Quantifying the future impact of the climate and land use change on hydrologic variables of the watershed is a thrust area of research in the field of hydrology to understand the availability and distribution of water for different uses in near future (Kumar et al., 2022). Further, to study the sustainability of existing structures and for the development of innovative water resource management plans for a specific basin or watershed can also achieved through such modelling studies. The spatiotemporal change of any landscape is a global concern given its impacts on human and animal lives and their complex relationships. Land use/land cover change (LULCC) and climate change are two important factors which affects the dynamics of hydrology of any local, regional as well as global level and ultimately affecting the humans and environment. The knowledge of historical and projected LULC and climate change are the fundamental concerns to the sustainable development of any watershed management. It is expected that due to modification of catchment LULC and climate change have been altered the hydrological condition of the area (Huyen 2017; Boru et al., 2019). The change in climatic variability can also alter the flow routing time, peak flows and overall volume of the watershed (Changnon and Demissie, 1996; Prowse et al., 2006). However, LULC change because of deforestation, urbanization, and cultivation with different tillage practices resulted into alteration of the surface runoff and ultimately causes change of flood frequency (Crooks and Davies, 2001; Binh and Trung 2005; Brath et al., 2006;), severity (De Roo et al., 2001), base flow (Tang et al., 2006), and annual mean discharge (Costa et al., 2003) of any watershed.
The computer simulation models can capture the interactions in time and space between the environment and humans for improved understanding of land use change (Veldkamp and Verburg, 2004). Modelling of land use dynamics using land use change models greatly improved our understanding about the cause and response relationship. Different biophysical drivers and the role of land use policies can be modelled using simulation models to explore, evaluate and visualize alternative futures for better development (Choudhari and Clarke, 2013). The cellular automata (CA) models divide a landscape into smaller cells and the behaviour is determined by transitions rules to capture the uncertainties of real-world system and the pattern and process (Batty, 2005). The CA-Markov models are popular due to simplicity, comprehensiveness, theoretical foundation, modelling techniques, overall simple structure, dynamics, data requirements, calibration and validation, operationality, and, applicability for prediction of future pattern and trends (Wegner, 1995; Torrens and Sullivan, 2001). The CA models simulate spatial distributions whereas Markov Chain models simulate temporal changes (Singh et al., 2015; Tariq & Shu, 2020). Singh et al. (2018) used the integrated CA-MCM model to predict spatial and temporal land use pattern using earth observation data-sets of Tons River Basin, Madhya Pradesh, India and concluded that sub-basin changes are largely influenced by the socio-economic and biophysical drivers. Munthali et al. (2020) successfully applied CA-MCM model for a district of Malawi and reported that change in landscape patterns is a complex interaction of factors namely biophysical, demographic, socio-economic, technological and cultural and reported that forest land will exhibit the highest pressure of humans. Singh et al. (2022) applied artificial neural network (ANN) embedded with Land Change Modeler (LCM) in TerrSet to predict future LULC of year 2030 and 2050 and they reported that open forest and the built-up land area will increase and this will have impact on the surrounding environment. The CA-MCM was applied to model urban growth and to measure urban footprints of a city (Kushwaha et al., 2021). Hence, MCM model is independent of time step and the predicted land use map of future can be used in future modelling studies.
Literature review outlined that there are many hydrological models available for simulating the hydrological behavior of a watershed or basin of varying degrees of complexity and used for either isolated impact investigation (LULCC or Climate Change separately) or combined (LULCC and Climate Change) (Sharma et al., 2022; Yin et al., 2017). The Variable Infiltration Capacity (VIC) model is a semi-distributed macro-scale hydrological model, it balances both the water and surface energy (Gao et al., 2010) and applied for India (Chawla and Mujumdar, 2015; Garg et al., 2019). The MIKE-SHE is an integrated model using for analyzing evapotranspiration, recharge, surface water and groundwater. The VIC model applied for water balance computation and water quality improvement for Yanghe Basin, China (Lie et al., 2022) and analyzing blue and green water quantities for watershed of China (Li et al., 2022). The SPHY model is a raster-based water balance model, researchers applied for investigation of climate change response for high mountain rivers (Khanal et al., 2021). The Water Erosion Prediction Project (WEPP) is a process-based model used to predict water balance and sediment yield was employed for an ungauged watershed in Shivalik foot-hills (Yousuf et al., 2022). Groundwater Loading Effects of Agricultural Management Systems (GLEAMS) is a continuous and field-scale hydrological and can simulate surface runoff, percolation, leaching, erosion and sedimentation and applied for simulation of pesticides for GKb catchment, Flanders region of Belgium (Rathjens et al., 2022). The SWAT is a semi-distributed and physically-based model which has been applied around the for-stream flow modeling, water balance estimation, tillage operation, water allocation studies due to its simplicity, simple model structure (Sharma et al., 2022) and ease of operation and perform uncertainty and sensitivity analysis through different methods for improving the robustness of a forecasting model (Xie and Lian, 2013).
Few scholars have investigated the isolated impact of future LULC on water balance component and suggested that catchment scale modifications in LULC affects different water balance components of a basin (Kumar et al., 2018; Sertel et al., 2019; Kumar et al., 2022). Others have explored the isolated impact of climate on hydrologic components (Gashaw et al., 2018; Gong et al., 2019; Zhang et al., 2020). The CA-Markov model has been used to predict LULC change along with evolution and simulation of future land use with quantitative method (Qui, 2017; Kurniawan Ramadhan and Supriatna, 2019) whereas, SWAT model has been used to predict water balance of the watershed using LULC, soil and slope. Guajar et al., (2022) investigated the impact of LULC using SWAT model on surface flow at three gauging sites without considering the dams, barrages and irrigation water uses and reported that the average annual flows were observed 40% higher for the Gomti River Basin, India. The climate change impact is more pronounced on SWAT sensitivity parameters compared to the LULC change for Dharoi catchment, India (Sharma et al., 2022). SWAT in combination with copula statistics was applied to detect drought onset for Gomti River basin, India (Bhatt et al., 2022). Many previous studies have been conducted using combination of SWAT model and CA-Markov model to predict future water balance of the area (Gong et al., 2019; Ji et al., 2021; Ghodichore et al., 2022). Kumar et al., (2018) investigated the future LULC change impact on the water resources of the Tons Basin, India and observed that there is a decrease in surface runoff and lateral flow and slight increase in water yield for the simulated LULC year 2035. Kumar et al. (2022) applied SWAT and CMIP6 model to investigate the environmental flow and highlighted the extreme low flows will be dominant in the Ghaghara river basin, India. Climate change impact on runoff for Qinhuai River was assessed using SWAT model and CMIP6 scenarios and finding suggest that there will be increases in both rainfall and runoff hence there is an increased risk of flooding (Sun et al., 2022). The aim of this study was to simulate the future runoff change under the combined effect of future LULC and climate change.