Assessing climate change impacts on irrigation water requirements in the Lower Mahanadi Basin: A CMIP6-based spatiotemporal analysis and future projections

Climate changes directly impact socio-economic activities, affecting agriculture, water, energy, and food security, limiting growth (Hochman et al. 2018). Water scarcity constrains Indian agriculture, and climate change will impact agricultural water use. Studying climate effects on irrigation water requirement (IWR) is vital for addressing both food security and sustainable water use (Makar et al. 2022). India holds the leading position in paddy production, with around 84% of cultivation occurring during the Kharif season. This season necessitates rainfall exceeding 1,000 mm and an ideal temperature of 25 °C (Vyankatrao 2017). The total water requirement is influenced by key factors such as rainfall quantity, distribution, temperature fluctuations, crop variety, and soil composition. Additionally, alterations in the duration and intensity of rainfall, fluctuations in diurnal temperatures, and shifts in weather patterns significantly impact the production of Kharif crops (Mall et al. 2007). It is important to highlight climate change assessments for a better grasp, especially because future predictions for regional monsoons are uncertain. Thorough investigations are necessary due to the significant impact of global warming on the sensitivity of monsoons (Moon & Ha 2020).

Erratic rainfall, water scarcity, extreme weather, and saline intrusion pose significant challenges to paddy cultivation, crucial for the nation's food supply. Paddy irrigation in India is threatened by climate change, impacting agriculture and food security. This underscores the urgent need for proactive measures to safeguard paddy irrigation. To address these challenges, studies have explored various mitigation and adaptation strategies aimed at enhancing the resilience of paddy cultivation to climate change impacts (Suryavanshi et al. 2012; Shanabhoga et al. 2020; Malhi et al. 2021). Water-saving methods like sprinklers and drip systems aid agriculture's adaptation to climate change. Crop diversification with saline-tolerant varieties boosts paddy farming resilience. Rising temperatures increase evapotranspiration rates, requiring more water for paddy fields. Erratic rainfall and longer dry spells make water availability unpredictable, harming rice cultivation. These factors intensify pressure on traditional irrigation methods, reducing crop yields, and worsening water scarcity (Kumar et al. 2012). A report (Nelson et al. 2009) indicated that climate change will affect both the direct and indirect aspects of IWR and crop yields, respectively. Climate change's adverse effects on agriculture have made global food security a paramount challenge in the 21st century. McGuire (2015) studied climate change effects on agriculture and food security using different scenarios. Results show similar impacts for moderate emissions of representative concentration pathways (RCP4.5 and RCP6.0) but worsen under higher emissions (RCP8.5). Shared socio-economic pathways SSP1 and SSP2 have smaller impacts than SSP3, and RCP4.5/RCP6.0 have less impact than RCP8.5. New models perform better than older Coupled Model Intercomparison Project Phase (CMIP5) in predicting future climate conditions. For instance, a study by Hamed et al. (2022) compared the performance of two CMIP5 scenarios (RCP4.5 and RCP8.5) with their CMIP6 counterparts (SSP2-4.5 and SSP5-8.5) in Egypt. The results indicated that CMIP6 models exhibit less uncertainty in accurately simulating changes in seasonal air temperatures and rainfall compared with CMIP5 models. Ming et al. (2021) described the key message from the Sixth Assessment Report (AR6) of Intergovernmental Panel of Climate Change (IPCC), global surface temperatures are projected to increase by a minimum of 2.1 °C by 2100 compared with 1850–1900 under all emission and global precipitation is expected to rise by up to 13% by 2100 compared with 1994–2010, but not uniformly, with some regions, like parts of the subtropics, facing decreased rainfall. So, it is necessary to examine regional-scale climate patterns to enhance the understanding of localized climate behaviour. To grasp future regional climate change and its effects, global climate models/general circulation models (GCMs) serve as reliable tools. These models project various hydro-meteorological variables under different potential climate scenarios. Utilizing projections from multiple climate models is essential for gaining insights into model uncertainties and for the development of effective risk management strategies (Kang et al. 2009; Hannah 2022).

In the Lower Mahanadi Basin (LMB), characterized by high variability in basin yield and river discharge, floods occur during wet years while water scarcity prevails in dry years. Rapid economic growth, industrialization, urbanization, and population growth in states like Chhattisgarh and Odisha exacerbate water stress. Future droughts could lead to conflicts over water allocation within and between states. Climate change, with an increased frequency of extreme events, further compounds these challenges (Kumar & Bassi 2021). The LMB is a crucial agricultural region known for its fertile lands and significant agricultural productivity. With heavy reliance on irrigation for crop production, accurate projections of IWRs are essential for optimizing agricultural practices and ensuring sustainable water use. Climate change exacerbates these challenges, highlighting the urgency of adaptive water management strategies. This study's projections not only address immediate agricultural needs but also inform policy decisions and long-term planning efforts (Naha et al. 2022).

This paper contributes to the existing literature by concentrating on the establishment of a framework for determining the seasonal boundaries of maximum (max) and minimum (min) values for rainfall, temperature, and IWR in an LMB situated in the state of Odisha in India for future periods near-future (2025–2050), mid-future (2051–2075), and far-future (2075–2100). As different models are expected to give different results, to enhance result reliability, various models are applied in this study. Six models demonstrated superior performance based on statistical parameters in the analysis of 13 bias-corrected climate models from the more recent CMIP6 framework, introduced to address limitations in CMIP3 and CMIP5 models (Gupta et al. 2020), utilized for examining regional climate impacts under two scenarios, SSP370 and SSP585. By integrating advanced climate models and rigorous analysis, the research offers valuable insights into the nuanced impacts of climate change on paddy crop irrigation. It enhances understanding of future climate risks and informs targeted adaptation strategies crucial for safeguarding food security and livelihoods in vulnerable regions like the LMB. It fills a crucial gap by examining the specific challenges faced by smaller basins, which are often overlooked but disproportionately affected by climate variability. This will help us grasp the lasting effects of climate change on development and mitigation. It will also assist local water managers, researchers, and policymakers in creating plans to adapt local agricultural water management.

The LMB spans across 20 districts of Odisha, delineated in Supplementary Figure S2. Below the Hirakud dam, it is primarily nourished by major tributaries such as the Ong and Tel rivers. The total population of the Mahanadi Basin, according to the 2001 census, amounts to approximately 2,83,22,294 (approximately 3 million), with over 1.49 million residing in the LMB. This region plays a vital role in Odisha's socio-economic fabric. Paddy farming dominates the region, making a substantial contribution to the state's agricultural production, which is primarily supported by its rural populace. Approximately three-fifths of its workforce is involved in agriculture, a sector contributing around one-sixth of the state's gross product. Cultivated lands cover roughly one-third of the state's total area, with about three-fourths of these lands dedicated to paddy cultivation. This emphasizes the significant role of agriculture, particularly paddy farming, in Odisha's economy and livelihoods (Das & Huke 2024). Prioritizing water conservation, infrastructure development, and sustainable agricultural practices is essential to ensure a prosperous future for the state's agricultural sector and rural communities. Additionally, conducting spatiotemporal studies of IWR in the basin is crucial for understanding water availability and usage patterns, and addressing potential challenges effectively.

In this study, meteorological datasets, involving daily Tmax, Tmin, and rainfall, were acquired over a 34-year period (1981–2014) as a baseline from the India Meteorological Department (IMD) (http://www.imdpune.gov.in/) for 84 geographical latitudes and longitudes which comes under the LMB. Rainfall data were accessible at a grid resolution of 0.25° × 0.25° while the temperature data was taken at the spatial resolution of 1° × 1°. This dataset is set as an observed dataset for the study. A short summary of the temporal period and other details of data used in the study is shown in Table 1.

To project the future climate change on IWR, the downscaled CMIP6 GCMs datasets, subject to bias correction, were developed by Mishra et al. (2020) for South Asia were taken. The available datasets include daily bias-corrected data generated using the Empirical Quantile Mapping approach. These datasets encompass historical data from 1951 to 2014 and future projections from 2015 to 2100 across four distinct scenarios, shared socio-economic pathways (SSP126, SSP245, SSP370, and SSP585). The dataset employed for analysis consists of downscaled high spatial resolution data measuring 0.25° × 0.25° from 13 CMIP6 GCMs (refer to Supplementary Table S1). Three time horizons were taken for the future climate change detection and their implication on IWR under two climatic scenarios (SSP370 and SSP585) in LMB, which are near- (2025–2050), mid- (2051–2075), and far- (2076–2100) future and the baseline period is 1981–2014. SSPs forecast future greenhouse gas emissions up to the year 2100, considering various global socio-economic changes. They outline trajectories based on a no-climate-policy baseline scenario post-2010, resulting in a temperature increase of 3.00–5.00 °C by 2100. These SSPs, akin to RCPs, can be associated with climate policies, offering diverse outcomes at the century's end with radiative forcing levels of 2.6, 3.4, 4.5, 6.0, 7.0, and 8.5 W/m² in 2100 (Riahi et al. 2017).

In the CMIP6 phase, the scenarios of the RCPs, namely RCP2.6, RCP4.5, RCP6.0, and RCP8.5 from CMIP5, have been replaced with SSPs. These include SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5, each considering radiative forcing levels for 2100. The climate change research community introduced SSPs to facilitate integrated analysis of future climate vulnerabilities, impacts, adaptation, and mitigation (Riahi et al. 2017). SSPs encompass five scenarios: (1) a trajectory marked by sustainability-focused growth and equality (SSP1); (2) a scenario where trends broadly adhere to historical patterns (SSP2); (3) a world characterized by ‘resurgent nationalism’ and fragmentation (SSP3); (4) a scenario defined by escalating global inequality; and (5) a situation of rapid and unconstrained growth in energy use and economic output (SSP5), developed by the climate change research community, the SSPs represent an innovative scenario framework (Kriegler et al. 2014). This framework aims to streamline the comprehensive analysis of forthcoming climate impacts, and vulnerabilities, as well as adaptation and mitigation strategies. Developed collaboratively over recent years, the pathways delineate plausible global trajectories that would, in time, pose distinct challenges for climate change mitigation and adaptation. These SSPs hinge on five narratives delineating alternative socio-economic progressions, encompassing sustainable development, regional competition, inequality, fossil-fuel-driven growth, and moderate development.

The analysis aimed to calculate the future irrigation water needs for Kharif paddy in the LMB. It also examined seasonal rainfall and temperature variations across 84 grid stations from 2025 to 2100.

The approach of the study involved the analysis of gridded data on rainfall, Tmax, and Tmin. This dataset facilitated the computation of reference crop evapotranspiration using a computational program. Subsequently, the same data informed the assessment of water requirements, encompassing various agricultural activities such as field preparation, nursery operations, transplantation, and percolation processes. The basin's average annual rainfall based on the gridded rainfall data of the IMD for a baseline period of 34 years is approximately 1440.039 mm. The spatial distribution of the seasonal average values for both rainfall and IWR can be seen in Supplementary Figure S1, most of the area encounters a seasonal rainfall ranging from 1,063 to 1,359 mm. However, specific locations, such as certain parts of the northern and southern regions, experience higher rainfall within the range of 1,500–1,800 mm. A similar trend is evident in the spatial distribution of IWR, where areas with high rainfall exhibit a narrower range of IWR, typically between 290 and 340 mm. In contrast, most locations fall within the broader range of 340–426 mm.

Seasonal rainfall exhibits unique values for each GCM across different time spans. In the mid-future, a distinct pattern emerges in min rainfall projections. Notably, INM-CM-5-0, MPI-ESM1-2-HR, and NorESM2-MM depict a decrease in rainfall for SSP370 compared with the near-future. Conversely, for SSP585, there is an upward trend in rainfall values across all time spans for each GCM. Reflecting on these findings, it is noteworthy that INM-CM-5-0 stands out with its unique temperature patterns, emphasizing the significance of considering individual GCM behaviour (Hannah 2022). This reinforces the importance of understanding GCM-specific variations to enhance the reliability of climate projections.

By consolidating data from various GCMs, a comprehensive range for average seasonal max and min IWR, rainfall, and temperature has been established. This detailed information is presented in Table 4 for each period and emission scenario, highlighting notable trends. In the SSP370 scenario from near-future to far-future, min rainfall exhibits fluctuation, ranging from 1217.07 to 1313.82 mm, signifying a 23.7% increase from the baseline. Conversely, max rainfall experiences more significant variation, spanning 1640.47 to 2011.37 mm, reflecting an 11.5% increase compared with the baseline period. Under the SSP585 scenario, max and min IWR decrease by 9.53 and 20.2% and max rainfall increases by 28.9% relative to the baseline period in the far-future. Despite the overall increase in rainfall and a decrease in IWR anticipated in the near-, mid-, and far-future, the exploration through spatiotemporal mapping reveals specific locations where IWR is increasing. This emphasizes the critical importance and precision of spatiotemporal studies in detecting regional climate change impacts. However, a distinct pattern emerges in temperature trends, indicating a more pronounced increase in Tmin compared with Tmax in both SSP370 and SSP585 scenarios. In the SSP370 scenario, Tmin ranges from 25.7 to 27.1 °C, while Tmax shows a comparatively narrower variation, ranging from 34.1 to 35.3 °C. This parallel variation is observed in the SSP585 scenario as well. This emphasizes the importance of recognizing the temperature dynamics for effective climate adaptation planning. Both the average seasonal Tmax and Tmin are going to increase by 3–4 °C by the end of the 21st century with respect to the baseline period in both scenarios (Mishra et al. 2020; Salunke et al. 2023).

In the LMB with diverse terrain like hilly regions and coastal plains, the basin's water availability is influenced by hard sedimentary and alluvial systems (Kumar & Bassi 2021). Understanding these complexities is crucial for effective water resource management and evaluating climate models' performance in replicating key variables like rainfall and temperature.

The results of the present study show that the spatial distribution of rainfall influenced the IWR, with areas of high rainfall exhibiting a narrower range of IWR (290–340 mm) compared with locations with lower rainfall (340–426 mm) (Bassi 2023). Initially, the baseline study indicates a tendency for irrigation requirements to decrease as rainfall increases. However, as the study progresses into future time periods, the situation becomes more complex. Each selected GCM model presents a distinct projection for IWR. While some GCMs suggest an increase in IWR for the future, others indicate a decrease. This is because unexpected extreme events might change the situation down the road, as the Tmax and Tmin are also showing increasing trends in the future. This divergence underscores the variability and uncertainty inherent in future climate projections and emphasizes the importance of considering multiple models in assessing future irrigation needs (N'guessan et al. 2023).

Analysis of the performance of various GCMs in the study area highlights the importance of selecting suitable models for climate change research and water resource management, because relying solely on the outcomes of a single climatic model is insufficient. So, to establish a more reliable and comprehensive range for climatic parameters like rainfall, Tmax, and Tmin, it is prudent to incorporate results from multiple GCMs (Ahmed et al. 2019). Further study of average rainfall and temperature reveals that one or two GCMs stand out with unique temperature patterns, emphasizing the importance of understanding individual GCM behaviour for reliable climate projections (Thao et al. 2022).

As discussed in the result section, under both the intense forcing SSP (SSP370 and SSP585) emission scenarios, all GCMs indicate an increase in rainfall (Abbas et al. 2023). The findings suggest that rainfall increases will be more pronounced in the far-future which further indicates that adaptation strategies should account for long-term changes in water availability and the uneven spatial distributions of rainfall imply that certain regions could face challenges in managing water resources, necessitating targeted planning efforts (McCrystall et al. 2021). Examining spatiotemporal shifts in IWR under diverse climate scenarios unveils model-specific trends. SSP585 indicates heightened extreme IWR events, with distinct spatial variations. Certain areas show elevated IWR levels. Understanding these shifts is crucial for effective water resource management and agricultural planning (Ahmad et al. 2023).

All these trends suggest that the region may experience more extreme weather events in the future, potentially leading to increased water stress and challenges in water management. Irrigation constitutes a substantial portion of this usage, highlighting the significance of efficient water allocation and management practices to sustain agricultural activities in the region (Samuel et al. 2017). Studies suggest that benefit-sharing mechanisms can play a crucial role in improving water allocation among riparian states facing water stress, as demonstrated in the case of the Mahanadi Basin. Such approaches could be instrumental in mitigating water deficits and enhancing the equitable distribution of water resources among stakeholders (Bassi & Chaturvedi 2024).

The selected GCMs collectively indicated an overall increase in rainfall, with distinctive variations among them. Notably, MPI-ESM1-2-LR projected a significant rise, while EC-Earth3 displayed a more modest increase. Intriguingly, some models forecasted a temporary decline in rainfall during the mid-future under SSP370, followed by a substantial increase in the far-future. The range of IWR values is approximately similar for both scenarios. However, a closer examination of spatial and temporal analyses reveals diverse trends, with MPI-ESM1-2-LR predicting the highest decrease in the mid- and far-future. The spatial distribution of extreme (highest and lowest) values of IWR varied between scenarios, SSP585 is associated with more extreme IWR events. There are places which have high rainfall as well as higher values of IWR, suggesting that the region may experience more frequent and intense rainfall events, underlining the critical role of spatiotemporal analyses in understanding regional climate change impacts.

Temperature projections further emphasized the unique characteristics. Importantly, findings project a more pronounced increase in Tmin compared with Tmax in both SSP370 and SSP585 scenarios, with a projected 3–4 °C rise by the far-future relative to the baseline period. The projected increase in temperature suggests that the region may experience more heatwaves, which could lead to increased heat-related illnesses.

The research contributes vital information for climate modelling and adaptation planning. The study highlights the need for careful consideration of spatiotemporal analysis in understanding climate change. It serves as a useful guide for policymakers, agricultural researchers, and water resource managers, aiding them in adopting effective, long-term crop planning strategies, mitigation and adaptation strategies, stressing the importance of using diverse climate models for reliable predictions.

The authors would like to acknowledge the Central University of Jharkhand for pursuing this research.

The authors received no funds, or grants for the preparation of this manuscript.

All authors contributed to the study's conception and design. Data collection and analysis were performed by R.K.J. and H.P.S. Program-based IWR analysis was first done by R.K.J. The spatiotemporal analysis of all the climatic parameters and IWR was first done by P.K. along with the preparation of the first draft of the manuscript and all authors commented on previous versions of the manuscript.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.