2.8.1. Climate Scenario Analyses Setting and Simulation

An Arc-SWAT-based modeling approach to analyzing the impacts of climate change in the sub-basins of the CRV lakes region, and optimum agricultural water use and optimization strategies with respect to the identified impacts were carried out. Separate modeling for the selected sub-basins was performed. The climate scenarios (CSc) were set to analyze the impacts of climate change on the components of the water balance in the near-term (2031–2060) and in the long-term (2070–2099) periods for each of the regional concentration pathway (RCP) emission scenarios. The emission scenarios are RCP2.6 (low emission scenario), RCP4.5 (medium emission scenario), and RCP8.5 (high emission scenario). The simulations were categorized into seven CSc analyses, including the baseline data as listed in Table 5. The options for agricultural water use management are indicated based on the resulting water balance components affected by the changes in climate for each sub-basin.

The climate data were downscaled, bias corrected, analyzed, and simulated in an integrated manner with WGEN, CMhyd, and Arc SWAT. The WGEN software interlinks station coordinates and elevations with their respective data. All data statistics, such as average, standard deviation, mean, variance, etc., for each of the weather components downscaled were calculated and synchronized to their respective stations with WGEN. Rain Years, dew point, and other important variables useful for calculating the water balance components were also calculated and generated in WGEN. Finally, these climate data were imported into the SWAT models and simulated to see the changes in the components of the water balance that are especially useful for surface water sources.

#### 2.8.2. Data Downscaling

Climate data stored in the World Climate Research Program (WCRP) databases were used. The data are from the experiments of CMIP5–RCP (RCP2.6-CMIP5, RCP4.5-CMIP5, and RCP8.5-CMIP5). These data were derived by the MIROC-RCA4 ensemble driving climate models under the GCM. The GCM data of these RCP data variables were regionalized to the regional climate model (RCM) with the Coordinated Regional Downscaling Experiment (CORDEX) for Africa, CORDEX-AFR-44. Both, historical data as well as the data of RCP2.6, RCP4.5, and RCP8.5 were downscaled by RCA4 models. RCA4 is the fourth version of the Rossby Center Regional Atmospheric model. It was originally developed by the Swedish Meteorological and Hydrological Institute within the CORDEX initiative. It is a dynamic downscaling method widely used with the CORDEX [23,53]. The downscaled datasets were daily precipitation, daily maximum near-surface air temperature, daily minimum near-surface air temperature, daily sunshine duration, near-surface relative humidity, and near-surface wind speed for future periods from 2006 to 2100. The duration of daily sunshine in units of seconds (s) was extracted from the model and adjusted to daily solar radiation with the units of kilowatt per square meter (KW/M2) for SWAT use and to the SWAT input data standard units using Angstrom techniques [54].

#### 2.8.3. Bias Correction

The data for precipitation and temperature were bias-corrected via linear scaling methods with CMhyd software, which is a SWAT community tool, before they were applied in the SWAT simulation. The need for bias correction is mainly due to onshore and offshore trade wind disturbances. The historical data from the model and the observed locational dataset from six stations in the study region were applied to the software. Data ranges from 1990 to 2006 were applied from the historical dataset of the climate model. Furthermore, observed datasets from the same periods were used to correct the biasedness created due to trade winds in the climate models. Parameters or correction factors for each month were developed in relation to the observed data range of the same time periods. Based on the parameters, the software adjusted the predicted rainfall and temperature values from the downscaled data. The corrected data values were applied to WGEN for statistical analyses and then to SWAT for simulation.

**Table 5.** Applied climate scenarios for analyzing the impacts of climate change on the major components of the water balance in the sub-basins.


Note: NT = Near term and LT-Long term.

#### **3. Results and Discussion**
