found a reduction in the amount of sediment yield that leave the catchment. Polanco et al. (2017) applied SWAT model to investigate the effect of discretization on its performances in the entire Upper Blue Nile Basin. Their results showed that increasing the number of sub-basins would affect the magnitude of streamflow and other water balance components and the authors suggested increasing the spatial resolution by increasing the number of sub-basins is important to improve the performances of the SWAT model in the basin. Another hydrological assessment study by Desta and Lemma (2017) in the Lake Ziway watershed, Rift valley, Ethiopia used SWAT, and their findings showed that the annual base flow is decreasing, while actual ET is increasing over time. Overall, all studies proved that SWAT model is capable of simulating the hydrological processes in Ethiopia as a whole and in the Upper Nile basin in particular. However, all the studies considered the water balance components at the outlets of the study catchments ignoring spatial distributions of water balance components. In addition to this, the seasonal and spatial variability of the hydrological components and their relation to vegetation cover condition have not been analyzed.

3 Groundwater Dynamics and Climate Change

Aquifers are the largest storage of global freshwater: more than two billion people rely on groundwater (Cuthbert et al. 2019a). Groundwater (GW) accounts for one third of all freshwater withdrawals of the world, supplying an estimated 36, 42 and 27% of the water for domestic, agricultural and industrial purposes, respectively (Taylor et al. 2013). Although GW is vulnerable to depletion, it is being consumed faster than it is being naturally replenished (Rodell et al. 2009; Sutanudjaja et al. 2011). The rapid population growth, expansion of irrigation agriculture, and economic development globally increased the water demand, and leads to water stress in several parts of the world (Wada et al. 2011). Groundwater levels and fluxes are controlled by a dynamic interplay between recharge and discharge, with a variety of controls and feedback loops from climate, soils, geology, land cover and human abstraction (Cuthbert et al. 2019b). According to Abbaspour et al. (2015), the quality and quantity of GW in Europe are under heavy pressure and water levels have decreased. Compared to surface water, groundwater responds more slowly to changes in meteorological conditions (Bovolo et al. 2009). As a result, the laws governing groundwater rights are still mostly static even in developed countries (Rodell et al. 2009). This aggravates overexploitation of GW worldwide and very pronouncedly in arid regions (Hashemi et al. 2015). Higher standards of living, demographic changes, land and water use policies, and other external forces are increasing the pressure on groundwater resources. In Ethiopia, 80% of the national water demand is covered from groundwater resources (Kebede 2012). The total annual aquifer recharge is estimated to be more than 30 m3 (Berhanu et al. 2014; Kebede 2010). Due to the infancy of institutes and research capabilities in the country, our data about groundwater recharge, groundwater-surface water connection, and aquifer properties is still limited (Berhanu et al. 2014). Therefore, combined use of groundwater (GW) and surface water (SW), understanding the groundwater (GW)—surface water (SW) interaction, and investigating the effect of climate change on GW are urgent issues that need to be addressed.

So far, a number of models are developed to study the interaction between GW and SW. According to Zhou and Li (2011), 1966 was a year when numerical models were applied for the first time to simulate steady state regional flow patterns of hypothetical aquifer systems. Since then, numerous physically based models have evolved (e.g., MODFLOW, Harbaugh 2005; ParFlow, Kollet and Maxwell, 2006; GSFLOW, Markstrom et al. 2008) and have been applied to improve the understanding of process dynamics. Interests of developing and using coupled land surface and subsurface models are growing. Coupling of MODFLOW with SWAT is a recent advancement. Bailey et al. (2016) developed the coupled SWAT-MODFLOW model that was upgraded later into the graphical user interface model SWATMOD-Prep (Bailey et al. 2017). Ehtiat et al. (2018) integrated SWAT, MODFLOW, and MT3DMS to investigate how SW conditions affect the quality of the GW system in a non-coastal aquifer. Similarly, Park et al. (2019) developed a QGIS-based graphical user interface SWAT-MODFLOW model for Middle Bosque River Watershed in central Texas. These advancements of coupled model development are derived due to the deficiencies associated with surface water and groundwater models. Although SWAT model has a wide range of application history, it is limited in simulating groundwater dynamics below a threshold depth of 6 m (Neitsch et al., 2011; Guzman et al. 2015; Shao et al. 2019). It uses the hydrologic response units (HRUs) as its smallest spatial computation unit where there is no exchange of water between different HRUs (Chunn et al. 2019), and the HRUs are not spatially connected (Guzman et al. 2015; Bailey et al. 2016). Additionally, SWAT has been limited with regard to groundwater dominated low flows (Pfannerstill et al. 2014). The new and fully restructured SWAT model version, SWAT + (Bieger et al. 2017), overcomes some of these limitations. For example, new landscape units have been introduced to allow for routing water e.g., from uplands to lowlands, spatial objects including aquifers, reservoirs, channels, and HRUs are independent and interactions and processes within the catchment can be connected using connect files. The MODFLOW model also has its own limitation to simulate the surface hydrologic processes because its subroutines are designed to simulate flow processes occurring in the saturated zone such as GW recharge, GW discharge, and pumping (Harbaugh 2005). Thus, application of coupled model like SWAT-MODFLOW (Bailey et al. 2016, 2017) will provide additional information such as volumetric exchange rates between the surface water and groundwater, groundwater discharge and stream seepage areas, deep percolation to the aquifer, and distributed groundwater head. Thus, use of such coupled hydrological model is important in regions where there is a knowledge gap about the groundwater-surface water interaction.

Uncertainties related to water resources management are growing due to the effect of climate change (Abbaspour et al. 2015). For this reason, research efforts are increasing to understand the influence of climate change on water resources system. Past climate change impact studies primarily focused on surface water, and only a few of them addressed the effect of climate change on groundwater (Goderniaux et al. 2009; Kidmose et al. 2013). However, there are recent advancement towards investigating the impact of climate change on groundwater (e.g., Cuthbert et al. 2019a studied the global patterns and dynamics of climate–groundwater interactions; Cuthbert et al. 2019b investigated resilience of groundwater to climate variability in in sub-Saharan Africa). Chunn et al. (2019) studied the impacts of climate change and water withdrawal on the GW-SW interactions in West-Central Alberta using the integrated SWAT-MODFLOW model. The potential impact of climate change on GW varies both temporally and spatially. Both a decrease and increase of groundwater levels are reported. Ali et al. (2012) reported that the GW system in south-western Australia was less affected by the future drier climate than the surface water system, but projected water tables are expected to decline in all areas under a drier climate where perennial vegetation was present and able to intercept recharge or use groundwater directly. An increase of the GW table is expected for future climate conditions in irrigation-dominated areas in the Oliver region of the south Okanagan, British Columbia (BC), Canada (Toews and Allen 2009). A climate change impact study by Döll (2009) revealed an increase of groundwater recharge in northern latitudes, while a 30–70% decrease is expected in semi-arid zones, including the Mediterranean, north-eastern Brazil and south-western Africa between 1961–1990 and 2041–2070. A recent study conducted by Cuthbert et al. (2019b) in sub-Saharan Africa indicated that the multiyear continuous GW levels decline in Tanzania, Namibia, and South Africa, while long-term rising trends were reported for Niger. A climate change impact study carried out in the Tekeze basin of Ethiopian by (Kahsay et al. 2018) based on the Coordinated Regional Climate Downscaling Experiment (CORDEX) Africa datasets for Representative Concentration Pathways (RCPs) of RCP 2.6 and RCP 4.5 scenarios showed decreases in the projected GW recharge by 3.4% for RCP 2.6 and 1.3% for RCP 4.5, respectively. Another climate change study by Tigabu et al. (2021) in the Lake Tana Basin, Ethiopia also highlighted that contribution of groundwater to the streamflow is anticipated to decrease for the mid-term (2031–2060) and long-term (2065–2094) future under RCP 4.5 and 8.5. Overall, study results on the effect of climate change on GW show that the magnitude and scale of influences vary from one geographic location to another, and very few studies are available for Africa in general and Ethiopia in particular. This indicates that the hydrological processes controlling groundwater recharge and sustainability, and sensitivity to climate change in Africa are poorly understood and cause high uncertainties for future water resources management and planning (Cuthbert et al. 2019b). Hence, more research efforts are required with regard to climate change impact on GW to minimize the uncertainties related to future water management and planning in Africa.

4 Statement of the Problem and Research Questions

The Lake Tana Basin has significant national importance because the government of Ethiopia has identified it as a potential area for irrigation and hydropower development, which are vital for food security and economic growth in the country (Tegegne et al. 2017). Similarly, the Lake Tana Basin is of international importance as it is a head water source of the Nile River and an area of high biodiversity (Setegn et al. 2010). As a consequence, several hydrological models from simple conceptual to more advanced physically and semi-physically based hydrological models have been applied to understand hydrological processes and the water balance of the Lake Tana Basin (Dile et al. 2018). However, most of the past hydrological studies focus on the water balance at basin outlets and lack detailed mapping of water balance components on a spatial basis. In addition, the hydrologic studies do not investigate the hydrologic mass balance in relation to vegetation types (Van Griensven et al. 2012). Although the SWAT model was used to investigate the effect of land use change on the hydrology of the basin, none of the papers explicitly addressed the specific effects of crops on the water fluxes.

According to the Ethiopian government growth and transformation plan, agriculture irrigated from GW resources is expected to cover 2 million hectares by 2020 (Kebede 2012). Due to its proximity to the point of demand, GW provides 90% of the domestic water supply, 95% of the industrial use, and a small proportion of irrigation water demand. In general, 80% of the total national water supply comes from GW (Kebede 2012). Previous hydrological studies focused on the surface water use and hydrologic balance and do not adequately address issues of GW, particularly in the context of combining SW–GW models in the Lake Tana Basin. Dile et al. (2018) reviewed past studies in the upper Blue Nile basin. According to this review, application of integrated hydrologic models is missing, as most studies have focused on single model applications to estimate a single output such as streamflow or sediment loss at the basin outlet. Consequently, the authors recommended that future research should focus on the application of coupled models to predict multiple outcomes across multiple spatial and temporal scales in the basin. Additionally, Chebud and Melesse (2009), who applied a numerical model to investigate the groundwater flow system in the Gumara catchment, identified a research gap on the spatial and temporal distribution of percolation in the Gumara catchment in particular, and in Lake Tana Basin in general. A climate change impact study carried out by Taye et al. (2015) in the Lake Tana basin reported that low flows are expected to decline by up to 61% under the A1B and B1 emission scenarios for 2050s. Consequently, it is assumed that more GW will be used to overcome the freshwater constraint in Ethiopia in general and in the Lake Tana Basin in particular. Hence, we can conclude that the demand of groundwater for agricultural, domestic, and industrial uses are expected to increase in the future. However, the GW-SW flow dynamics on the spatio-temporal basis for current and future conditions are not well understood.

In summary, studies on hydrological modeling, GW-SW interaction and climate change impacts conducted so far in Ethiopia are not conclusive and some are only conducted at the macro scale level that lack spatio-temporal analysis (Asfaw et al. 2018; Tigabu et al. 2019). Moreover, findings are not in agreement. This implies that additional research efforts are needed to advance our understanding of the spatio-temporal variability of hydrological conditions, GW-SW interaction, and climate change impacts in the Lake Tana Basin. Therefore, the attempt of this research is to answer the following research questions:

  1. 1.

    How do the major water balance components vary on spatial and seasonal basis under different agricultural crops and land use/cover classes?

  2. 2.

    Do the groundwater and surface water systems interact with each other and how does the behaviour of GW-SW interactions vary in space and time?

  3. 3.

    How will future climate change affect the major water balance components of the Lake Tana Basin?

5 Materials and Methods

5.1 General Overview of the Study Area

The Lake Tana Basin is located in the Amhara regional state of Ethiopia (Fig. 24.1). It has the largest freshwater Lake for the landlocked country, Ethiopia. Lake Tana, which is the largest fresh water lake in Ethiopia and the third largest in the Nile Basin, is situated on the basaltic Plateau of the north–western highlands of the country (Vijverberg 2009). The Lake Tana Basin is the head water source of the Blue Nile River with a total catchment area of 15,321 km2 and the lake water body covers a surface area of about 3156 km2 that accounts for 20% of the catchment area (Alemayehu et al. 2010; Tigabu et al. 2020a). It contains about 50% of the country’s fresh water (Costa et al. 2014). Around 2 million people live in the Lake Tana Basin (Vijverberg 2009). There are many rivers and streams flowing into the Lake Tana. However, only the four major rivers, Gilgelabay, Gumara, Ribb and Megech contribute around 95% of the inflow to the Lake Tana. The mean annual inflow to the lake is 158 m3/s (Alemayehu, et al. 2010), but 86% of the water originates from Gilgelabay, Gumara, and Ribb (Tigabu et al. 2020a). The Blue Nile River is the only one through which surface outflow from the lake takes place. Annually, about 4 billion m3 (127 m3/s) of water leaves the lake at the lake outlet near Bahir Dar city (Setegn et al. 2008).

Elevation and movement of the inter-tropical convergence zone (ITCZ) determine the spatial and temporal variation of rainfall in the basin. The amount of summer rainfall in the basin is controlled by the position of the ITCZ. The basin rainfall is characterized by its high seasonal difference: July, August, and September are wet months with the highest amounts of rainfall when the ITCZ position is in the northern hemisphere, June and October are transition months between wet and dry seasons, November through March belong to the dry season, and April and May are months with small rainfall amounts. Similar to the seasonal differences, the basin annual, seasonal, and monthly rainfall amounts show high spatial variability because of the diverse topography and changes in the location of the ITCZ (Woldesenbet et al. 2017). Topography has a pronounced impact on rainfall amounts in the region. The landscape varies from lowland flood plain (1700 m) to high mountain ranges (4400 m). For this reason, the basin has different climatic zones and spatially varying annual rainfall (Melesse et al. 2011). As reported by Tigabu et al. (2020a), the amount of annual rainfall is directly related to elevation above mean sea level: high rainfall is observed in the highlands, whereas low rainfall is measured in the lowlands. Moreover, large (global) atmospheric circulation and sea surface temperatures such as large scale forcing through El Niño Southern Oscillation (ENSO), Quasi-Biennial Oscillation (QBO), as well as west–east sea surface temperature gradients over the equatorial Indian Ocean are significantly influencing rainfall variability (Awange et al. 2014).

5.2 Data Availability

Daily climate data records from 1980 to 2015 of several stations were collected from the National Meteorological Service Agency (NMA 2016) and streamflow and Lake water level records for the period 1980 to 2015 were provided by the Department of Hydrology, Ministry of Water, Irrigation and Electricity of the Ethiopian Government (MoWIE 2016). Spatial data including land use and land cover and soil data at a scale of 1:50,000 were collected from the Amhara Design and Supervision Works Enterprise (ADSWE 2017). Around 60% of the basin is cultivated. From the total cultivated area, intensively cultivated, moderately cultivated (26.7%), and farm villages, account 61.7%, 26.7%, and 11.6%, respectively. The study basin has highly heterogeneous soils class. Around 50% of the total area is covered by Eutric Leptosols and 13% by Eutric Nitosols. Besides the land use/cover and soil data, the Shuttle Radar Topography Mission (SRTM) global digital elevation model (DEM) data with a 30 m by 30 m resolution was used as topography input data (USGS 2016).

5.3 Model Construction

The standalone SWAT2012 and the coupled SWAT-MODFLOW models were used to investigate the impact of agricultural crops on the major water balance components, the groundwater-surface water interaction, and climate change impacts on the hydrologic processes of the Lake Tana Basin. We constructed three independent SWAT models for the catchments of Gilgelabay, Gumara, and Ribb based on available climate, land use, soil, and digital elevation model data (Tigabu et al. 2019). The three independent model setups were used to better represent each catchment, e.g. by a more precise representation of the stream network. A step by step model parametrization was carried out for each catchment before simulating the hydrologic components. The efficiencies of the models were evaluated based on measured streamflow data.

To investigate the GW-SW interaction, the three SWAT2012 models were linked with MODFLOW using the SWATMOD-Prep interface (Bailey et al. 2017). SWAT simulates the land phase hydrologic processes including in-stream processes and soil zone processes, whereas the three-dimensional groundwater flow and all major groundwater sources and sinks (e.g. recharge, pumping, discharge to tile drains and interaction with stream networks) can be simulated using MODFLOW-NWT. The source codes of the two models were combined into a single FORTRAN executable that runs independently (Bailey et al. 2016; Abbas et al. 2018). The MODFLOW-NWT (Niswonger et al. 2011) has a Newton solver that can only be used with the Upstream Weighting (UPW) package to allow drying and re-wetting of grid cells in a transient simulation. Each SWAT model was discretized into finite difference MODFLOW grid cells with a lateral dimension of 210 m by 210 m (Tigabu et al. 2020b). In this regard, the MODFLOW boundary condition for active and inactive cells were determined based on the boundaries of the catchments.

Because of the limitation of SWATMOD-Prep in creating a MODFLOW model with a single layer, only the first unconfined layer of the aquifer system was considered (Bailey et al. 2017). Here, the unconfined aquifers were modeled under the assumption that the groundwater flows horizontally and the GW discharge is proportional to the saturated aquifer thickness (Forchheimer 1986; Dupuit 1863). Moreover, two-dimensional groundwater flow condition under no head variation in the vertical direction was assumed.

Finally, the impacts of climate change on the major hydrologic components were investigated for the mid-term (2031–2060) and long-term (2065–2094) future periods. Calibrated SWAT2012 models were used to simulate the future water balance components. Predictions were carried out using projected rainfall and temperature data of multiple regional climate models (RCMs) from the CORDEX African domain. The ensemble mean approach was used to derive a representative prediction. Prior to using the projected rainfall and temperature data, downscaling and bias-correction methods were applied (Tigabu et al. 2021).

6 Result and Discussion

6.1 Impact of Agricultural Crops on Hydrologic Processes

Groundwater recharge: comparing the rate of groundwater recharge from agricultural crops and other land use/covers, we find that cereal crops like millet, teff, wheat, and barley show relatively higher recharge rates as compared to mixed crops, pulses, rice, grassland, and forest (Fig. 24.2). The changes on groundwater recharge are also reflected in the spatial patterns indicating variable effects of different LULC. Results show significantly higher GW recharge rates for agricultural areas than forest cover. These differences of the groundwater recharge rate among different LULC classes could be caused by the variability of the soil structure, water interception and transpiration demand of different vegetation covers. Agricultural activities can also increase infiltration due to tillage of the soil (Gumindoga et al. 2014).

Fig. 24.2
A set of 3 dot plots of E T, S U R Q, and recharge versus land use land cover illustrates E T from 500 to 800, S U R Q from 0 to 500, and recharge from 400 to 800 ranges.

Variability of actual ET, surface runoff, and groundwater recharge under different LULC conditions of the Gumara Catchment as simulated using the SWAT model. The abbreviations AGRL, FPEA, FRSE, RNGE, URLD, URMD, WBAR, WETL, and WWHT stand for general agriculture, pea, forest, range grass land, light density urban settlement, medium density urban settlement, barley, wetland, and wheat, respectively

Full size image

Surface runoff: the surface runoff also shows considerable variability among the different LULC and between the three catchments. Results show that hydrologic response units (HRUs) that include medium density urban settlements, farm settlements, and mixed cropland increased the magnitude of surface runoff. The differences in its magnitude vary significantly from one LULC to another. Compared to other LULC, the annual mean runoff magnitude of corn cover is above all other LULC. The differences vary from 14 to 83 mm and all are statistically significant. Contrary to this, cereal crops such as teff, wheat, and barely show relatively low runoff (Fig. 24.2). These differences can be explained by the lower runoff curve number value of cereal crops and the high curve number of corn. When we compare runoff between the three catchments, our results show that Gumara catchment experience the highest followed by Ribb.

Actual evapotranspiration (ET): similar to groundwater recharge, ET also varies among different LULC. The annual mean values decrease from forest cover to agricultural land for all the three catchments. Among the agricultural crops, teff has the smallest amounts of average annual ET. The differences in ET values among the different crops and other LULC could be a result of differences in water uptake, available water capacity, and leaf area indices.

Apart from changes of HRUs, the water balance components reflect the patterns of the rainfall distribution in the Gilgelabay catchment. The northern and south-western part of the catchment are characterized by high amount of groundwater recharge, as a result of the highest amount of rainfall in the catchment (Fig. 24.3). In the Gumara and Ribb catchments the spatial variation of the water balance components are not influenced by rainfall as only one rainfall station was used for the simulation (see Tigabu et al. 2019), so that LULC and soil units have more influence on the spatial patterns of hydrologic components than rainfall. We found also significant seasonal variability of the hydrologic components (Tigabu et al. 2019). Because of the influence of the ITCZ position, about 80% of the rainfall falls during the summer (wet season), and this controls water availability in the catchments. More than 90% of the groundwater recharge and surface runoff and about 49% of the ET take place during the wet season. Therefore, the seasonal anomalies of the recharge and surface runoff processes in the region are largely correlated with the rainfall anomalies between the dry and wet seasons, while the overall ET variability between the two seasons is not pronounced. Base flow substantially contributes to streamflow of the Lake Tana Basin.

Fig. 24.3
A set of three maps of Gilgelabay, Ribb, and Gumara for the annual deep aquifer recharge in millimeters. The discharge ranges from 40 to 420.

Spatial variability of simulated annual deep groundwater recharge of the Gilgelabay, Ribb, and Gumara, Catchments

Full size image

In all the three catchments streamflow is dominated by base flow. Setegn et al. (2008) also found that the base flow contribution to streamflow is very high. However, Setegn et al. (2008) reported a 59% base flow contribution for Gilgelabay catchment, while 73% is found in the current study. Because of the differences of fitted model parameters between the three catchments, the base flow contribution to the streamflow varies.

Generally, the results proved that hydrologic processes (groundwater recharge, surface runoff, actual evapotranspiration) are significantly varying between cereal crops, legume plants, grassland, and forest covers. As an example, the degree of vulnerability to surface runoff is the highest in Gumara, and groundwater recharge is highest in Gilgelabay catchment. The differences are weaker between the Gumara and Ribb catchments. Furthermore, the groundwater recharge, water yield, and surface runoff values varied between the dry and wet seasons in the three catchments.

6.2 Dynamics of GW-SW Interaction

The efficiencies of SWAT-MODFLOW of the measured monthly streamflow are very good (NS >0.7) for all the three catchments. Additionally, the model was able to mimic the monthly streamflow reasonably (Fig. 24.4). Results showed that GW-SW interactions varied both in space and time. Groundwater recharge zones are mainly located in the highland areas. The potential head difference between the aquifer and the stream and other properties such as hydraulic conductivity, specific storage and initial groundwater head are the main controlling factors for the spatial variability of GW-SW interaction (Kim et al. 2008). The current study found most interactions between GW and SW during the wet season (Fig. 24.5). This high variability of the GW-SW interaction between the dry and wet months influences the water level in the rivers.

Fig. 24.4
A set of three waveform graphs of streamflow versus year for Gilgelabay, Gumara, and Ribb catchments depicts the waveforms of observed and simulated streamflow. Each graph is divided into calibration and validation sections.

Comparison of simulated and observed monthly streamflow of the Gilgelabay, Gumara, and Ribb Catchments

Full size image
Fig. 24.5
A set of three graphs of the G W S W interaction versus months for Gilgelabay, Gumara, and Ribb catchments.

Seasonal differences of the GW-SW interaction at the catchment outlet points for Gilgelabay, Gumara, and Ribb catchments

Full size image

In Gumara and Ribb the exchange between aquifer and stream is bidirectional for all time steps (Fig. 24.5). The Gumara Catchment is largely controlled by the surrounding surface water system. In the downstream floodplain and reaches groundwater is discharged to the stream system. Dessie et al. (2014) reported that the floodplain of the Gumara Catchment was recharged by the groundwater during the rainy season. Chebud and Melesse (2009) also reported that 0.09 billion cubic meter of water flow from floodplain aquifer to the Lake Tana during the wet season. The bidirectional flow of the Ribb Catchment is more or less similar to the Gumara Catchment. Since the early 1990s the temporal patterns showed significant changes for the entire year and during the dry season (magnitude and flow direction) for the two catchments (Fig. 24.6). Before the early 1990s the flow direction was from the stream to the aquifer and changed later. This decrease of flow from the river network to the aquifer and the change of direction from the aquifer to the stream system (annual and dry season) over time could be related to the increase of runoff. Recently the number of rainy days in a year are decreasing while the total rainfall amount did not change significantly, which in turn leads to more intensive rainfall and results in more surface runoff generation and less seepage to the aquifer system. Another notable results from this investigation are the strong dynamics of the exchange fluxes between the aquifer and the stream system on the daily time step (Fig. 24.5). The high variability of the daily and seasonal GW-SW exchange rates might be attributed to a large variation in daily rainfall (coefficient of variation ranging from 100 to 200%).

Fig. 24.6
A set of 3 line graphs of G W S W interaction versus year for Gilgelabay, Gumara, and Ribb catchments. It depicts the curve of G W-S W becoming nearly constant after 1990 in the Gumara catchment.

Temporal dynamics of groundwater-surface water interaction of river reaches at the outlet subasins of Gilgelabay, Gumara, and Ribb Catchments

Full size image

Overall, we found a strong hydraulic connection between GW-SW systems and all the three catchments are characterized by a considerable spatio-temporal variability within and between the catchments. The annual temporal patterns and direction of the water fluxes in the Gilgelabay Catchment differs from the ones in the Gumara and Ribb Catchments where negligible seepage from the stream was found.

6.3 Impact of Climate Change on Water Balance Components

Climate change is likely to result in an alteration of the global hydrological cycle. Results showed that the mean annual temperature is expected to increase significantly while the ensemble annual mean rainfall from regional climate models shows only minor changes for the mid-term (2031–2060) and long-term (2065–2094) compared to the baseline period from 1985 to 2014 (Fig. 24.7). The increase of temperature in turn is expected to cause flow regime change.

Fig. 24.7
A set of two boxplots of delta rainfall and temperature versus the baseline period. The baseline periods are mid-R C P 4.5, mid-R C P 8.5, long R C P 4.5, and long-R C P 8.5.

Boxplots showing expected percentage changes on the ensemble mean annual rainfall and temperature with reference to the annual mean values of the baseline period in the case of the Gumara Catchment under RCPs 4.5 and 8.5 for the mid- and long-term periods

Full size image

With rising temperatures, one may expect more potential evapotranspiration (PET) and actual evapotranspiration (ET), which in turn could cause a drop of the groundwater contribution (GWQ) to the streamflow. Here, we analyse the model results for the major water balance components with the Mann Kendall trend test and the Tukey multiple mean comparison tests applied on the annual values of each hydrologic component.

The future annual mean temperature is expected to rise significantly (p-value<0.05) for both time windows and RCPs, which in turn will cause a significant increase of PET. To the contrary, our results show a decrease of the annual ET. The projected decrease is not statistically significant for Gumara, while it is significant for the Gilgelabay catchment (Table 24.1). The groundwater discharge (GWQ) is also anticipated to drop significantly (p-value <0.05) in Gumara (Fig. 24.8, Table 24.1), while no significant (p-value > 0.05) decrease is expected in Gilgelabay except for the mid-term of RCP4.5. Significant increases are expected for surface runoff (SURQ) in both catchments. The expected changes of SURQ are increasing even under conditions of slightly lower rainfall. But the number of projected rainy days will decrease in both catchments. Future rainfall is expected to be more intense than the rainfall during the baseline period (Fig. 24.9). As a result, more surface runoff is expected. Thus, the expected increase of runoff will have different implications on the other water balance components. Except for the mid-term of RCP4.5 in Gilgelabay, the groundwater contribution to the streamflow is expected to decrease for both catchments and both RCPs. The decrease in the Gumara Catchment is significant, whereas it is not significant in Gilgelabay. ET was expected to increase with increasing temperatures. However, we found a decrease of ET in both catchments. This decline was observed even for positive changes of the projected rainfall because water availability limited the responses to warming (Condon et al. 2020; Wagner et al., 2015). About 51% of ET takes place during the dry season (October to May), when less water is available. Expected changes of water yield are not significant as the increase of surface runoff is compensated by the decrease of GWQ in both catchments. Moreover, we found a decrease of annual percolation and soil moisture that could result in a drop of the water table of the shallow aquifer. Thus, a decrease of the water table and soil moisture in turn would cause a reduction of the groundwater discharge and evapotranspiration. Setegn et al. (2011) also reported a decrease on the future GWQ under A1B and B scenarios outputs from GFDL and MPI.

Table 24.1 Changes in the projected ensemble mean rainfall and temperature resulting from different regional climate models (RCMs) and SWAT simulated major water balance components with reference to the baseline period for the Gumara and Gilgelabay catchments
Full size table
Fig. 24.8
A set of two boxplots of delta G W Q and S U R Q versus the baseline period. The baseline periods are mid-R C P 4.5, mid-R C P 8.5, long R C P 4.5, and long-R C P 8.5.

Boxplots showing the projected percentage changes of groundwater discharge (GWQ) and surface runoff (SURQ) in Gumara catchment under RCPs 4.5 and 8.5 for the mid- and long-term periods compared to the baseline period

Full size image
Fig. 24.9
A set of two line graphs of rainfall intensity versus months for Gumara and Gilgelabay depicts the curves of baseline, mid-R C P 4.5, mid R C P 8.5, long R C P 4.5, and long R C P 8.5.

Expected changes in rainfall intensity for baseline, mid-term and long-term periods under RCP4.5 and RCP8.5 concentration pathways

Full size image

In general, the small changes on rainfall in the scenario data will result in a significant change of the major hydrologic components. The temperature effect seems smaller than the rainfall effect. However, potential evapotranspiration (PET) is significantly higher under the two RCPs. Overall, the significant increase of the projected temperature and the small changes of the rainfall amount and intensity are expected to change the major water balance components. Surface runoff and groundwater contribution to the streamflow are expected to change in opposite directions. Here, increased surface runoff primarily contributes to a reduction of groundwater contribution to the streamflow.

7 Conclusion

This research provides an additional understanding and new information about the present and future hydrologic perspectives of the Lake Tana Basin. The topics covered include impacts of agricultural crops on hydrology, groundwater and surface water interaction, and impact of climate change on water resource in the Lake Tana Basin. The physically based hydrologic models SWAT and SWAT-MODFLOW were set up to model the present and future hydrologic components in the Lake Tana Basin. The models were calibrated and validated with measured streamflow data from the Gilgelabay, Gumara, and Ribb catchments with satisfactory results. The key findings are:

The models SWAT and SWAT-MODFLOW help to understand the spatial and seasonal patterns of the hydrologic components under different LULC and climate change conditions. Modeled results revealed a substantial contribution of groundwater to the hydrologic systems in the three catchments. Moreover, the different agricultural crops had different influences on spatial and seasonal patterns of hydrological variables. While some e.g., cereal crops enhanced groundwater recharge, others such as legume plants increased surface runoff. Therefore, the future water availability is also affected by the types of cultivated crops.

The coupled hydrologic models indicated that the surface water and groundwater systems interacted with each other with a high spatio-temporal variability. The anomalies were significantly higher on daily time step than monthly or yearly. We used an ensemble average of 35 rainfall and temperature outputs from 35 regional climate model to minimize uncertainties associated with individual RCM (Kling et al. 2012).

Understanding the availability of water resources under global climate change conditions can provide new insights for policy makers and water managers. Our intention was to understand how the major water balances components respond to future rainfall and temperature changes. The mean temperature is expected to rise higher than the global average, while the rainfall is not expected to change significantly. We found that, the groundwater contribution to the streamflow and actual evapotranspiration tend to decrease whereas the surface runoff is likely to increase. Hence, appropriately accounting for the major water balance components in the context of climate change is of great importance when considering the future management and planning of water resources. Secondly, it is highly advisable to take appropriate measures to mitigate climate change impacts, e.g., implement water conservation measures, to store the expected increase of surface runoff and minimize soil erosion.

Overall, the hydrologic modeling methods and results that were based on specific agricultural crop cover and different climate scenarios have diverse applicability for similar areas worldwide that are dominantly influenced by agricultural crops to design proper future water management plans. Moreover, results from the coupled hydrologic model also enhance our level of understanding on the flow dynamics of groundwater and surface water interaction on different temporal scales in the current study area and it helps to explain the interconnected hydrologic processes in other catchments in Ethiopia and Africa. Thus, the methodological approaches followed in this study could be transferred to such other regions.

Use of hydrological research outputs for decision making processes requires understanding the level of uncertainty associated with applied research methodology, quality and available input data, and hydrologic model parameters. Uncertainties associated with hydrologic modeling results in data scarce regions like that of the Lake Tana Basin are expected to be high level. We minimized uncertainties associated with the hydrologic model parameters through model calibration and validation using the available limited data. Nevertheless, we believe that further improvement could be possible by enhancing the database through regular measurement of groundwater table elevation, withdrawal, and stream stage values.

In general, results proved that the hydrology of the region was changed overtime due to an increase of area and production intensity of agricultural systems, which in turn was driven by the rapid population growth. These changes are expected to be escalated in the future due to climate change unless appropriate agricultural practice and water resources management are designed to cope with the anticipated changes. In this research, crop types were identified according to their impact on groundwater recharge, surface runoff generation, and evaporative demands, which would be a good insight to design proper future agricultural and water management practices in the region and other similar areas that are under the pressure of intensive agriculture. The hydrologic models were developed considering agricultural crop variability considering their topographic variations and growing suitability of individual crops. They show up significant variances of influences on the hydrologic cycle. Thus, it is suggested to transfer this methodological approach to similar regions that are under rapid population growth pressure and agricultural intensification. Moreover, our findings showed that the groundwater and surface water systems are connected to each other. Hence sustainable water resource management requires to consider both, groundwater, and surface water resources.