Investigating the impact of climate change on irrigation and crop water requirements of Bhadra and Tungabhadra command area: A CMIP-6 GCMs and CROPWAT 8.0 approach

The severity and frequency of hydrological extremes have been altered by global climate change, resulting in increased stress on agricultural and socioeconomic development (Yuan et al. 2015; Piras et al. 2016; Malla & Arya 2023; Velpuri et al. 2023). The hydrological cycle, which is significantly influenced by climate change, plays a vital role in determining the availability of water resources and agricultural water demand (Tukimat et al. 2017; Vogeti et al. 2023). Climate change thus has a significant impact on these factors. Water is treated as an important commodity for the development of human society in semi-arid regions (Umesh Babu & Puttaiah 2013). India is an agrarian country, and the majority of the reservoir projects primarily focus on irrigation. The agricultural system mainly faces pressure due to climate variability, population growth, urbanization, and industrialization at the regional level (Dai et al. 2013; Yu et al. 2019; Ahmad et al. 2023). Climate parameters are the major factors that affect water requirements, crop growth, development, and crop production in a region (Meenu et al. 2013; Battude et al. 2016; Timlin et al. 2024). Due to anthropogenic activities and population growth, the availability of water resources and water demand have been changing dramatically in the future timestep. Consequently, it is critical to comprehend the variation in the irrigation water requirement (IWR) under climate change conditions for optimal operation of reservoirs in the long run.

Climate projections for future periods are generally obtained from the global climate models (GCMs), these models reproduce the physical process of the atmosphere through various mathematical equations. Intergovernmental Panel on Climate Change (IPCC) has initiated the Coupled Model Intercomparison Project Phase-6 (CMIP-6) that consists of various climate models from researchers throughout the world (Meinshausen et al. 2020). The uncertainty in the future water availability is obtained with the aid of CMIP-6 GCMs outputs under various shared socioeconomic pathways (SSPs) (Riahi et al. 2017; Wang et al. 2020; Katzenberger et al. 2021; Li et al. 2021). Uncertainty present in the CMIP-5 climate models has been lowered in the recent CMIP-6-based GCMs with a better understanding of the physics underlying the atmospheric process over a couple of decades (Eyring et al. 2016; Gusain et al. 2020; Priestley et al. 2020). To understand regional water availability various researchers have used the CMIP-6-based GCMs in recent years (Katzenberger et al. 2021; Karan et al. 2022; El-Rawy et al. 2023).

As per Jain & Singh (2020), there exists a strong correlation between global warming and global crop water requirement (CWR), and an increase in rainfall has reduced its benefit to irrigation. A total of 56% of the agricultural land in India is rainfed (Goyal & Surampalli 2018). The regional impact of climate variables on CWR is quintessential for sustainable management and development of the agriculture sector. The uncertainty associated with climate change can cause occurrences of droughts and floods in a highly variable manner affecting water availability and agriculture significantly (Das et al. 2020; Sunil et al. 2021). The principle that lies behind any irrigation project is an estimation of CWR (Dhamge et al. 2008). Appreciating the impact of climate change on CWR is most important in solving problems related to food security and optimal utilization of existing resources for irrigation in the future period (Zhou et al. 2017; Roushdi 2024).

Climate projections from GCMs, considering various climate scenarios, have been utilized in several studies to determine the future irrigation demand of a command area (Sunil et al. 2021; Sharma & Tare 2022). Different research works have focused on estimating reference evapotranspiration (ET_o) and CWR under changing climate parameters for various climate scenarios (Shen et al. 2013; Bouras et al. 2019; Anil 2020; Agrawal et al. 2023). CROPWAT 8.0, developed by the Water and Land Development Department of the Food and Agriculture Organization (FAO), is a widely used tool for estimating crop evapotranspiration (ET_c), CWR, IWR, and irrigation scheduling globally (Poonia et al. 2021; Sunil et al. 2021; Gabr 2022). In CROPWAT 8.0, ET_o is estimated using the FAO Penman–Monteith method (Allen et al. 1998, 2004), and the United States Department of Agriculture (USDA) soil conservation (SC) method is employed for effective rainfall estimation (Mehta & Pandey 2016; Djaman et al. 2018; Patidar et al. 2020).

Moreover, the estimation of command area level CWR assists the policymakers to lead better strategies under climate change impacts. In the present study, uncertainty in the future CWR and IWR of Bhadra and Tungabhadra (TB) command areas are studied by considering the output from CMIP-6-based climate models under two emission scenarios. Rehana & Mujumdar (2013) estimated the irrigation demand under future climate using a single climate model results from the study to show that CWR in future climate is going to increase and also suggested that the basin-level irrigation demand assessment would benefit the policymaker to decide on changing climate. A recent study suggested that there exists a discrepancy between the historical demand and the supply for the existing cropping system in the Bhadra command area (Kumar et al. 2022). From the CROPWAT model, estimated crop evapotranspiration of paddy crops in the Bhadra command area shows that tail-end water requirement was more for the period 1984–2010 (Abdul Karim et al. 2013). System productivity and profitability of the TB basin studied earlier will be greatly affected by the CWR of the region in the future period. Kumar et al. (2022) suggested that the existing cropping pattern needs to be modified to cater to the mismatch in the water demand and supply in the Harihara branch canal under the Bhadra reservoir. Therefore, there is a need for an effective plan and adaptation of integrated water resource management (IWRM) to attain sustainable development (Basavanneppa & Kumar 2020).

The climate projection of the TB basin for the future has been predicted to observe several droughts along with greater variability of streamflow, soil moisture, and evapotranspiration (Gosain et al. 2006; ACIWRM & WRD 2012; Chanapathi et al. 2020; Nyayapathi et al. 2023). Umesh Babu & Puttaiah (2013) show the possible concerns and limitations that exist in the water resource management of the TB basin and suggest a holistic IWRM to gratify the climate change impact. The literature mentioned earlier shows how important it is to analyze the CWR of the Bhadra and TB command areas under changing climate scenarios. The present study builds upon this foundation by examining the uncertainty in future CWR and IWR for the Bhadra and TB command areas using output from CMIP-6-based climate models under two emission scenarios. Notably, the dearth of evidence in the existing literature for comprehensive studies estimating future irrigation demands at the command area level, specifically using CMIP-6-based best-performing GCMs under different SSPs. This emphasizes the novelty of the current research (Surendran et al. 2015; Sunil et al. 2021; Gabr 2022; Sharma & Tare 2022).

The aim is to bridge the research gap by analyzing the effect of climate change on future irrigation demand in the Bhadra and TB command areas under SSP-245 and SSP-585 scenarios, utilizing the outputs from the five best-performing GCMs. The anticipated variation in future CWR and IWR holds substantial implications for policymakers, providing valuable insights for future irrigation planning and reservoir operation modeling within the command area. By addressing this gap in the literature, our study contributes to the broader understanding of climate change impacts on ET_o, CWR, and IWR at the command area level.

The Bhadra and TB command areas are situated within the TB basin, which is a subbasin of the Krishna Basin. TB subbasin covers an area of 28,177 km² and is located between 74°46′52″ E to 78°01′29″ E longitude and 13°8′60″ N to 16°13′35″ N latitude in the state of Karnataka. Tungabhadra River is formed by the convergence of the Bhadra and Tunga rivers near Koodli village in the Shivamogga district. River Bhadra and Tunga start from Gangamoola in the Varahaparvatha hills of the Western Ghats.

The Indian Meteorological Department (IMD) provided the observed precipitation, maximum temperature (T_max), and minimum temperature (T_min). The daily precipitation data at a grid resolution of 0.25° × 0.25° are available (Pai et al. 2014). The daily T_max and T_min data have a spatial resolution of 1° × 1° (Srivastava et al. 2009). The data collection period spans from 1975 to 2015. The wind speed data and relative humidity data were obtained from the National Centre for Environmental Prediction (NCEP)/National Center for Atmospheric Research (NCAR) re-analysis data (https://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.pressure.html).

GCMs are the most suitable tool for studying regional climate change and its future impacts. These models project different hydro-meteorological variables for future periods, taking into account various climatic scenarios. Recent studies utilize the latest set of GCMs provided by the CMIP-6 to generate future hydro-meteorological datasets.

Within the CMIP-6-based GCMs, a new set of climate scenarios called SSPs were considered. These scenarios take into account various socioeconomic components (Riahi et al. 2017). Despite that, the GCM projections obtained directly are not suitable for capturing regional climate due to their coarse spatial resolutions (>200 km²). Therefore, it becomes essential to correct the inherent biases present in the raw dataset. In this study, a bias-corrected dataset is employed to estimate future CWR. Mishra et al. (2020) addressed the issue of bias correction for climate data by developing a bias-corrected climate dataset for six countries in South Asia using the empirical quantile mapping (EQM) method. This approach provides a solution for correcting the biases in the climate data, enhancing their suitability for regional analysis.

In the present study, the climate projections from CMIP-6-based bias-corrected 13 GCMs for SSP-245 and SSP-585 scenarios for the period 2023–2100 at a spatial resolution of 0.25° were used. It is important to note that the historical dataset for the period 1975–2015 from all the models was extracted along with future data for the study area. The selected GCMs along with their originating institute and spatial resolution are given in Supplementary Material Table S1.

To ensure an accurate estimation of future CWR, it is essential to consider the region-specific performance of GCMs. Hence, it is required to select a subset of GCMs that exhibit consistent performance in representing the observed data for base period simulation. In the present study, a multi-criteria decision-making method is employed to rank GCM's performance. Three performance evaluation metrics, namely, the Nash–Sutcliffe coefficient of efficiency, skill score, and percent bias, are utilized to assess the degree of agreement between observed and simulated data. These metrics provide useful details about the match between observed and simulated values. To facilitate the ranking process, a payoff matrix is constructed, with each GCM treated as an alternative and the aforementioned metrics as the criteria. Since the computed metrics have different ranges, a technique called linear sum normalization is applied to normalize them. Finally, a normalized payoff matrix is derived using Equation (1).

Each grid within the study area assigns rankings to the GCMs, and the ranking schemes may vary across grids. To enable impact assessment studies for the entire study area, a group decision-making approach is employed to determine the suitable set of GCMs. This approach aims to extract a consensus among the grids and identify the GCMs that are appropriate for the entire study area. In this method, the total ranks obtained are divided into two parts (i.e., upper and lower) by arranging the ranks in a decreasing order.

Let m represent the position of a GCM in the lower portion, where m = z for grid points i and m = 0 otherwise. z signifies the position in the lower portion, which ranges from the first position to the xth position, encompassing all rankings in the lower portion. The variable i denotes a grid point, with i ranging from the first to the last grid point within the system.

Finally, GCMs are ranked using values for the GCM with the highest value is considered the best suitable GCM. The group decision-making is employed at each grid to extract five best-performing GCMs for further analysis.

The total irrigation requirements of the crops grown in the study area under future periods were estimated using the CWR obtained. The net irrigation requirement is estimated as the difference between the CWR and P_eff. Finally, the total irrigation water demand for major crops is obtained.

Mean annual precipitation increases under both scenarios, but the median precipitation shows the existence of less increment in the future precipitation projection. A similar kind of pattern was observed in the Bhadra command area, and this may be the result of more frequent short-duration high-intensity rainfall in the study area. T_min shows a higher increment than T_max under both emission scenarios, and the SSP-585 scenario shows a greater increment. The increase in future temperature may increase the rate of evapotranspiration in the study area, which might affect the vegetation. From the results, it was seen that future temperature and precipitation projections are increasing under both Bhadra and TB command areas under future climate scenarios with the SSP-585 scenario showing greater variation.

The CWR for various crops shows both positive and negative trends in the future period, but the variation is maximum in the case of the SSP-585 scenario. For example, the CWR of kharif rice shows an increasing trend in the future under the SSP-245 scenario, and the CWR increases to 4.39, 5.27, and 6.38% in the near, middle, and far future, respectively. In the case of the SSP-585 scenario, the CWR increases by about 5.55, 6.93, and 6.82% in the near, middle, and far future, respectively. Rabi rice shows an average decrement of CWR of about 3.82% in the SSP-245 scenario and 5% in the SSP-585 scenario. A significant increment in future CWR is seen in the case of vegetables up to 4.9% in the SSP-245 scenario and 7.07% in the SSP-585 scenario. Perennial crop sugarcane in the Bhadra command area is showing an average decrement of 4.95% CWR in the future period. The remaining crops show negligible variation in CWR for the future period under both SSP-245 and SSP-585 scenarios. The simulations of crop water demand for future scenarios are obtained using climatic parameters and crop coefficient K_c, the value varies for each crop, and further variation also exists within the growth stages of the crop. CWR obtained from the present study shows the variation present in the individual crop water requirements under future periods, and it is the result of climatic effects as well as the effect of crop development stages.

The CWR for the TB command area during the kharif season reveals a decreasing trend in all the crops for SSP-245 and SSP-585 scenarios. From Figure 6, it is found that a significant decrement in future CWR is observed in the kharif season, and among the scenarios, the SSP-585 scenario shows a greater reduction in future CWR for the future periods. For rice, there is an average reduction of 22.74% future CWR in the SSP-245 scenario and the average decrease of 23.10% CWR in the SSP-585 scenario, and variation among future periods is negligible. Cotton, maize, and millet show an average decrease in future CWR of 18, 23.5, and 21%, respectively. Sunflower shows a greater reduction in CWR up to 24.13% in the SSP-585 scenario. For the annual crop sugarcane, the variation in future CWR to historical value is much less (i.e., up to 1%) in the TB command area.

Crops growing under the TB command area show less variation in CWR in the rabi season than that in the kharif season. From Figure 7, it is observed that there is a decrease in CWR for all the crops, and the decrement is maximum in the case of the SSP-585 scenario. The overall decrease in CWR during the rabi season ranges between 0.5 and 3.35%. The results in Section 4.1 indicate that there is an increase in precipitation, T_max, and T_min in the future period, which might affect the estimation of future CWR, as climate parameters are the only parameters that affect the estimation of ET_o and the increase in T_max and T_min projections leading to an increase in ET_o. Furthermore, ET_o is used to estimate ET_c along with crop coefficient (K_c), which is a constant factor obtained from the historical period simulation, and the trend obtained in ET_o will continue for further calculation.

It is necessary to know the variation in CWR of any command area for the development of optimal cropping patterns for sustainable development. Thus, policymakers can make reliable decisions to decide the operation policy of the reservoir under the command area. From the results, it is found that the existing cropping pattern may not be suitable and sustainable for the future period, and there is a need to optimize the cropping pattern to increase irrigation efficiency. The uncertainty in the climate projections will affect future crop water demands and irrigation demands. It is mandatory to estimate CWR for the command area in the future period to know the effect of climate change on individual crops. It is also necessary to update CWR for the future period with the view of developing an optimal release policy to increase reliability and crop yield.

The results depicted that the future precipitation projections and IWR show an increasing trend in the study area. The soil moisture that exists in the command area before monsoon will not be sufficient to meet CWR during the early growth stages of kharif crops. Since soil water requirements during the initial and development stages are high, the increase in temperature projections increases evapotranspiration resulting in soil moisture content.

The monthly IWR for the TB command area is shown in Figure 9. It shows that the overall monthly future irrigation requirement is increasing in both SSP-245 and SSP-585 scenarios except in the case of January during the SSP-245 scenario. The highest increment in monthly IWR was observed in a hotter month, May, and in the case of June the increase in temperature and a decrease in water availability cause an increase in IWR to a greater extent. During monsoon season though the precipitation projections were showing increasing trends in a future period, and increasing temperature projections caused increasing evapotranspiration, thereby an increase in CWR, the variation is variable among future periods, and this is more in the case of the SSP-585 scenario.

Near-future IWR projections are higher than middle and far-future projections, and this is due to inherent uncertainty prevailing in the future climate. In the present study, five best-performing GCMs are selected to estimate future CWR to obtain reliable IWR under future climate. As GCM output uncertainty was considered a main contributor in climate change studies, the present study guides the policymakers to develop an appropriate release policy for the existing reservoirs in Bhadra and TB command areas. The variation in demand patterns led to the use of multi-objective optimization to augment the release of the Bhadra and TB reservoirs. Since both reservoirs are designed to meet multiple purposes, the dynamic behavior of the irrigation water demand from the study will help to optimize the operation policy along with the enhancement of hydroelectric production in the study area.

Bhadra and TB command areas, situated under the TB sub-basin, are considered vital sources of irrigation, hydropower, industrial needs, and drinking water supply to the surrounding districts. There exists a mismatch between demand and supply at the command area level (Kumar et al. 2022). Hence, estimating the crop water demands and IWR under climate change scenarios is quintessential for formulating the future release policy of the reservoirs within the TB basin. Limited research has been reported so far analyzing the climate change impact on crop water demands and IWR of the Bhadra and TB command areas (Abdul Karim et al. 2013; Rehana & Mujumdar 2013; Kumar et al. 2022). Furthermore, there has been no evidence thus far of considering CMIP-6-based GCMs in estimating CWR and IWR under different climate scenarios (i.e., SSPs). Therefore, in the current study, CROPWAT 8.0 software is used to estimate CWR and IWR under future periods using five best-performing GCMs under the CMIP-6 dataset. GCMs are selected based on CP and group decision-making techniques, which is widely accepted in the process of GCM selection in climate change studies (Srinivasa Raju et al. 2017). The future climate projections under both command areas were found to be increasing under both SSP-245 and SSP-585 scenarios (Rudraswamy et al. 2023). The predicted mean annual precipitation is expected to vary from 300 to 2,850 mm in the future period (2023–2100). Both T_min and T_max indicate an increasing trend, with T_min showing a significant increment (i.e., 0.62–1.62 °C) compared to T_max. The FAO Penman–Monteith method is applied in the CROPWAT model to estimate reference evapotranspiration, a widely accepted method for various crops (Allen et al. 1998; Poonia et al. 2021; Sunil et al. 2021; Gabr 2022).

The temperature rise will significantly contribute to the increase in ET_o losses (Shahid 2014), ET_o losses during the initial growth phase and development stages of crops, due to greater water demand for various physiological processes (Gangwar et al. 2019). From the results, ET_c for the Bhadra command area shows dynamic variation for rice crops in the kharif and rabi seasons. During the kharif season, ET_c for rice shows an increasing trend in the near, middle, and far future under both SSP-245 and SSP-585 scenarios. However, during the rabi season, ET_c is decreasing under both SSPs, and an increase in temperature projections will affect the ET_c during the kharif season though the rainfall shows an increasing trend in that period. CWR shows a mixed trend in the study area. Rice shows an average increment of 5.35 and 6.43% in both SSP-245 and SSP-585 scenarios in the kharif season whereas in the rabi season, the average decrement of 3.82 and 5% is observed for SSP-245 and SSP-585 scenarios, respectively. A significant rise in CWR is observed for the vegetable crop (i.e., 4.9–7.07%) and for the annual crop sugarcane, and the CWR decreases significantly in the future period for both SSPs. Monthly IWR obtained under the Bhadra command area shows a maximum deviation from historical values. In the monsoon period except for June, IWR shows a significant decrement of 84.12, 93.69, and 63.61% in July, August, and September, respectively, for SSP-245 scenario; 4 and 84.46% for SSP-245; and 59.28% during July, August, and September for SSP-585. Greater reduction in IWR during monsoon season is due to higher precipitation projections. However, summer IWR is increasing, indicating the rise in temperature projection increases evapotranspiration in the command area, and irrigation requirements are higher in April and May.

Results obtained for the TB command area show a decrease in CWR for both rabi and kharif seasons, and the SSP-585 scenario shows a greater decrement than the SSP-245 scenario. Kharif crops such as rice, cotton, maize, and millet show an average decrement of 23, 18, 23.5, and 24.13%, respectively. Rabi season shows a negligible decrement than kharif season (i.e., 0.5–3.35%) due to greater variability in climate projections. Contrastingly, the IWR shows an increasing trend in all the months for the TB command area, which indicates that future irrigation releases should be optimized using a better operation policy to meet the rising irrigation requirements of the command area. The average monthly irrigation requirements of SSP-245 were higher than SSP-585 due to the higher precipitation projections obtained under SSP-585 future periods.

The findings recommend a notable impact of climate change on CWR and IWR, which is in line with previous studies (Abdul Karim et al. 2013; Kumar et al. 2022). In addition, the uncertainty in climate projections and the ET_o estimation method plays a vital role in estimating future irrigation requirements (Multsch et al. 2015). The study specifically focuses on the command area, emphasizing the relevance of the aforementioned considerations, but the cropping pattern and soil profile are assumed to be constant for future periods, which can be addressed in the future study. Overall, the results obtained give valuable insights for the command area development authority, facilitating sustainable development by optimizing cropping patterns and reservoir operation policies to enhance irrigation efficiency in response to changing climatic conditions.

The present study explores the impact of climate change on CWR and IWR in Bhadra and TB command areas for CMIP-6 GCMs under SSP-245 and SSP-585 scenarios. The conclusions of this study are as follows:

• The future climate projections from the best-performing CMIP-6 GCMs show a consistent increasing trend from 2023 to 2100 in both command areas under SSP-245 and SSP-585 scenarios.

• Analysis of future CWR of the Bhadra command area reveals a rise in kharif rice CWR by 4.39, 5.27, and 6.38% in near, middle, and far future under SSP-245 and by 5.55, 6.93, and 6.82% under SSP-585, respectively. Contrastingly, the rabi rice crop reveals a decrease; vegetables show significant increments of 4.9 and 7.07% in SSP-245 and 585 scenarios, respectively; and perennial crop sugarcane shows an average decrement of 4.95% CWR in the future period.

• In the TB command area, both the kharif and rabi seasons exhibit a decreasing trend in CWR for SSP-245 and SSP-585 scenarios, with a significant reduction in kharif rice CWR by 22.74 and 23.10% for SSP-245 and SSP-585 scenarios, respectively. Kharif crops show a more pronounced decrease compared to rabi crops (i.e., 0.5–3.35%). Higher uncertainty in the SSP-585 scenario resulted in more variability in the estimation of CWR within the command area.

• Future monthly IWR for the Bhadra command area indicates a substantial reduction for July, August, and September, in line with the increase in monsoon precipitation, which could satisfy CWR effectively in the command area. Meanwhile, the rise in temperature projections in the future period increases IWR during April, May, and June.

• Monthly IWR under the TB command area exhibits a significant increase for all the months under future climate scenarios, despite the decrease in CWR. This emphasizes care to be taken in deciding cropping patterns and implementing effective management strategies to enhance irrigation efficiency in the command area.

• The study addresses the uncertainty in future CWR and IWR stemming from climate variables. It employs the best-performing GCMs and FAO Penman–Monteith method, a widely accepted method, leading to increased irrigation efficiency and optimizing the cropping pattern and reservoir releases for future periods.

The outcome of the study would provide a foundation for adaptation strategies, including enhancing irrigation efficiency, managing water demand, installing water harvesting schemes, optimizing the future cropping pattern, and irrigation releases from the reservoir based on future WA for various climate scenarios in the TB basin.

All relevant data are available from an online repository or repositories: https://www.imdpune.gov.in/cmpg/Griddata/Rainfall_1_NetCDF.html – For gridded precipitation data, https://www.imdpune.gov.in/cmpg/Griddata/Max_1_Bin.html – For gridded temperature data, https://esgf-node.llnl.gov/projects/cmip6 – future climate dataset. https://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.pressure.html – For wind speed data and relative humidity data.

The authors declare there is no conflict.