ABSTRACT
Using an ensemble of three global climate models, the current study aims to estimate climate change and quantify the changes in the aridity and evapotranspiration of two distinct areas in Ethiopia. To adjust for bias in the climate dataset, the Hydrological Modeling Tool (CMhyd) was used. These studies were initially run using station data and employed the shared socioeconomic pathways scenarios for the short-range (2011–2040) years and medium-range (2041–2070) years. Climate-related aridity is measured using the De Martone and United Nations Environment Programme aridity indices. The study's findings for the western and eastern catchments for the reference period (1981–2010) years reveal an average annual temperature rise of 1.5 and 0.06 °C and a drop in annual precipitation of 15.73 and 3.68 mm/year, respectively. These alter the climate in the geographical areas that have historically supported drought. Evapotranspiration in the western and eastern catchments may grow by up to 24.6 and 21.6%, respectively, by the 2070s. The observation implies that the western catchment may experience more pronounced climate change and volatility than the eastern one. The consequences of this observation influence agriculture, water resource management, and the social and economic well-being of those living in drought-prone areas.
HIGHLIGHTS
Using CMIP6 climate models.
Comparing two hydro-climate regions.
Using two basic aridity index parameters.
Evapotransportation estimation.
Semi-arid area analysis.
INTRODUCTION
Nowadays, effective management of the natural environment and creating efficient plans for making reliable judgments about upcoming climate uncertainty becomes more and more significant for the scientific world. The continuous alteration of the usual atmospheric conditions that characterize local, regional, and global climates is referred to as global warming (Naz et al. 2022). Changes and oscillations in the climate may have a significant influence on the ecosystem and many areas of our society. Long-term or short-term climatic conditions with human or natural origins have an impact on the total water resources, both the surface and underground water (Mancosu et al. 2015; Alemayehu et al. 2016).
Scholars have begun a variety of programs to decrease the environmental impact of these developments. Currently, the most modern global circulation models are being utilized to explain the present and future implications of climate change and direct policy. The Coupled Model Intercomparison Project (CMIP) has continued to release new CMIP model results since it began in 2013 (https://esgf-node.llnl.gov/projects/cmip6/). When compared to CMIP5, CMIP6 produces data with greater geographical resolution and better algorithms.
Examples of climatic phenomena that aid in our understanding of the variation of crucial climate parameters (precipitation and temperature) are aridity and drought. Aridity is a long-term climatic state in contrast to drought, which is a transient, short-term weather phenomenon that may occur in a range of climates (Beštáková et al. 2023). The aridity indices (AIs) of an area influence how variable its water resources are (Haile et al. 2022). The carrying capacity of ecosystems is negatively impacted by the imbalance in water availability that is brought on by dryness. Low overall moisture, considerable temporal and geographic variability, and a low yearly average rainfall all contribute to this disparity (Proutsos et al. 2021).
A variety of metrics may be used to measure the aridity of the climate. However, one index has done well in a certain area and climatic zone, while another has done well in other areas. This indicates that the indices vary in terms of the data's accessibility and the weather. De Martone's AI, as well as the United Nations Environment Programme (UNEP) AI (modified Thornthwaite approach) (Tabari & Aghajanloo 2013), were used among the AIs. These indices are ideal for examining the climatic conditions in semi-arid locations because there are no many reliable data points available. The other indicator for measuring precise climatic and bioclimatic classifications is potential evapotranspiration (PET) (Tu & Yang 2022). Studies of meteorology and hydrology that are used to calculate the actual water consumption rate of different crops must consider PET (Senatilleke et al. 2022). Understanding how climatic variability impacts hydrology requires understanding the effects of climate change on conventional evapotranspiration (ETo) (Milly & Dunne 2016; Ajjur & Al-Ghamdi 2021; Singer et al. 2021; Stefanidis & Alexandridis 2021; Al-Hasani & Shahid 2022).
Ethiopia has gone through several severe and prolonged dry spells in recent years (Edossa et al. 2010; Library et al. 2017; Kassaye et al. 2021). In comparison to other climate disasters, drought is regarded as one of the most harmful extreme climatic occurrences (Edossa et al. 2010; Haile et al. 2022). According to the research, climate change is the main factor contributing to Ethiopia's drought and famine (Kloos 1982; Zegeye 2018; Peng et al. 2021). According to Wilhite and others (2007), both Ethiopia's surface water resources and subsurface water supplies are susceptible to the effects of climate change. One of the most pressing requirements for increasing the accessibility of water resources is expectations of future events, particularly the direction and intensity of environmental change. The effects of the drought and its primary causes in various regions of Ethiopia have been the subject of extensive research (Gebremedhin et al. 2018b; Tadesse et al. 2018; Abrha & Hagos 2019; Meaza et al. 2022).
The research area is located in semi-arid regions in western and eastern Ethiopia having huge climate variability and potable water scarcity. Gambela plain (western) and Fafen-Jerer and Shinile subbasin (eastern) Ethiopia are the focus of this particular study. Studies (Gebremedhin et al. 2018a; Tadesse et al. 2018; Alemayehu et al. 2020) have confirmed the existence of drought in the aforementioned region. However, these studies focused on the socioeconomic aspects of a few small semi-arid regions and lacked detailed climate analysis and the application of recent global climate model (GCM) outputs. Furthermore, it is unknown and poorly documented how the prevalence of climatic aridity varies through time and space in the research locations. Before now, there has not been regional research that has established the expected climate of the western and eastern catchments. These were the driving forces for our investigation into the risk assessment of climatic droughts in the research locations and our projection of potential outcomes. This study differs from previous comparable studies in that it considers two catchments with unique topo-ecological conditions, land-use patterns, and human lifestyles. Comparisons across different scenarios highlight the importance of developing climate change mitigation policies to stop droughts from occurring more frequently, staying longer, and increasing. Due to this, a study was conducted to examine the variation of AIs and PET using CMIP6-GCM data for the reference period (1981–2010) years and the short- and mid-term periods (2011–2040 and 2041–2070), and the study used the update of Representative Concentration Pathway (RCP)4.5 and 8.5 based on Shared Socioeconomic Pathways (SSP)2 and SSP5. Researchers, governmental organizations, or stakeholders will utilize this study in the future to design preventative strategies for drought, land degradation, and desertification anticipation and mitigation efforts. It will also assist water resource planners and policymakers in taking action to understand the future effects of climate change.
MATERIALS AND METHODS
The purpose of this study is to determine the PET and the dryness of the climate in the reference period and future epochs using temperature and precipitation data. Because of this, the primary goal of the current study is to assess the high-resolution climatological data downscaled from seven GCMs while showcasing the expected changes in climate.
The study area
The topographic elevation and meteorological station locations: (a) western catchment and (b) eastern catchment.
The topographic elevation and meteorological station locations: (a) western catchment and (b) eastern catchment.
The eastern catchment spans 90,319 km2 in eastern Ethiopia and extends from north to south. Geographically, the eastern catchment extends from 41°06′ to 45° 05′ east and from 6°15′ to 11°02′ north. The region encompasses the Fafen-Jerer and Shinile sub-basins. The primary issue is climate fluctuation due to the semi-arid environment of the eastern basin. Demographically, the majority of the territory is part of Ethiopia's Somali regional state. The altitude range is 360 to 3,000 m above the mean sea level (Figure 1(b)). Harar, Dire Dawa, and Jigjiga are the major cities and regional capitals in the area.
Climatic data
The National Meteorological Institute (NMI) gathered the historical daily climatic data that are currently accessible. There were a total of 20 weather stations that had a daily record for the years that were accessible. Some of the stations lack data on maximum and minimum temperatures as well as values for the past recorded daily precipitation. The station's nearest neighbor approaches and gridded datasets were used to fill up the data gaps (https://power.larc.nasa.gov/data-access-viewer/). The GCM output was utilized to analyze the climate in the future. The data of the GCM were obtained from the website World Climate Research Program (https://esgf-node.llnl.gov/projects/cmip6/). For the short-range (2011–2040) years and mid-range (2041–2070) years, the GCM dataset from CMIP of the IPCC's sixth assessment report subsection RCPs (SSP2-4.5 and SSP5.8.5) was utilized. Climate change scenarios from GCM are generally large and need to be scaled down to obtain relevant information for areas of interest (Ashfaq et al. 2022; Shiru et al. 2022; Wang & Tian 2022).
To accurately evaluate the regional consequences of climate change on precipitation and temperature, downscaling the GCM from CMIP6 results to a higher geographical resolution is necessary. The climatic data from the GCM model were downscaled, and these biases were corrected using the CMhyd model (climatic Model Data for Hydrological Modeling) (Panahi et al. 2022). The model included bias correction techniques, linear scaling, delta change correction, local scaling of precipitation intensity, conversion of precipitation power, scaling of temperature dispersion, and distribution mapping of precipitation and temperature. In this study, linear scaling bias correction was applied. In many different applications, the CMhyd model is frequently used to correct temperature and precipitation biases (Haider et al. 2021; Hordofa et al. 2021; Yeboah et al. 2022).
Data analysis
The daily temperature and precipitation outputs from the three GCMs (CMCC-ESM2, FGOALS-g3, and MIROC6) were integrated for each study by averaging the various datasets. The reference period (1981–2010) years, the present and forthcoming (2011–2040) years, and the mid-range (2041–2070) years for each region are also compared in the study. In this study, potential ETo was calculated using the Thornthwaite approach (Thornthwaite 1948; Proutsos et al. 2021), and aridity was assessed using the De Martone and UNEP AIs.
Potential evapotranspiration
Aridity indices
Numerous studies have demonstrated that the dryness index, sometimes known as humidity, can accurately depict both humid and dry climatic conditions (Şarlak & Mahmood Agha 2018; Pellicone et al. 2019; Beštáková et al. 2023). The degree to which a climate lacks sufficient, life-supporting moisture is referred to as aridity, which in the context of climate is the opposite of humidity. In this particular study, two geographically separate areas of Ethiopia were examined using the indices listed below.
De Martone index
The De Martone AI is used to categorize the climate of a region (Table 1) (Gavrilov et al. 2019; Pellicone et al. 2019).
The climate classification according to De Martone aridity and UNEP aridity index value (UNEP 1992; Baltas 2007)
Index . | Type of climate . | Index value . |
---|---|---|
De Martone aridity index | Arid | DMI < 10 |
Semi-arid | 10 ≤ DMI < 20 | |
Mediterranean | 20 ≤ DMI < 24 | |
Semi-humid | 24 ≤ DMI < 28 | |
Humid | 28 ≤ DMI < 35 | |
Very humid | 35 ≤ DMI ≤ 55 | |
Extremely | DMI > 55 | |
UNEP aridity index | Hyper-arid | <0.05 |
Arid | 0.05_0.20 | |
Semi-arid | 0.20_0.50 | |
Dry subhumid | 0.50_0.75 | |
Humid | >0.75 |
Index . | Type of climate . | Index value . |
---|---|---|
De Martone aridity index | Arid | DMI < 10 |
Semi-arid | 10 ≤ DMI < 20 | |
Mediterranean | 20 ≤ DMI < 24 | |
Semi-humid | 24 ≤ DMI < 28 | |
Humid | 28 ≤ DMI < 35 | |
Very humid | 35 ≤ DMI ≤ 55 | |
Extremely | DMI > 55 | |
UNEP aridity index | Hyper-arid | <0.05 |
Arid | 0.05_0.20 | |
Semi-arid | 0.20_0.50 | |
Dry subhumid | 0.50_0.75 | |
Humid | >0.75 |
UNEP aridity index
RESULTS AND DISCUSSION
This section presents climate analysis results for two districts' catchments based on baseline and future periods SSP2-RCP4.5 and SSP5-RCP8.5 scenarios. Climate change indicators are defined as the difference between the periods 2011–2040, 2041–2070, and 1981–2010.
Model selection and future climate projection
It is subjective and difficult to select a specific model for a specific area of study. In this particular study, seven GCM models are evaluated for the climate projection of two district catchments. The analysis confirms that most of the model works of the study region climate project, Euro-Mediterranean Centre on Climate Change with Climate and Earth System Models (CMCC-ESM2), the Earth System Model EC-Earth3 with carbon cycle (EC-Earth3-CC), Flexible Global Ocean-Atmosphere-Land System Model Grid-Point Version 3 (FGOALS-g3), Model for Interdisciplinary Research on Climate version 6 (MIROC6), and Max Planck Institute Earth System Model (MPI-ESM1-2-HR), Meteorological Research Institute Earth System Model Version 2.0 (MRI-ESM2-0), and NUIST Earth System Model version 3 (NESM3) were evaluated (Table 2) models with meteor observations of 20 stations data in the study regions for the reference period (1981–2010). To select the best model projection, the study used model evaluation parameters: correlation coefficient (R2), Nash–Sutcliffe efficiency (NSE), percent bias (PBIAS), and root-mean-square error (RMSE). The results showed that CMCC-ESM2, FGOALS-g3, and MIROC6 have the best agreements with the study region's meteorological observation (Table 3) for both precipitation and temperature datasets. We used the bias-corrected average of these three models' outputs for both SSPs. Other similar studies in regions confirm that CMIP6-GCM models are effective in projecting future climate in Ethiopia (Alaminie et al. 2021, 2023; Balcha et al. 2022). This is the first attempt in the study regions for a climate projection that can be modified with additional scenarios and an extended estimation period.
The GCMs applied, the institution, and the sources
No . | GCM . | Institution . | Resolution (lon.* lat., deg) . | References . | Remark . |
---|---|---|---|---|---|
1 | CMCC-ESM2 | Euro-Mediterranean Centre on Climate Change, Italy | 1°*1° | Lovato et al. (2022) | Selected |
2 | EC-Earth3-CC | European EC-Earth Consortium | 0.7°*0.7° | Farhat et al. (2022) | |
3 | FGOALS-g3 | Chinese Academy of Sciences, China | 2°*2.3° | Li et al., (2021) | Selected |
4 | MIROC6 | Japan Agency for Marine-Earth Science and Technology, Japan | 1.4°*1.4° | Tatebe et al. (2019) | Selected |
5 | MPI-ESM1-2-HR | Max Planck Institute for Meteorology, Germany | 0.94°*0.94° | Müller et al. (2018) | |
6 | MRI-ESM2-0 | Meteorological Research Institute, Tsukuba, J + C34apan | 1.1°*1.1° | Yukimoto et al. (2019) | |
7 | NESM3 | Nanjing University of Information Science and Technology, China | 1.9°*1.9° | Yang et al. (2020) |
No . | GCM . | Institution . | Resolution (lon.* lat., deg) . | References . | Remark . |
---|---|---|---|---|---|
1 | CMCC-ESM2 | Euro-Mediterranean Centre on Climate Change, Italy | 1°*1° | Lovato et al. (2022) | Selected |
2 | EC-Earth3-CC | European EC-Earth Consortium | 0.7°*0.7° | Farhat et al. (2022) | |
3 | FGOALS-g3 | Chinese Academy of Sciences, China | 2°*2.3° | Li et al., (2021) | Selected |
4 | MIROC6 | Japan Agency for Marine-Earth Science and Technology, Japan | 1.4°*1.4° | Tatebe et al. (2019) | Selected |
5 | MPI-ESM1-2-HR | Max Planck Institute for Meteorology, Germany | 0.94°*0.94° | Müller et al. (2018) | |
6 | MRI-ESM2-0 | Meteorological Research Institute, Tsukuba, J + C34apan | 1.1°*1.1° | Yukimoto et al. (2019) | |
7 | NESM3 | Nanjing University of Information Science and Technology, China | 1.9°*1.9° | Yang et al. (2020) |
Note: ‘Selected’ in the remark is to state that the data from the three models were applied for the projected climate parameters analysis with both RCP4.5 and RCP8.5 cases.
Monthly observations of precipitation of the GCM models validation with ground observations (1981–2010)
Statistical analysis . | . | Correlation coefficient . | RMSE . | R2 . | PBIAS . | NSE . |
---|---|---|---|---|---|---|
Model | ||||||
CMCC-ESM2 | 0.99 | 13.33 | 0.98 | 0.00 | 0.84 | |
EC-Earth3-CC | 0.94 | 13.86 | 0.88 | 0.00 | 0.69 | |
FGOALS-g3 | 0.98 | 18.72 | 0.97 | 0.00 | 0.84 | |
MIROC6 | 0.99 | 28.46 | 0.98 | 0.00 | 0.83 | |
MPI-ESM1-2-HR | 0.97 | 25.29 | 0.94 | 0.00 | 0.77 | |
MRI-ESM2-0 | 0.99 | 14.58 | 0.97 | 0.00 | 0.83 | |
NESM3 | 0.67 | 33.64 | 0.50 | 0.01 | −0.18 |
Statistical analysis . | . | Correlation coefficient . | RMSE . | R2 . | PBIAS . | NSE . |
---|---|---|---|---|---|---|
Model | ||||||
CMCC-ESM2 | 0.99 | 13.33 | 0.98 | 0.00 | 0.84 | |
EC-Earth3-CC | 0.94 | 13.86 | 0.88 | 0.00 | 0.69 | |
FGOALS-g3 | 0.98 | 18.72 | 0.97 | 0.00 | 0.84 | |
MIROC6 | 0.99 | 28.46 | 0.98 | 0.00 | 0.83 | |
MPI-ESM1-2-HR | 0.97 | 25.29 | 0.94 | 0.00 | 0.77 | |
MRI-ESM2-0 | 0.99 | 14.58 | 0.97 | 0.00 | 0.83 | |
NESM3 | 0.67 | 33.64 | 0.50 | 0.01 | −0.18 |
Historical and projected analysis of climate variables
Precipitation and temperature
Temperature distribution in (a) western and (b) eastern catchment for the baseline period (1981–2010).
Temperature distribution in (a) western and (b) eastern catchment for the baseline period (1981–2010).
The baseline (1981–2010) precipitation distribution in (a) western and (b) eastern catchment.
The baseline (1981–2010) precipitation distribution in (a) western and (b) eastern catchment.
Mean air temperature (°C) (line graph) and average monthly precipitation (mm) (bar graph) for 1981–2010 for the three specific catchments.
Mean air temperature (°C) (line graph) and average monthly precipitation (mm) (bar graph) for 1981–2010 for the three specific catchments.
The trend precipitation and temperature (1981–2010) years
Historical analysis of precipitation and temperature: (a) western catchment and (b) eastern catchment.
Historical analysis of precipitation and temperature: (a) western catchment and (b) eastern catchment.
Projected precipitation and temperature trends
The change of projected precipitation and temperature in the catchments relative to the baseline period
. | . | Scenarios . | |||
---|---|---|---|---|---|
. | . | Western catchment . | Eastern catchment . | ||
Mean precipitation (mm/year) . | Time . | RCP4.5 . | RCP8.5 . | RCP4.5 . | RCP8.5 . |
Baseline | 1981–2010 | 1,588.35 | 1,588.35 | 595.32 | 595.32 |
Short range | 2015–2040 | 1,743.52 | 1,803.05 | 668.70 | 674.01 |
Mid-range | 2041–2070 | 1,820.95 | 1,907.86 | 698.89 | 755.09 |
Change short range (%) | 9.77 | 13.52 | 12.33 | 13.22 | |
Change mid-range (%) | 14.64 | 20.12 | 17.40 | 26.84 | |
. | . | Scenarios . | |||
. | . | Western catchment . | Eastern catchment . | ||
Mean precipitation (mm/year) . | Time . | RCP4.5 . | RCP8.5 . | RCP4.5 . | RCP8.5 . |
Baseline | 1981–2010 | 23.37 | 23.37 | 22.38 | 22.38 |
Short range | 2015–2040 | 23.95 | 24.01 | 22.74 | 22.93 |
Mid-range | 2041–2070 | 24.63 | 25.09 | 23.51 | 24.09 |
Change short range (%) | 2.51 | 2.76 | 1.62 | 2.44 | |
Change mid-range (%) | 5.40 | 7.39 | 5.05 | 7.64 |
. | . | Scenarios . | |||
---|---|---|---|---|---|
. | . | Western catchment . | Eastern catchment . | ||
Mean precipitation (mm/year) . | Time . | RCP4.5 . | RCP8.5 . | RCP4.5 . | RCP8.5 . |
Baseline | 1981–2010 | 1,588.35 | 1,588.35 | 595.32 | 595.32 |
Short range | 2015–2040 | 1,743.52 | 1,803.05 | 668.70 | 674.01 |
Mid-range | 2041–2070 | 1,820.95 | 1,907.86 | 698.89 | 755.09 |
Change short range (%) | 9.77 | 13.52 | 12.33 | 13.22 | |
Change mid-range (%) | 14.64 | 20.12 | 17.40 | 26.84 | |
. | . | Scenarios . | |||
. | . | Western catchment . | Eastern catchment . | ||
Mean precipitation (mm/year) . | Time . | RCP4.5 . | RCP8.5 . | RCP4.5 . | RCP8.5 . |
Baseline | 1981–2010 | 23.37 | 23.37 | 22.38 | 22.38 |
Short range | 2015–2040 | 23.95 | 24.01 | 22.74 | 22.93 |
Mid-range | 2041–2070 | 24.63 | 25.09 | 23.51 | 24.09 |
Change short range (%) | 2.51 | 2.76 | 1.62 | 2.44 | |
Change mid-range (%) | 5.40 | 7.39 | 5.05 | 7.64 |
Projected precipitation and temperature of western catchment: (a) SSP2-RCP4.5 and (b) SSP5-RCP8.5.
Projected precipitation and temperature of western catchment: (a) SSP2-RCP4.5 and (b) SSP5-RCP8.5.
Projected precipitation and temperature of eastern catchment: (a) SSP2-RCP4.5 and (b) SSP5-RCP8.5.
Projected precipitation and temperature of eastern catchment: (a) SSP2-RCP4.5 and (b) SSP5-RCP8.5.
Table 5 shows that with p-values of 0.0966 and 0.0033, respectively, the Mann–Kendall test findings show that significant precipitation takes place in the eastern and western catchments under SSP2-RCP4.5. As a result, we can say that there is a clear increasing pattern in the western basin. In both catchments, the precipitation of SSP2-RCP8.5 showed p-values of 0.0025 and 0.0048, which is predicted to show that both catchments show an increase in precipitation. Accordingly, SSP2-RCP4.5 temperature indicates no significant increase in both basins. In the western catchment, the temperature increases significantly (p = 0) under the SSP2-RCP8.5 scenario, while the eastern catchment shows no significant rise in temperature.
Mann–Kendall trend test parameters for baseline and predicted precipitation and temperature, PET, AI, and DMI
Basin . | Eastern . | Western . | ||||||
---|---|---|---|---|---|---|---|---|
Time . | 1981–2010 . | 2015–2070 . | 1981–2010 . | 2015–2070 . | ||||
Models . | Average . | RCP4.5 . | RCP8.5 . | Average . | RCP4.5 . | RCP8.5 . | ||
SSP scenarios . | Historical . | SSP2 . | SSP5 . | Historical . | SSP2 . | SSP5 . | ||
Mann–Kendall trend test result | Precipitation | P | 0.42500 | 0.09660 | 0.00250 | 0.16030 | 0.00330 | 0.00480 |
Trend | No trend | No trend | Increasing | No trend | Increasing | Increasing | ||
S | 1.32500 | 1.19800 | 2.68000 | 5.56000 | 3.24000 | 3.73000 | ||
Temperature | P | 0.06300 | 2.40030 | 4.44000 | 0.00028 | 1.75410 | 0.00000 | |
Trend | No trend | Increasing | Increasing | Increasing | Increasing | Increasing | ||
S | 0.01060 | 0.02800 | 0.04260 | 0.02690 | 0.02400 | 0.03800 | ||
PET | P | 0.00200 | 1.03890 | 2.24200 | 0.00061 | 1.22000 | 1.22000 | |
Trend | Increasing | Increasing | Increasing | Increasing | Increasing | Increasing | ||
S | 1.70160 | 3.64000 | 5.91000 | 5.50050 | 5.08800 | 5.08000 | ||
AI | P | 0.26800 | 0.24930 | 0.99400 | 0.09420 | 0.80460 | 0.15100 | |
Trend | No trend | No trend | No trend | No trend | No trend | No trend | ||
S | 0.00140 | 0.00079 | 2.57000 | 0.00650 | 0.00021 | 0.00160 | ||
DMI | P | 0.26800 | 0.16800 | 0.16800 | 0.10820 | 0.15540 | 0.22680 | |
Trend | No trend | No trend | No trend | No trend | No trend | No trend | ||
S | 0.12500 | 0.04500 | 0.04500 | 0.18340 | 0.04800 | 0.04640 |
Basin . | Eastern . | Western . | ||||||
---|---|---|---|---|---|---|---|---|
Time . | 1981–2010 . | 2015–2070 . | 1981–2010 . | 2015–2070 . | ||||
Models . | Average . | RCP4.5 . | RCP8.5 . | Average . | RCP4.5 . | RCP8.5 . | ||
SSP scenarios . | Historical . | SSP2 . | SSP5 . | Historical . | SSP2 . | SSP5 . | ||
Mann–Kendall trend test result | Precipitation | P | 0.42500 | 0.09660 | 0.00250 | 0.16030 | 0.00330 | 0.00480 |
Trend | No trend | No trend | Increasing | No trend | Increasing | Increasing | ||
S | 1.32500 | 1.19800 | 2.68000 | 5.56000 | 3.24000 | 3.73000 | ||
Temperature | P | 0.06300 | 2.40030 | 4.44000 | 0.00028 | 1.75410 | 0.00000 | |
Trend | No trend | Increasing | Increasing | Increasing | Increasing | Increasing | ||
S | 0.01060 | 0.02800 | 0.04260 | 0.02690 | 0.02400 | 0.03800 | ||
PET | P | 0.00200 | 1.03890 | 2.24200 | 0.00061 | 1.22000 | 1.22000 | |
Trend | Increasing | Increasing | Increasing | Increasing | Increasing | Increasing | ||
S | 1.70160 | 3.64000 | 5.91000 | 5.50050 | 5.08800 | 5.08000 | ||
AI | P | 0.26800 | 0.24930 | 0.99400 | 0.09420 | 0.80460 | 0.15100 | |
Trend | No trend | No trend | No trend | No trend | No trend | No trend | ||
S | 0.00140 | 0.00079 | 2.57000 | 0.00650 | 0.00021 | 0.00160 | ||
DMI | P | 0.26800 | 0.16800 | 0.16800 | 0.10820 | 0.15540 | 0.22680 | |
Trend | No trend | No trend | No trend | No trend | No trend | No trend | ||
S | 0.12500 | 0.04500 | 0.04500 | 0.18340 | 0.04800 | 0.04640 |
Potential evapotranspiration
The variation of projected evapotranspiration (mm) in the study regions for both SSP scenarios
. | . | Scenarios . | |||
---|---|---|---|---|---|
. | . | Western catchment . | Eastern catchment . | ||
Evapotranspiration variations in mm/year . | Time . | RCP4.5 . | RCP8.5 . | RCP4.5 . | RCP8.5 . |
Baseline | 1981–2010 | 1,451.32 | 1,451.32 | 1,202.13 | 1,202.13 |
Short range | 2015–2040 | 1,548.82 | 1,562.72 | 1,257.93 | 1,283.72 |
Mid-range | 2041–2070 | 1,695.27 | 1,806.71 | 1,364.73 | 1,462.11 |
Change short range (%) | 6.72 | 7.68 | 4.64 | 6.79 | |
Change mid-range (%) | 16.81 | 24.49 | 13.53 | 21.63 |
. | . | Scenarios . | |||
---|---|---|---|---|---|
. | . | Western catchment . | Eastern catchment . | ||
Evapotranspiration variations in mm/year . | Time . | RCP4.5 . | RCP8.5 . | RCP4.5 . | RCP8.5 . |
Baseline | 1981–2010 | 1,451.32 | 1,451.32 | 1,202.13 | 1,202.13 |
Short range | 2015–2040 | 1,548.82 | 1,562.72 | 1,257.93 | 1,283.72 |
Mid-range | 2041–2070 | 1,695.27 | 1,806.71 | 1,364.73 | 1,462.11 |
Change short range (%) | 6.72 | 7.68 | 4.64 | 6.79 | |
Change mid-range (%) | 16.81 | 24.49 | 13.53 | 21.63 |
Potential evapotranspiration of (a) western and (b) eastern catchment for the reference period (1981–2010).
Potential evapotranspiration of (a) western and (b) eastern catchment for the reference period (1981–2010).
Evapotranspiration trend of the western and eastern catchment for the historical period.
Evapotranspiration trend of the western and eastern catchment for the historical period.
Aridity indices
Different aridity/humidity conditions with a common number of climate classes were observed in the western and eastern catchments. The western catchment characteristic tilted to humidity, while the eastern catchment tilted to aridity conditions. In other words, western catchments are wetter, and the eastern catchments are frequently drier.
De Martone aridity index
De Martone aridity index with the climate classification of the two catchments for the historical period (1981–2010)
DMI value . | Climate class . | West catchment . | East catchment . | ||
---|---|---|---|---|---|
Area (km2) . | % . | Area (km2) . | % . | ||
DMI < 10 | Arid | – | – | 30,299.86 | 33.54 |
10 < DMI < 20 | Semi-arid | – | – | 42,003.90 | 46.49 |
20 ≤ DMI < 24 | Mediterranean | – | – | 12,953.65 | 14.34 |
24 ≤ DMI < 28 | Semi-humid | 1,022.94 | 2.50 | 4,716.14 | 5.22 |
28 ≤ DMI < 35 | Humid | 9,462.05 | 23.14 | 367.72 | 0.41 |
35 ≤ DMI ≤ 55 | Very humid | 15,237.95 | 37.27 | – | – |
DMI > 55 | Extremely humid | 15,161.18 | 37.08 | – | – |
DMI value . | Climate class . | West catchment . | East catchment . | ||
---|---|---|---|---|---|
Area (km2) . | % . | Area (km2) . | % . | ||
DMI < 10 | Arid | – | – | 30,299.86 | 33.54 |
10 < DMI < 20 | Semi-arid | – | – | 42,003.90 | 46.49 |
20 ≤ DMI < 24 | Mediterranean | – | – | 12,953.65 | 14.34 |
24 ≤ DMI < 28 | Semi-humid | 1,022.94 | 2.50 | 4,716.14 | 5.22 |
28 ≤ DMI < 35 | Humid | 9,462.05 | 23.14 | 367.72 | 0.41 |
35 ≤ DMI ≤ 55 | Very humid | 15,237.95 | 37.27 | – | – |
DMI > 55 | Extremely humid | 15,161.18 | 37.08 | – | – |
Bold values are to show the majority coverage in the area.
The DMI value of the study area for reference periods (1981–2010): (a) western and (b) eastern catchment.
The DMI value of the study area for reference periods (1981–2010): (a) western and (b) eastern catchment.
UNEP aridity index
Based on the UNEP aridity index, the classification of study area's climate for the historical period (1981–2010)
AI value . | Climate class . | West catchment . | East catchment . | ||
---|---|---|---|---|---|
Area (km2) . | % . | Area (km2) . | % . | ||
AI < 0.05 | Hyper-arid | – | – | – | – |
0.05–0.20 | Arid | – | – | 19,429.75 | 21.51 |
0.20–0.50 | Semi-arid | 1,536.28 | 3.76 | 41,651.57 | 46.12 |
0.5–0.75 | Dry sub-humid | 8,603.37 | 21.04 | 24,305.72 | 26.91 |
AI > 0.75 | Humid | 30,746.46 | 75.20 | 4,927.37 | 5.46 |
AI value . | Climate class . | West catchment . | East catchment . | ||
---|---|---|---|---|---|
Area (km2) . | % . | Area (km2) . | % . | ||
AI < 0.05 | Hyper-arid | – | – | – | – |
0.05–0.20 | Arid | – | – | 19,429.75 | 21.51 |
0.20–0.50 | Semi-arid | 1,536.28 | 3.76 | 41,651.57 | 46.12 |
0.5–0.75 | Dry sub-humid | 8,603.37 | 21.04 | 24,305.72 | 26.91 |
AI > 0.75 | Humid | 30,746.46 | 75.20 | 4,927.37 | 5.46 |
Bold values are to show the majority coverage in the area.
The UNEP-AI value of the study area for reference periods (1981–2010): (a) western and (b) eastern catchment.
The UNEP-AI value of the study area for reference periods (1981–2010): (a) western and (b) eastern catchment.
Trends of aridity indices
Historical aridity indices for (a) western catchment and (b) eastern catchment.
Historical aridity indices for (a) western catchment and (b) eastern catchment.
The spatial variability of aridity indices in the western catchment: (a) and (b) De Martone aridity index and (c) and (d) UNEP aridity index.
The spatial variability of aridity indices in the western catchment: (a) and (b) De Martone aridity index and (c) and (d) UNEP aridity index.
The spatial variability of aridity indices in eastern catchment: (a) and (b) De Martone aridity index and (c) and (d) UNEP aridity index.
The spatial variability of aridity indices in eastern catchment: (a) and (b) De Martone aridity index and (c) and (d) UNEP aridity index.
DMI and UNEP-AI values of western catchment for future scenarios: (a) SSP2-RCP4.5 and (b) SSP-RCP8.5.
DMI and UNEP-AI values of western catchment for future scenarios: (a) SSP2-RCP4.5 and (b) SSP-RCP8.5.
DMI and UNEP-AI values of eastern catchment for future scenarios (a) SSP2-RCP4.5 and (b) SSP-RCP8.5.
DMI and UNEP-AI values of eastern catchment for future scenarios (a) SSP2-RCP4.5 and (b) SSP-RCP8.5.
The spatial variability of aridity indices in western catchment for projected climate change DMI (a–d) and UNEP-AI (e–h).
The spatial variability of aridity indices in western catchment for projected climate change DMI (a–d) and UNEP-AI (e–h).
The spatial variability of aridity indices in eastern catchment for projected climate change DMI (a–d) and UNEP-AI (e–h).
The spatial variability of aridity indices in eastern catchment for projected climate change DMI (a–d) and UNEP-AI (e–h).
Comparison with previous studies
A study by Haile and others (2022) examining drought using the standardized ETo and drought index shows that the impact of droughts increases from time to time. Another study (Gebremedhin et al. 2018a) showed an increasing trend in all future temperatures. Examination of past drought balances also revealed that the lion's share of the western catchment is covered by wet climates and the eastern catchment is overwhelmed by semi-arid climates (Abera et al. 2019). Several studies share common sense about the future impact of climate change on Ethiopia's hydroclimatic parameters (Funk et al. 2012; Abrha & Hagos 2019).
Change of climate aridity and water resources
Changes in temperature, precipitation, and other atmospheric parameters are the main climate factors that affect dryness and water loss from plant transpiration and evaporation (Bibi et al. 2018; Li & Quiring 2021). Various researchers examine how climate change affects the availability of water in various Ethiopian regions (Frederick & Major 1997; Watts et al. 2015; Abrha & Hagos 2019; Tabari 2020). The impact of climate change on water resources is from domestic water supply to hydroelectric power generation and from agricultural use to industrial demand. Climate change affects water availability in both the eastern and western parts of Ethiopia (Abera et al. 2019). The precipitation and temperature changes in the study regions are contrary, resulting in opposite changes in the hydrological regimes. Increasing precipitation means increasing runoff, while increasing temperature leads to decreasing runoff. Consistent increasing temperature projections indicate that potential ETo may simultaneously increase and lead to a reduction in streamflow. Study shows that runoff in tropical zones is sensitive to precipitation and temperature change (Hasan et al. 2018). PET is also increasing for both the reference period and for the projected climate case, which plays a significant role in the water balance system, including surface runoff, water resource management, harvesting, and ecological water needs (Tadese et al. 2020). Population progress joined with an increase in potential ETo, and a decrease in water accessibility will make droughts and food insecurity more common. Consistently increasing temperature projections indicate that potential ETo may simultaneously increase and lead to a reduction in surface flow (Taye et al. 2015). The increasing aridity could cause even more water to evaporate into the atmosphere, which would eventually diminish river streamflow and affect how much water is available for crops and livelihood. Investigating how aridity is changing and the role of climate variables will improve our ability to predict long-term runoff, improve the efficiency of water resource management, and provide further insights into drought resilience and disaster mitigation. In addition, changes in aridity significantly affect the water cycle, water resource management, and desertification. Our results are in good agreement with other studies on projected climate change looking at other areas (Funk et al. 2012; Alaminie et al. 2021).
CONCLUSION
This research aims to identify trends in the spatial distribution of climate parameters related to water availability in semi-arid Ethiopia based on modeled temperature and precipitation data from 1981 to 2070. Bias-corrected multiple GCMs (CMCC-ESM2, FGOALS-g3, and MIROC6) of CMIP6 are selected to perform short-range (2011–2040) and mid-range (2041–2070) high-resolution climate projections concerning the reference period of 1981–2010 under SSP2-RCP4.5 and SSP5-RCP8.5. We first evaluated the seven GCM outputs of the simulated daily temperature and precipitation results from CMIP6 and selected the three best models for analysis. The results of this study conclude that CMCC-ESM2, FGOALS-g3, and MIROC6 are in the best agreement with ground observation data (1981–2010) for the study area. The most important climatic parameters (precipitation) have decreased in the last decades for both catchment areas. The temperature rises in the other results of both catchments for the historical record. In general, the decrease in precipitation and the increase in temperature are of high intensity in the western catchment compared to the eastern catchment. The AIs followed a similar downward trend with higher intensity in the western catchment than in the eastern one. The potential ETo in the reference period increased alarmingly by 14 mm/year in the western catchment. This is a result of recent decades' rising temperature intensity. The findings indicate that both study locations' climates are expected to become noticeably warmer than they are now within the next few decades. The temperature increment is higher for the mid-range (2041–2070) than short-range (2011–2040) for both SSPs. The results show that UNEP-AI is better than DMI in representing the coming climate conditions for both catchments. The overall spatial pattern of drought remains stable between 1981 and 2070, but important quantitative shifts toward more drought will occur in the western catchment and in the northeastern and southern parts of the eastern catchment. Finally, the researchers hope that the results of the present study will prove helpful in future planning, assessment, and management of water resources to plan and act accordingly to the coming impacts of climate change and to focus on climate-vulnerable areas through necessary precautions. We believe that researchers go through the climate change impact on water resources of the study regions and contribute to a better understanding of the spatial and temporal variability of drought on water resources.
ACKNOWLEDGEMENTS
The author expresses gratitude to all administrative bodies for providing the information required for this research project. I want to give deep thanks to the anonymous reviewer for their supportive and constructive reviews, which significantly improved the quality of the paper.
STATEMENTS AND DECLARATIONS
To the utmost of my understanding, I verify that the information encompassed herein is precise and all-inclusive. I hereby state that this task has not been formerly issued and has not been proposed for publishing in any journal.
AUTHOR CONTRIBUTIONS
Both authors contributed to the study's conception and design. Material preparation, data collection, and analysis were performed by Mr Tesema Kebede and Mrs Kidist Demessie. The first draft of the manuscript was written by Mr Tesema Kebede. Mrs Kidist Demessie read and approved the final manuscript.
FUNDING
The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
CONFLICT OF INTEREST
The authors declare there is no conflict.