Climate change impact on crop stress and food security in a semi-arid river basin

Arid and semi-arid regions occupy a significant portion (31%) of the land surface on the Earth. These regions are water constrained due to lower rainfall and higher evaporative losses. The higher population density in such areas increases water consumption, resulting in water scarcity issues. Over the last century, an average global temperature rise of 1 °C is noticed compared to the pre-industrial era due to increased greenhouse gas emissions, among which 2015–2019 were recorded as the hottest five years (IPCC 2018). The elevated concentration of carbon dioxide and global industrialization are mainly responsible for the emissions of greenhouse gases and, thus, global climate change (IPCC 2021). Climate change is estimated to increase globally in the coming decades and will probably reach or exceed 1.5 °C, resulting in shorter cold seasons, long warm seasons, and increasing heat waves, which threatens agriculture and human health (IPCC 2021). Zhao et al. (2021) investigated the semi-arid Indus River basin in India using the Coupled Model Intercomparison Project Phase 5 (CMIP5) under various warming scenarios. They reported tremendous warming (∼3 °C) in the basin, specifically in RCP8.5, threatening the people living there. Further warming may accelerate regional climate shifts and increase aridity (Zarch et al. 2015). Global warming directly impacts agriculture, which influences global food security (Ainsworth & Ort 2010). Temperature regulates the rate of metabolic processes in plants, which, in turn, impacts the production of grains, fruits, and biomass (Mondal et al. 2016). Thus, the agricultural sector is considered the most crucial and immediate concern in terms of the societal impacts of climate change (Lipper et al. 2014). Therefore, analysing the effects of climate change on agriculture is vital in terms of water availability in the region and food security.

Climate change is also causing uncertainty and irregularity in precipitation patterns in terms of simultaneous floods and droughts in many regions (Seneviratne et al. 2012). The climate is strongly influenced by changes in rainfall and temperature, which ultimately shape the dynamics of wet (humid) and arid (dry) conditions. Thus, assessing long-term aridity can help monitor the impacts of climate change and drought conditions to effectively manage agricultural lands through various strategies (Marani-Barzani et al. 2017). Aridity can be defined as ‘the degree to which a climate lacks moisture at a given location; it is contrary to humidity’ (Ramachandran et al. 2015). It helps quantify the gap between rainfall and water demand (Salvati et al. 2012). Higher aridity implies more dryness (i.e., implying water shortages) and vice versa. It is worth mentioning that the increase in global temperature leads to an increase in aridity (Straffelini & Tarolli et al. 2023), which may cause severe future droughts in large food grain production regions, especially the United States, Canada, China, and India (Bruins & Bu 2006). Rainfall and temperature are the functions that define the aridity index (AI). Therefore, detecting the long-term changes in future rainfall and temperature trends needs to be adequately addressed to understand water resource distribution (dryness/wetness) by assessing the historical and future changes in climatic variables under various representative concentration pathways (RCPs).

Several investigators have analysed the future changes in climate variables with respect to the historical period for different study domains, viz., Northeast Italy (Straffelini & Tarolli 2023), Iran (Mirmehdi et al. 2023), and India (Koteswara Rao et al. 2020), using CMIP5 projections. At a global scale, Avila-Diaz et al. (2020) investigated the climate extremes using CMIP5 datasets and reported significant warming under RCP4.5 and RCP8.5 scenarios in Brazil by the end of the century. The consecutive wet days (consecutive dry days [CDDs]) are expected to decrease (increase) in the region. To ensure the robustness of general circulation models (GCMs), many researchers agreed on adopting a multi-model ensemble (MME) mean approach (Ahmed et al. 2020). The spatio-temporal variability of the AI in Greece will be investigated from 2021 to 2100 by adopting the MME mean approach (Nastos et al. 2013). The study reported a decrease in the AI, indicating drier conditions in the region. Arid environments are characterized by sufficient energy and limited moisture conditions, which present severe challenges to crop production. Milovanović et al. (2022) studied the variations in rainfall, temperature, and AI for RCP4.5 and RCP8.5 scenarios in Serbia, Europe. They indicated much higher temperatures and drought risk under the RCP8.5 emission scenario. Straffelini & Tarolli (2023) reported a prominent hydroclimatic transition from wetter to drier conditions in Northeast Italy while analysing the past and future changes in rainfall, temperature, and AI. In the Indian context, Rao et al. (2020) utilized a multi-model mean of 18 CMIP5 GCMs and agreed with the tremendous warming and heat stress across India by the end of the 21st century for both RCP4.5 and RCP8.5 scenarios. Kumar et al. (2021) used CMIP5 GCMs to analyse the variability in climate extremes across India and reported the increasing flood risk in the future for Sabarmati, Tapi, Brahmani, Baitarni, Subarnarekha, and north-eastern river basins. Rani et al. (2022) studied the semi-arid Banas River basin in India to explore the role of aridity in modulating the cropping patterns of the region. The study reported fluctuations in the AI due to changing rainfall and temperature patterns from 1971 to 2013, which intensified the aridity and drought conditions, potentially affecting the region's cropping patterns in the future. Choudhary et al. (2023) investigated the spatial variability of aridity across India from 1969 to 2017. They indicated a shifting climate towards an arid regime in northeast India, whereas a more humid regime may prevail in western India. This transition in climatic conditions in western India was attributed to frequent floods in recent years.

Crop production is more susceptible to climate change and extreme weather events due to widespread spatio-temporal fluctuations in temperature, rainfall, frequent floods, and droughts (Chandio et al. 2020). The damages due to heat stress in crops are notably pronounced during their critical development stage, specifically, the reproductive period, and affect its development, productivity, and growth (Kamkar et al. 2023; Khan et al. 2023). Therefore, examining the potential alterations in future climate extremes is imperative for crop water management and food security. This aspect demands significant attention from various water resource managers to effectively address local disaster risks and formulate adaptive measures. According to a report from the Ministry of Earth Sciences of India, agricultural productivity, freshwater resources, and natural ecosystems are facing significant stress due to variability in temperature and rainfall patterns (Krishnan et al. 2020). Extreme heat stress adversely affects crop productivity, especially during crop reproduction (Teixeira et al. 2013). Li et al. (2023) examined the impact of escalated temperature on wheat and reported a positive correlation between temperature and wheat grain yield up to ∼33 °C. A declination in the correlation is noted when the maximum temperature exceeds a particular threshold, i.e., >33 °C. However, investigating the heat stress levels on Kharif and Rabi crops in a changing climate has not received wide attention in the scientific literature. Moreover, smaller basins exhibit higher susceptibility to regional climate change, which considerably impact the crops (Karan et al. 2022). Though the previous studies mainly focused on the variability of rainfall and temperature projections, limited discussion is available focusing on the linkages between climate change, aridity, and crop stress at the basin scale.

The Sabarmati River basin (SRB) is classified as an economically water-scarce basin with the lowest per-capita water availability (i.e., 240 m³ per-capita per year), resulting in the food production deficit (Amarasinghe et al. 2005). The livelihood of people in this densely populated basin depends primarily on agriculture, which covers around 73% of the basin area. Areas with high population density are specifically vulnerable to climate change and heat stress (Diffenbaugh et al. 2007; Akerlof et al. 2015). Therefore, to effectively address the impact of climate-related heat stress, it is crucial to assess and quantify the spatio-temporal variation of the climatic variables and their implications on agriculture. Moreover, a district-level assessment of climate change and crop stress would help stakeholders make informed decisions and foster climate-resilient agriculture, ultimately benefiting communities, economies, and the environment. Keeping this in mind, the key objectives of the present analysis are as follows:

• (i)

  Analyse the spatio-temporal changes in future aridity across the SRB with reference to the baseline period.

• (ii)

  Explore the dependency between climate indices and aridity for the baseline and future periods.

• (iii)

  Assess the temporal propagation of crop stress at the district level in the SRB.

Daily gridded rainfall and temperature data for the baseline period of 1951–2019 were obtained from the Indian Meteorological Department (IMD), Pune. The rainfall data were available at 0.25° × 0.25° spatial resolution (72 grid points), while the temperature data, originally available at 1.0° × 1.0° resolution, were re-gridded to 0.50° × 0.50° resolution (25 grid points) using the bilinear interpolation method for better spatial coverage (Table 1). The digital elevation model at 30 m spatial resolution was obtained from the Shuttle Radar Topography Mission (SRTM) (Figure 1(a)). Based on the satellite images, the land use is classified into six classes for the study region: forest, water bodies, agricultural land, built-up area, fallow land, and scrubland (Figure 1(b)). The present study uses rainfall and temperature datasets of five statistically downscaled GCMs at 0.25° × 0.25° and 0.50° × 0.50° spatial resolution, respectively, for the future period (2020–2100) (Table 2). These downscaled datasets are available on the web portal of the National Institute of Hydrology, Roorkee, India (http://117.252.14.227/). These CMIP5 models (BNU ESM, CCCmaCanESM2, CNRMCM5, MPIESM-LR, and MPIESM-MR) are bias corrected using Kernel regression-based statistical downscaling method (Kannan & Ghosh 2013) for RCP4.5 and RCP8.5 scenario. RCP4.5 represents a scenario focused on mitigation and stabilization (Thomson et al. 2011), while RCP8.5 represents the strongest scenario without any climate mitigation objective (Riahi et al. 2011). It is worth mentioning that, under the RCP8.5 scenario, the impact of global atmospheric warming on the global monsoon is more pronounced (Kitoh et al. 2013). Thus, the present study considers RCP4.5 and RCP8.5 scenarios to investigate the climate change impacts on the hydrology of the SRB.

The CMIP5 provides concentration pathways for greenhouse gases and radiative forcing, making it more directly relevant to our investigation of climate variables and their trends (Frame et al. 2018). On the other hand, the shared socioeconomic pathways of the CMIP6 offer valuable insights into socioeconomic and policy-driven scenarios that influence the emissions pathways (Frame et al. 2018). Gusain et al. (2020) compared the CMIP5 and CMIP6 models to analyse the Indian summer monsoon rainfall (ISMR) variability for the baseline period (1951–2005). The study highlighted that CMIP5 and CMIP6 demonstrated similar capabilities in simulating ISMR variability, with a few regional exceptions. Furthermore, their analysis also confirmed that the majority of CMIP5 models successfully captured the climatological pattern of the ISMR.

The non-parametric modified Mann-Kendall (MMK; Hamed & Rao 1998) test explores the statistical significance of trends in climate variables. The null (⁠⁠) and alternate (⁠⁠) hypotheses of the MMK test state the absence and presence of a monotonic or non-monotonic trend in the climate time series, respectively. The Spearman's rank correlation (Lehmann & D'Abrera 1975) is employed to investigate the statistical association between the climate and aridity indices for the baseline and future periods. Details of these methods are available in Supplementary Material Sections S1.1 and S1.2.

There are 27 climate indices recommended by the expert team on climate change detection indices to assess the alterations in the characteristics of climate extremes and widely accepted by the research community (Zhang et al. 2011; Sharma et al. 2020). The present study adopts three rainfall and temperature indices to analyse their linkages with aridity and their space–time variability across the SRB. Details of climate indices adopted in this study are discussed in Table 3. In this study, rainy days (RDs) are defined as the days on which rainfall equals or exceeds 2.5 mm. Precipitation (PRCP) is denoted by the total rainfall occurring during RDs in a year.

The United Nations Environment Programme (UNEP) has proposed the AI, another climate index used in the present study to investigate the transition in the climatic conditions of the basin. AI is a dimensionless parameter representing the measure of water availability in the ecosystem of a region (Subramanya 2008) (Table 3). Lower AI values correspond to a drier (or moisture-constrained) climate and vice versa. This index is useful for recording the evolution of the drought phenomenon. This study calculates AI for 17 grids common to rainfall and temperature in the SRB. Here, potential evapotranspiration (PET) is calculated using the Hargreaves method (Hargreaves & Samani 1985), a temperature-based approach, which is widely adopted for semi-arid and arid regions (Allen et al. 1998). The criteria given by UNEP (1997) are utilized in the present study to classify the AI: hyper-arid (AI < 0.05), arid (AI ≈ 0.05–0.20), semi-arid (AI ≈ 0.20–0.50), sub-humid (AI ≈ 0.50–0.65), and humid (AI > 0.65).

In this analysis, the MME mean approach is used to investigate the future changes in climate indices. The MME mean approach reduces the biases and uncertainties in future climate simulations generated by several GCMs. The complex physics involved in GCMs may lead to inherent errors and uncertainties (Ahmed et al. 2020). The MME mean approach addresses individual errors to offer a better fit for the analysis compared to individual GCMs. Sun et al. (2020) reported robust performance of the MME mean approach to the individual models using CMIP5 GCMs. The RCP4.5 and RCP8.5 ensembles of climate indices are calculated using the simple arithmetic mean method with no allocation of weights to individual GCMs during the ensembling process to avoid bias.

The physiological responses to heat stress exhibit consistent patterns among different crop species. For instance, under heat stress conditions, there is a reduction in the number of flowers per plant, impaired development of pollen tubes, limited release of pollen, and decreased viability and fertility of flowers (Suzuki et al. 2013). The present study calculates the impact of crop stress on selected Kharif (rice, groundnut, maize, and cotton) and Rabi (wheat) crops for the baseline and future periods at the district level. The crop stress is calculated considering two crucial temperature parameters, namely, critical temperature (⁠ and limiting temperature (⁠ The threshold temperature at which adverse effects on crop growth and development begin to occur is known as ⁠. On the other hand, the extreme threshold temperature beyond which the crop cannot survive or sustain growth is defined as ⁠. The present analysis follows two assumptions as indicated by Teixeira et al. (2013): (i) the sensitivity of crops is more during the reproductive phase of development (thermal-sensitive period [TSP], days), and (ii) yield damage resumes when > ⁠, and maximum impact will occur in crops when > Here, refers to the daily maximum temperature during the TSP of each crop. and for different crops are taken from the study by Teixeira et al. (2013) and given in Table S1 (Supplementary Material). The impact of increasing temperature on the crops is assessed using anthesis heat stress (AHS) during the reproductive phase of the crops. High AHS will reduce pollination, impair flower development, develop poor-quality seeds, and eventually lower the yield. In the present study, AHS is classified as very low (AHS = 0), low (AHS < 0.05), medium (0.05 ≤ AHS < 0.15), high (0.15 ≤ AHS ≤ 0.30), and very high (AHS > 0.30) (Sun et al. 2019). Detailed information about AHS is given in Section S1.3 (Supplementary Material).

In this section, the results of trend analysis of climatic parameters for the baseline (1951–2019) and future (2020–2100) periods are presented (Section 4.1). The grid-wise projected climatic indices (rainfall, temperature, and aridity) are derived using MME of five GCM models of CMIP5 for both RCP4.5 and RCP8.5 scenarios. The percentage deviation of the projected AI with respect to the baseline period is also estimated (Section 4.2). The association between the climate indices and AI is examined using Spearman's rank correlation test at a 5% significance level (Section 4.3). Finally, the district-level crop stress is derived for the baseline and future periods (Section 4.4).

The RDs over the SRB vary in the range of 28–43 days for the baseline period, most of them occurring during the monsoon or the Kharif period (Figure 3(d)). Like PRCP, the southern SRB experiences fewer RD than the rest of the basin. However, a significant rise in the RDs is projected in the future for RCP4.5 and RCP8.5 scenarios to the tune of 42–75 days and 40–82 days, respectively, with similar spatial patterns as that of PRCP (Figure 3(e) and 3(f)). It is evident that with the progression of time, the mean rainfall may likely increase by the end of the 21st century, implying increased water availability in the SRB.

The CDDs is an important indicator defining the sustenance of deficit moisture deficit conditions in the region. CDD varies between 171 and 216 days for the baseline period, indicating almost 6–7 months of dry spells in the basin during a year, particularly during the Rabi and summer months (Figure 3(g)). The headwater regions of the SRB experience fewer dry spells, which increase as we traverse from north to south. However, the future projections indicate a considerable reduction in the dry spells coherently for both the RCP scenarios (Figure 3(h) and 4(i)). Thus, SRB is projected to experience wetter meteorological conditions under both warming scenarios, ensuring soil moisture sufficiency to take up agricultural activities in the basin.

The annual mean maximum (T_max), minimum (T_min), and average (T_avg) temperatures across the SRB are 33, 19.5, and 26 °C, respectively, in the baseline period (Figure 4(a), 4(d) and 4(g)). The T_max and T_min across the SRB are projected to rise to 33 (36 °C) and 22 (23 °C), respectively, in the future period of RCP4.5 (RCP8.5) ensemble (Figure 4(b), 4(c), 4(e), and 4(f)). A significant rise in the temperature is projected over the SRB by the end of the 21st century. T_min is exhibiting a rather significant spatial rise compared to T_max and T_avg (Figure 4). By the end of the 21st century, a considerable increase in climatological temperature becomes evident due to the escalating radiative forcing pathway that reaches 8.5 Wm⁻² by 2100 in the RCP8.5 scenario. On average, the basin would experience around 1 °C (2 °C) rise in the warming levels by 2100 compared to the baseline period for RCP4.5 (RCP8.5). The increase in the minimum temperature, in particular, would adversely impact the ecosystem and agriculture through a sustained warming environment. Higher temperatures would drive higher amounts of rainfall across the SRB. According to the Clausius-Clapeyron (C-C) relation, for every 1 °C increase in temperature, the atmosphere's capacity to hold moisture increases by approximately 7% (Trenberth 2011). Thus, the increase in projected rainfall may be attributed to rising sea and land surface temperatures, which will accelerate evaporation and intensify the rainfall. The current study corroborates with the findings of Dash & Maity (2023), who reported a positive correlation between dew point temperature and rainfall during the monsoon period in north-western and peninsular parts of India for the period 1991–2020. Hence, a trade-off between increased temperature and rainfall must be considered in formulating effective water resource management strategies in a changing climate.

Our results corroborate with the findings of Goswami et al. (2006), who reported a significant increase in rainfall over central India post-1981, and Mukherjee et al. (2018), who indicated an increase in rainfall prominently in southern and central India. Our results corroborate with Panda et al. (2023), who reported an increase in summer monsoon rainfall by 43.73% in arid and semi-arid regions of India by 2080s using RCP4.5 and RCP8.5 scenarios of CMIP5. Murugavel et al. (2022) reported rapid warming of land mass in the pre-monsoon season and the gradual heating of the Arabian Sea. Due to the land–sea surface temperature contrast, the moisture-carrying wind will travel from the Arabian Sea to the Indian subcontinent, substantially increasing rainfall during the monsoon season (Mishra et al. 2021; Jasmine et al. 2022). SRB has witnessed an increase (decrease) in rainfall (dry spell length). These findings are consistent with Yaduvanshi et al. (2021), who also reported a significant increase in RDs over southern and northwest India (SRB lies in northwest India). Our findings also align with Feng & Fu (2013), who reported an areal expansion of the global drylands by ∼10% by 2100 (except in India and northern tropical Africa) and projected an increase in their wetness. Mondal et al. (2022) linked significant rainfall changes with increased radiative forcing under higher emission scenarios due to enhanced anthropogenic activities. Our findings regarding increased future warming corroborate with a previous study from India (Vijay et al. 2021) and are also consistent with the study on a global scale (Lawin et al. 2019). Increasing warming may impact both human well being and agricultural productivity, especially in semi-arid and water-scarce regions like SRB, where heat-related stress is already a significant concern (Yaduvanshi et al. 2021).

The MMK trend assessment for the AI shows a heterogeneous trend (i.e., 58.8, 11.8, and 29.4% of grids show increasing, significant increasing, and decreasing trends, respectively) in the baseline period (Figure 6). On the other hand, 58.8% (41.2%) of grids indicated a significant decreasing (increasing) trend in RCP4.5 (RCP8.5) ensemble (Figure 6). Thus, the role of moisture availability in leading the fluctuation in AI plays a vital role in the warming climate of the semi-arid SRB. The increasing AI in the SRB indicates a transition from a semi-arid to a sub-humid climate. The projected increase in rainfall (i.e., moisture) over the SRB may modulate the effect of rising temperatures. Thus, the higher rates of increase in rainfall overshadow the temperature escalation and, thus, the PET rise, which ensures sufficient moisture supply in the basin. The projected AI at the end of the 21st century, using ensemble mean from five GCMs, shows that humid conditions are expected to establish in some parts of the SRB, particularly the southern SRB. With increasing AI in the SRB, the water deficit may decrease, which would help in ecosystem restoration and agricultural productivity in the basin (Du et al. 2013).

The monsoon months account for 85–90% of the annual rainfall across the SRB, which overlaps with the Kharif period, while the drier conditions are prevalent in the Rabi period. On the other hand, the Rabi season experiences milder temperatures than the Kharif season; hence, the evaporative losses are relatively lower in the former season. The temperature variation in the Rabi (October to April) and Kharif (May to September) periods are investigated for RCP4.5 and RCP8.5 warming scenarios with respect to the baseline period (Table 4). It is observed that the average temperature during the Kharif period rises by ∼5.6% (∼8.3%) in RCP4.5 (RCP8.5) (Table 4). Comparatively, higher temperature rise of around ∼8.6% (∼13.5%) in RCP4.5 (RCP8.5) is noted during the Rabi period. The severity of changes in the average temperature is higher for RCP8.5 than RCP4.5 for all the districts of SRB (Table 4). Schwalm et al. (2020) reported that conditions associated with the RCP8.5 scenario may prevail in the future. Therefore, considering the escalating temperature episodes in the future, specifically RCP8.5, all districts of SRB are likely to face adverse effects by the end of the 21st century. Our result corroborates with the study by Sun et al. (2019), who reported crop stress in India due to increasing temperatures, posing a threat to agricultural production and food security by 2100. Madan et al. (2012) reported negative impacts of increasing temperature, such as reduced grain yield, increased spikelet sterility, altered grain quality, delayed flowering, and increased susceptibility to pests and diseases due to increasing temperature in rice. Further, a rise in temperature will reduce wheat yield, as reported by Mostafa et al. (2021). On the other hand, Gunawat et al. (2023) stressed the usage of frequent irrigation and high fertilizer application to increase the wheat yield in the semi-arid region of Rajasthan, India. In summary, our findings indicate the increasing risk of agricultural production due to heat stress in SRB for both Kharif and Rabi crops. Notably, agricultural lands encompass around 73% of the total SRB area. Hence, this condition could adversely impact food security, with larger socioeconomic implications, specifically for the cultivators, agro labourers, and agro-based industries. It may also escalate food grain prices due to a widening gap between supply and demand. On the other hand, farmers may adopt unsustainable practices, such as excessive irrigation or increased use of agrochemicals, to mitigate the effects of heat stress, which may negatively impact the environment in the long run.

Adaptation and mitigation strategies are required to overcome the negative impacts of AHS on crops, including the development of heat-tolerant crop varieties, sustainable and innovative agricultural practices, intelligent water management practices, and policies that promote sustainable farming systems. Such measures can reduce the socioeconomic and environmental consequences of increasing AHS and enhance resilience in changing climate conditions in the basin.

The study explored the spatiotemporal variation of aridity and climatic indices across the semi-arid SRB using daily rainfall and temperature values for baseline (1951–2019) and future (2020–2100) periods under two warming scenarios (RCP4.5 and RCP8.5). The dependency between climate indices and AI for baseline and future periods is also explored. Further, crop stress is estimated at the district level for baseline and future periods. Based on our findings, the key conclusions are as follows:

• (i)

  The projected total rainfall and RDs (dry spells) have been found to increase (decrease) with respect to the baseline period in the basin for both RCP4.5 and RCP8.5 scenarios, resulting in a wetter meteorological regime in the SRB. The significant increase in the minimum, maximum, and average temperatures in the future with reference to the baseline period is noted. These escalated temperature episodes are likely to bring more heat waves, catalyse higher water losses, and affect agricultural productivity and human health.

• (ii)

  The prevalence of wetter meteorological conditions overshadows the warming tendency in the basin. As a result, an increase in the AI is reported, which marks a transition from the present semi-arid climate towards the sub-humid type. Thus, the SRB is projected to experience moisture-sufficient conditions at the end of the 21st century.

• (iii)

  Strong interdependence of climatic indices and AI is noted under higher anthropogenic warming levels (i.e., the RCP8.5 scenario) compared to the baseline period and RCP4.5 scenario. However, AI is more associated with rainfall than temperature in the baseline and future periods.

• (iv)

  Increasing temperature episodes during the reproductive stage of the Kharif and Rabi crops, particularly in RCP8.5, exhibit high to very high antithesis heat stress levels, posing a threat to crop production in the region. However, it is worth noting that the increase in projected temperature during the Rabi period (non-monsoon) is more than that of the Kharif period. These changes may affect the water and food security in the SRB and, in turn, the socioeconomic conditions of the people primarily relying on agriculture for their livelihoods.

Suppose the emission path outlined in the RCP8.5 scenario becomes a reality. In that case, most of the SRB may face a significant challenge to ensure food security due to high temperatures and a rapidly growing population by the end of the 21st century. Therefore, policymakers must implement innovative agricultural practices to address and mitigate these challenges in the SRB.

The authors are thankful to the Directorate, Indian National Committee on Climate Change (INCCC), Ministry of Jal Shakti, Government of India (GoI), for funding the research project entitled ‘Impact of Climate Change on Water Resources of Sabarmati Basin’. The authors are grateful for the infrastructural support provided by the Centre of Excellence (CoE) on ‘Water Resources and Flood Management,’ TEQIP-II, Ministry of Human Resources Development (MHRD), Government of India, for conducting the study reported in the present paper. The authors are thankful to the India Meteorological Department (IMD), Pune, and National Institute of Hydrology (NIH), Roorkee, India, for providing the essential data required to conduct the study.

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

The authors declare there is no conflict.