Temperature extremes (ExT) have risen significantly in many major land regions, including India, in recent decades. Projections suggest a more frequent occurrence of high-intensity ExT in India by the end of the 21st century, impacting ecosystems, health, agriculture, and the economy. Observations show a link between ExT occurrence and large-scale atmospheric circulation anomalies, as well as regional land-atmosphere feedback processes driven by soil moisture (SM) variations. SM influences ExT by modulating surface energy partitioning and evapotranspiration (ET), and its persistent nature (memory) can significantly impact near-surface temperature and precipitation variability. This study uses the MRI-AGCM3.2 model to investigate the impact of SM perturbations on ExT over India for the historical period (1951–2010) and future projections (2051–2100) under a 4K warming scenario. Previous studies on SM sensitivity experiments over India have mostly focused on SM-precipitation feedback, leaving the impact of SM perturbations on ExT unclear. This research aims to address this gap.

The introduction extensively reviews existing literature on increasing global and regional heatwaves, focusing on India's vulnerability. It highlights previous research on the link between extreme heat events and large-scale atmospheric circulation, as well as regional land-atmosphere feedback processes influenced by soil moisture variability. The review also points out the limited understanding of the impact of soil moisture perturbations on temperature extremes over India, specifically the lack of studies focusing on the direct impact on extreme temperature characteristics.

The study utilized the Meteorological Research Institute (MRI) high-resolution (~60 km) atmospheric general circulation model (MRI-AGCM3.2) for simulations covering the historical period (1951–2010, HIST) and future projections (2051–2100, FUT) under a 4K warming scenario. The model's simulation of key surface hydro-meteorological variables (precipitation, SM, ET, and maximum temperature) was evaluated against observational datasets (IMD, GLDAS, LDAS), employing a bias correction method. Soil moisture-temperature (SM-T) coupling was assessed using linear regression, identifying north-central India (NCI) as a hotspot. Six long-term simulations were conducted: HIST, HIST-20% (SM decreased by 20%), HIST+20% (SM increased by 20%), FUT, FUT-20%, and FUT+20%. Extreme temperature indices (frequency, duration, intensity) were calculated using daily maximum temperature data exceeding the 90th percentile (HIST) for three or more consecutive days. The Generalized Extreme Value (GEV) theory was used to analyze extremes over NCI, incorporating SM as a covariate in a non-stationary GEV model. Soil moisture memory (SMM) was quantified using the e-folding time-scale method, examining its response to wet and dry SM perturbations. Statistical significance testing (t-test, Pearson correlation) was used throughout the analysis.

Model evaluation showed reasonable agreement with observations after bias correction. North-central India (NCI) was identified as a region of strong SM-T coupling. The historical simulation indicated at least 4 ExT events per year over India, with an average duration of 5-6 days and intensity around 47°C. The future simulation (4K warming) projected a dramatic increase in ExT events, duration, and intensity. SM sensitivity experiments revealed that drier conditions intensified ExT, while wetter conditions reduced them. This effect was more pronounced in the historical simulation. In NCI, a 20% decrease in SM increased ExT frequency by approximately 5 events per year, duration by 1.8 days per event, and intensity by 0.71°C (HIST). Conversely, a 20% increase in SM led to substantial ExT reductions. The GEV analysis confirmed the significant impact of SM on ExT, with a 20% decrease in SM leading to a 1.25°C increase in the 50-year return level of yearly maximum temperature in NCI. Changes in surface energy partitioning, particularly reduced latent heat flux and increased sensible heat flux under drier conditions, explained the impact of SM on ExT. Soil moisture memory also played a crucial role. The impact of SM perturbations on ExT frequency and duration diminished in the future scenario due to increased precipitation and soil moisture.

The findings demonstrate the significant influence of soil moisture on temperature extremes over India, particularly in north-central India. The observed changes in surface energy partitioning and soil moisture memory provide a mechanistic explanation for the impact of soil moisture on extreme temperatures. The reduction in the impact of soil moisture perturbations on extreme temperature frequency and duration under future warming scenarios is linked to increased precipitation and soil moisture, leading to reduced temperature differences between the surface and atmosphere. This highlights the complex interplay between soil moisture, temperature, and precipitation in a changing climate.

This study highlights the crucial role of soil moisture in modulating temperature extremes over India, particularly in north-central India. The findings emphasize the importance of considering soil moisture variability when assessing future climate change impacts. Further research could explore the impact of other land-atmosphere interactions and the potential for improved land surface modeling to better predict future temperature extremes.

The study is based on a single model simulation (MRI-AGCM3.2), limiting the generalizability of the results. Other models with varying representations of land-atmosphere coupling might yield different outcomes. The use of a 4K warming scenario represents a simplified representation of future climate change; more complex scenarios could provide a more comprehensive picture. While the study acknowledges model biases and performs bias correction, these corrections do not perfectly account for model uncertainties.