Baseline Temperature Variability Shaping Geographical Distribution of Future Hot Extremes under Anthropogenic Warming

Baseline Temperature Variability Shaping Geographical Distribution of Future Hot Extremes under Anthropogenic Warming | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Physical Sciences - Article Baseline Temperature Variability Shaping Geographical Distribution of Future Hot Extremes under Anthropogenic Warming Xiaohui Ma, Zhili Tang, Shenghui Zhou, Lixin Wu, Wenju Cai, Zhao Jing, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5393056/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 26 Nov, 2025 Read the published version in Communications Earth & Environment → Version 1 posted You are reading this latest preprint version Abstract Hot extreme events are among the most devastating disasters affecting human health and the natural environment. While there is broad consensus on an increasing severity of these events under anthropogenic warming, their geographical distribution exhibits substantial spatial heterogeneity, and its driving factors remain uncertain. Here, utilizing an eddy-resolving high-resolution climate model alongside multiple simulations from Coupled Model Intercomparison Project Phase 6, we find baseline temperature variability as a key factor shaping the global distribution of projected hot extremes, with over 80% of the global increase in hot extremes anticorrelated with baseline temperature variability. We further demonstrate that the baseline temperature variability is anchored by persistent land-atmosphere coupling, which endures over century timescales and sustains the spatial heterogeneity of future hot extremes. Our findings suggest that baseline temperature variability could serve as a potential indicator for future hot extreme distribution, offering valuable insights for developing targeted adaptation strategies and improving regional resilience. Earth and environmental sciences/Climate sciences/Climate change Earth and environmental sciences/Climate sciences/Climate change/Projection and prediction Figures Figure 1 Figure 2 Figure 3 Figure 4 1. Introduction Hot extreme events, characterized by prolonged periods of excessively high temperatures exceeding the average, have devastating impacts on human health, agriculture, and the natural environment 1–4 . One of the most severe events, the 2003 European heatwave, led to approximately 70,000 deaths and more than €15 billion in economic loss 5,6 . Similarly, the 2010 Russian heatwave resulted in over 55,000 deaths and a 25% reduction in crop production 7 . Under climate change, these events are expected to increase in frequency, intensity and duration 8 . Importantly, the geographical response of hot extremes to anthropogenic warming is highly heterogeneous, with some regions experiencing far more intense extremes than the global average, leading to disproportionately socioeconomic impacts 9–11 . Understanding the drivers behind the spatial heterogeneity is essential for systematically improving regional climate projections for adaptation strategies. While there is broad consensus regarding the increasing occurrence, magnitude of hot extremes under anthropogenic warming 12–15 , the physical processes driving their spatial heterogeneity are regionally dependent and mechanistically diverse 16 . One key driver widely discussed is a soil moisture-temperature feedback, where decreasing soil moisture leads to rising mean temperature due to reduced evaporation, and vice versa 10,17,18 . The feedback plays a crucial role in amplifying regional hot extremes, particularly in midlatitude regions. In high-latitude regions, snow/ice-temperature feedback modifies surface albedo, further affecting warming 19 . Additionally, atmospheric dynamical processes, such as blocking events and large-scale circulation patterns, significantly contribute to the regional variability of hot extremes. For example, intense European heatwaves have been linked to weakened atmospheric circulation and increased atmospheric blocking 20–22 . Sea surface temperatures (SSTs) may also impact adjacent land hot extremes by altering atmospheric circulation and heat advection 23–25 . These drivers are intricately connected—soil moisture can modulate atmospheric circulation, exacerbating hot extremes 26 ; while changes in atmospheric circulation can feedback onto soil moisture and SST 27 , creating a complex interaction system. Identifying the dominant factors and unraveling how these drivers contribute to the geographical distribution of projected hot extremes remain a challenge. Moreover, large uncertainties in the future evolution of soil moisture, snow/ice and atmospheric circulation in current climate models 8,28–30 add difficulty to accurately projecting how hot extremes will manifest regionally under anthropogenic warming. The issue of whether there is a systematic predictive indicator for the projection of regional variability in future hot extremes is critical yet unresolved. In this study, we analyze outputs from an unprecedented set of eddy-resolving, high-resolution simulations from Community Earth System Model (HR-CESM) with ~ 0.25° atmospheric and ~ 0.1° oceanic resolution and 10 climate models in Coupled Model Intercomparison Project Phase 6 (CMIP6, see “High-Resolution CESM and CMIP6 simulations” in Methods; Supplementary Table S1 ), to investigate the dominant factors and associated physical processes driving the spatial heterogeneity of future hot extremes in warming climate. 2. Results 2.1 Simulated and Projected Hot Extremes in HR-CESM and CMIP6 We first evaluate the models’ ability to simulate the observed characteristics of hot extreme events, including occurrence frequency, intensity, total duration, and cumulative heat (a hot extreme event is defined as at least three consecutive days with daily temperatures exceeding the 90th percentile local threshold during the summer season, see “Definitions of Hot Extremes and Related Metrics” in Methods for detailed definitions). The simulations are compared against ERA5 reanalysis data (the fifth-generation reanalysis from the European Centre for Medium-Range Weather Forecasts 31 ) for the period 1979 to 2019. Both HR-CESM and CMIP6 faithfully reproduce the spatial distribution of hot extreme total duration (our primary measure for hot extreme events, see “Definitions of Hot Extremes and Related Metrics” in Methods), with only slight differences in magnitude (Fig. S1 a-c). The same holds true for the occurrence frequency, intensity and cumulative heat distributions as well. In addition, we evaluate the models’ ability to represent observed trends in hot extremes over the historical period (Fig. S1 d-f). While both HR-CESM and CMIP6 capture regions with significant increases in hot extreme duration, they tend to slightly overestimate the trends. Overall, the models show high fidelity in simulating key characteristics of hot extremes. However, we note that increasing the resolution to eddy-resolving scales in HR-CESM provides limited improvement in their climatological simulation compared with CMIP6, a finding consistent with previous research 32 . We next examine the projected changes of hot extremes in HR-CESM under the high emission scenario of Representative Concentration Pathway 8.5 (RCP8.5). By the end of the 21st century (2081–2100), substantial global increases in the total duration, occurrence frequency, intensity, and cumulative heat of hot extremes are projected compared to the historical period (1981–2000, Fig. 1 a-d). The global averaged total duration of hot extremes is expected to rise from 2.57 days per summer to 66.46 days per summer, a more than 25-fold increase (Fig. 1 e). The increase in duration, together with a higher intensity, drives a substantial rise in cumulative heat, with future levels projected to exceed current values by over 70 times (Fig. 1 e). In addition, there is a near linear-relationship between the rise in hot extremes and summer mean temperature increase from 1979 to 2100, along with a hemispheric asymmetry in the response (Fig. 1 f). The more pronounced increase in hot extremes in the Northern Hemisphere is attributed to enhanced warming, driven by the greater land coverage in this region. These projected features align with previous findings 33–35 , confirming the robustness of hot extreme projections under warming conditions by HR-CESM. The projected changes in hot extreme events exhibit pronounced spatial heterogeneity (Fig. 1 a-d). Regions such as northwestern and eastern coastal North America, the Amazon region in northern South America, southern Europe, the Middle East, northern and central Africa, and South Asia experience larger increases in total duration (Fig. 1 a). In contrast, regions like the central U.S., northeastern Canada, southern South America, northern Eurasia, the Sahel, and Australia show a suppressed rise. The contrast in regional variability aligns closely with changes in intensity and cumulative heat, while the frequency of events shows an opposite trend. The relatively smaller increase in frequency in regions with prolonged duration response is attributed to more persistent hot extremes in a warming climate. We verify the hot extreme projections using CMIP6, across different warming periods and thresholds. The projected rising trend and hemispheric asymmetry for 2031–2050 are evident in CMIP6, albeit with a lower amplitude (Fig. S2 a-f). By mid-century, the global averaged hot extreme duration is projected to be approximately 11 times the current value (Fig. S2e), representing 40% of the increase expected in HR-CESM by 2100. Importantly, the spatial variability of projected hot extremes in CMIP6 closely mirrors that of HR-CESM (Fig. S2a-d), displaying similar hot and cold spot patterns, though the increase is more subdued in southern Europe. The above projections are based on a fixed threshold referenced to the historical period (T90 hist ). We also validate the results using a varying threshold referenced to future warming periods (T90 future , see “Definitions of Hot Extremes and Related Metrics” in Methods). Compared to the fixed threshold, the varying threshold results in an overall shorter but still statistically significant increase in hot extreme total duration (Fig. S3a). The spatial distribution of hot extreme increase remains similar, except for a less pronounced response in South Asia (Fig. S3a). To ensure consistent identifications of hot and cold spots across HR-CESM, CMIP6, and different thresholds, southern Europe and South Asia are excluded from the hot spot regions in the subsequent analysis when comparing temperature probability density functions (PDFs) between these regions. The exclusion is also supported by the K-means clustering results (see “K-means Clustering Analysis” in Methods) as discussed later. 2.2 Drivers of Spatial Heterogeneity in Future Hot Extremes Changes in temperature mean and variability are widely recognized as key factors influencing projections of hot extremes under warming 10,36,37 . However, the spatial distribution of projected hot extremes and mean temperature changes show little similarity, with only a weak and insignificant positive correlation under the RCP8.5 scenario in HR-CESM (Fig. S4a,b). Similarly, no clear correlation is found between the spatial distribution of projected hot extremes and temperature variability changes (Fig. 2 b, d). Instead, 83% of projected changes in global hot extremes are negatively correlated with historical temperature variability (Fig. 2 a). A linear regression shows a significant negative correlation, with a coefficient of -0.73 above the 99% confidence level (Fig. 2 c). This negative correlation also holds across over 71% of the global area when applying a varying threshold (Fig. S3). These findings are further validated in CMIP6 (Fig. S4c,d, S5), revealing a consistent negative correlation between the increase in hot extremes and baseline temperature variability across 93% of the global area. Together, these results suggest the robust role of baseline temperature variability in shaping the future geographical distribution of hot extremes. To better assess the relative contributions of temperature mean and variability in driving regional differences in hot extreme responses, we analyzed temperature PDFs for both high- and low-response regions. To minimize uncertainties in region selection, we apply the K-means clustering approach (see “K-means Clustering Analysis” in Methods), which produces high- and low-response regions that closely match the hot and cold spots consistently identified across HR-CESM, CMIP6, and various threshold definitions (Fig. S6a), confirming the robustness of the classification. The temperature PDFs of these two groups are distinct, with high-response regions exhibiting much narrower distribution and lower historical temperature variability than low-response regions (blue lines in Fig. 2 f, g). Specifically, the standard deviation of historical temperatures (the square root of variance) in high-response regions is only one-quarter that of the low-response regions (blue bars in Fig. 2 f, g). However, the projected total duration of future hot extremes in high-response regions is approximately twice that of low-response regions (red bars in Fig. 2 f, g). Under warming conditions, both temperature PDFs shift rightward, leading to a significant increase in the proportion of hot extreme days (red lines in Fig. 2 f, g). Projections show that T90p, representing the proportion of temperatures exceeding the 90th percentile, will rise to 79.8% in high-response regions in the future, compared to 45.3% in low-response regions. Interestingly, the mean temperature shifts are comparable in both regions, with a slightly greater warming observed in low-response regions (5.6°C increase in high-response regions and 5.9°C in low-response regions). This suggests that the greater increase in hot extremes in high-response regions is not attributed to differences in mean temperature shifts. The impact of baseline temperature variability on the increase in T90p is further quantified using idealized Gaussian temperature distributions (Fig. 2 e). For the same 6°C warming, T90p for temperature profiles with low baseline variability (one standard deviation) is expected to rise to nearly 100%, while for profiles with high baseline variability (four standard deviations), T90p increases to only 41.3%. The difference in response magnitude aligns with model projections, confirming that baseline temperature variability is the primary driver of the regional differences in future hot extreme intensification. 2.3 Physical Processes Influencing Baseline Temperature Variability To identify the physical processes contributing to the spatial variability of baseline temperature variability, we examine latent heat flux (LH), sensible heat flux (SH), and downwelling radiation (RD), all of which influence the surface energy balance that drives land surface temperature and, consequently, temperature variability 10,36 . Spatial comparison shows that across most of the globe, baseline temperature variance closely aligns with LH variance, except in the northern high latitudes (north of 50 °N), where it tends to be more influenced by SH and RD (Fig. 3 a-c). After excluding this region, scatter plots clearly reveal that the temperature variance is primarily driven by LH variance, with a smaller contribution from SH and a minimal impact from RD (Fig. 3 e, f). Soil moisture has been widely reported to influence the partition between LH and SH through land-atmospheric coupling 17,27,38 , emphasizing its potential role in driving temperature variability. This is corroborated by the strong spatial resemblance between temperature and soil moisture variance (Fig. 3 d). A closer examination of the scatter plots reveals two distinct regimes (Fig. 3 g). In the low-temperature variance regime, the scatter cloud is nearly vertical, indicating the predominant role of soil moisture. In the high-temperature variance regime, the scatter cloud aligns at a 45° angle, suggesting joint impacts from soil moisture and LH. The difference is likely linked to local dry or wet conditions, which affect the sensitivity of evapotranspiration responses to soil moisture, as suggested by previous studies 27,38 . These findings, consistent with prior research, support the critical role of soil moisture in modulating temperature variability. 2.4 Century-Scale Persistence in Land-Atmosphere Coupling The role of baseline temperature variability in driving the spatial distribution of future hot extremes over the next century raises an important question: what is the key driver behind this "century memory"? The above analysis reveals a strong link between soil moisture and temperature variability, primarily operating through a negative soil moisture-evaporation-temperature feedback according to previous studies 27 . This feedback is often referred to as land-atmosphere coupling and can be effectively captured by the correlation between the detrended evaporation and surface temperature 39 . Regions characterized by both low hot extreme response and high baseline temperature variability coincide with areas of active land-atmosphere coupling (Fig. 4 a). Notably, the coupling remains remarkably stable across historical and future simulations (Fig. 4 a, b), with nearly 95% of global areas showing persistent patterns (Fig. 4 c). The spatial match of active and inactive land-atmosphere coupling areas between past and future simulations shows a high correlation of 0.86 (Fig. 4 d), underscoring the persistence of this distribution over a century timescale. This enduring pattern is also corroborated by multiple CMIP6 models, indicating that it is not model-specific (Fig. 4 e). The results suggest that the spatial distribution of baseline temperature variability is intrinsically tied to land-atmosphere coupling, which operates over a long time and anchors the spatial heterogeneity of future hot extremes. 3. Discussion Using eddy-resolving HR-CESM simulations and multiple model results from CMIP6, we identify baseline temperature variability as a key factor shaping the uneven distribution of future hot extremes, a result validated across various climate models, warming levels, and thresholds. The baseline temperature variability is anchored by persistent land-atmosphere coupling, which operates over century timescales and sustains the spatial heterogeneity of future hot extremes. Moreover, the coupling primarily influences the temperature variability through soil moisture and latent heat flux, consistent with the current understanding of soil moisture-temperature feedback dynamics, and further supporting the robustness of the results. Existing studies have emphasized the role of changes in temperature mean and variability in determining hot extreme projections under warming 10,36,37,40–42 , underscoring the need for accurate representations of these factors for both historical and future periods to ensure reliable estimates of future hot extremes. However, the task remains challenging due to substantial uncertainties in projected changes in temperature mean and variability in future periods. Our findings show that regional variations in projected hot extremes are only weakly connected to changes in temperature mean or variability under warming. Instead, baseline temperature variability primarily determines the global pattern of hot extreme projections, highlighting it as a potential predictor for assessing regional variations in future hot extreme responses. These findings provide a simplified framework for assessing the geographic distribution of future hot extremes based on historical state without integrating into the future, offering valuable insights for developing more targeted adaptation strategies for regions most vulnerable to extreme heat under climate change. However, the results also stress the importance of more precise and comprehensive observations to quantify historical temperature variability, as well as improved climate models’ capability of accurately simulating this variability, to enhance projections of regional variations in future hot extremes. The findings may be subject to uncertainties due to their reliance on the century-scale persistence of land-atmosphere coupling, while accurately representing and projecting the coupling remains a major challenge for current climate models 16,28,43 . Methods High-Resolution CESM and CMIP6 Simulations The eddy-resolving high-resolution Community Earth System Model (HR-CESM) simulations with ~ 0.25° atmosphere and ~ 0.1° ocean components developed by the National Center for Atmosphere Research (NCAR) are analyzed 44 . The simulations consist of a 500-year preindustrial control run and a 250-year simulation covering historical (1850–2005) and future (2006–2100) climates. Historical forcing is applied from 1850 to 2005, followed by the RCP8.5 scenario (a high greenhouse gas concentration pathway) forcing from 2006 onwards. Daily outputs are used to assess hot extreme events and related variables, comparing two periods: a historical period from 1981 to 2000, and a future period from 2081 to 2100. Ten climate model simulations from the Coupled Model Intercomparison Project Phase 6 (CMIP6) are examined, including five from the High-Resolution Model Intercomparison Project (HighResMIP 45 ) and five pairing low-resolution simulations. The oceanic resolution of these models ranges from 0.1° to 1°, while atmospheric resolution ranges from 0.25° to 1°. The selected models and their specific resolutions are detailed in Table S1 . Each model includes a 100-year historical and future (under RCP8.5 scenario) simulation from 1950–2050. Hot extreme events during the period overlapping with ERA5 and HR-CESM are analyzed and compared. Definition of Hot Extremes and Related Metrics Following previous studies 14,46 , a hot extreme event is defined as at least three consecutive days with daily temperatures exceeding the 90th percentile threshold during summer (JJA in the Northern Hemisphere and DJF in the Southern Hemisphere). The 90th percentile is calculated for each calendar day at each grid point using a 15-day moving window centered around the targeted day over the reference periods 46 . Two thresholds are computed: one reference to the historical period (1979–2019, T90 hist ) and another to future warming periods (2060–2100, T90 future ). Correspondingly, hot extreme projections under anthropogenic warming are evaluated in two ways. For the fixed threshold, hot extremes in both historical and future periods are identified using T90 hist . For the varying threshold, hot extremes are identified using T90 hist for the historical and T90 future for the future period. Projected changes in hot extremes using the fixed threshold account for contribution from mean temperature increases between the historical and future periods, while the varying threshold excludes this effect. The occurrence frequency is defined as the number of hot extreme events per summer. The intensity of each event is measured by the maximum temperature anomaly above the 90th percentile threshold during the event and then averaged over the summer season. Total duration refers to the cumulative number of days during which hot extreme events occur per summer. Cumulative heat is the sum of temperature anomalies exceeding the 90th percentile threshold across all hot extreme events 14 . Notably, total duration is an effective indicator of projected hot extreme changes, capturing both occurrence frequency and mean duration response (Fig. 1 a). It is also the dominant contributor to cumulative heat increase (Fig. 1 d), a finding consistent with prior research 14 . Therefore, we use total duration as the primary measure of hot extreme responses in our analyses. K-means Clustering Analysis K-means clustering is applied to identify regions with high or low hot extreme responses based on projected changes in total duration and historical temperature variance 47 . Both variables are normalized (ranging from − 2.5 to 2.5 standard deviation) and divided into five categories, labeled from − 2 to 2, with each category spanning one standard deviation. This ensures comparability between the variables and prevents one from dominating the clustering due to scale differences. The K-means algorithm is then applied to the processed data and repeated 1,000 times with random initial cluster centroids, yielding an optimized clustering of 5. The two clusters with the largest and opposite variations in hot extreme changes and temperature variance are selected, which gives the high- and low-response regions shown in Fig. S6. As illustrated in the spatial map of hot extreme change and temperature variance (Fig. 1 a and Fig. 2 a), high-response regions correspond to areas with low historical temperature variance, and vice versa. These regions also closely match the hot and cold spots consistently identified across HR-CESM, CMIP6, and different threshold definitions, confirming the robustness of the classification. Note that in all identified regions, historical temperature variance is significantly lower in high-response regions compared to low-response regions. However, due to differing geographical locations, the temperature PDFs in these regions exhibit distinct mean temperatures, which can obscure the overall PDF shape when combined. To enhance clarity, the temperature PDFs for North Africa and Northern Eurasia (regions with distinct mean temperatures) are plotted separately (Fig. S6b, c), while the remaining regions are shown in Fig. 2 f, g. Declarations Data Availability ERA5 reanalysis can be downloaded from https://doi.org/10.24381/cds.bd0915c6. The CESM simulations can be achieved through https://ihesp.github.io/archive/products/ds_archive/Sunway_Runs.html. The CMIP6 data can be downloaded from https://pcmdi.llnl.gov/CMIP6/. Code Availability Python and Matlab codes to reproduce the analyses are available upon request from the corresponding author. Acknowledgments This research is supported by the National Natural Science Foundation of China (42376025), Science and Technology Innovation Program of Laoshan Laboratory (LSKJ202300302, LSKJ202202503), Shandong Provincial Natural Science Foundation (ZR2022YQ29), Taishan Scholar Funds (tsqn202103028). We thank Laoshan Laboratory in Qingdao and the National Supercomputing Center in Jinan for providing the high resolution CESM simulations and high performance computing resources that contributed to the research results reported in this paper. Author Contributions Z. T. performed most of the analyses under X. M.’s instruction. S. Z. detected the hot extremes and assisted with data processing. X. M. conceived the central idea, designed the study and wrote the manuscript. L. W. supervised the project. W. C. contributed to discussions on physical mechanisms influencing temperature variability. Z. J., Z. C. and B. G. contributed to interpreting the results and improving the manuscript. Competing Interests The authors declare no competing interests. References Lobell, D. B. & Field, C. B. Global scale climate–crop yield relationships and the impacts of recent warming. Environ. Res. Lett. 2 , 014002 (2007). Campbell, S., Remenyi, T. A., White, C. J. & Johnston, F. H. Heatwave and health impact research: A global review. Health & Place 53 , 210–218 (2018). Domeisen, D. I. V. et al. Prediction and projection of heatwaves. Nat Rev Earth Environ 4 , 36–50 (2022). Yin, J. et al. Future socio-ecosystem productivity threatened by compound drought–heatwave events. Nat Sustain 6 , 259–272 (2023). Robine, J.-M. et al. Death toll exceeded 70,000 in Europe during the summer of 2003. Comptes Rendus Biologies 331 , 171–178 (2008). García-León, D. et al. Current and projected regional economic impacts of heatwaves in Europe. Nat Commun 12 , 5807 (2021). WMO. Atlas of Mortality and Economic Losses from Weather, Climate and Water Extremes . vol. 1123 (2014). Seneviratne, S. I. et al. Weather and Climate Extreme Events in a Changing Climate. Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change 1513–1766 (2021) doi:10.1017/9781009157896.013. Seneviratne, S. I., Donat, M. G., Pitman, A. J., Knutti, R. & Wilby, R. L. Allowable CO2 emissions based on regional and impact-related climate targets. Nature 529 , 477–483 (2016). Vogel, M. M. et al. Regional amplification of projected changes in extreme temperatures strongly controlled by soil moisture‐temperature feedbacks. Geophys. Res. Lett. 44 , 1511–1519 (2017). Turnau, R., Robinson, W. A., Lackmann, G. M. & Michaelis, A. C. Model Projections of Increased Severity of Heat Waves in Eastern Europe. Geophysical Research Letters 49 , (2022). Perkins, S. E. & Alexander, L. V. On the Measurement of Heat Waves. Journal of Climate 26 , 4500–4517 (2013). Russo, S. et al. Magnitude of extreme heat waves in present climate and their projection in a warming world. JGR Atmospheres 119 , (2014). Perkins-Kirkpatrick, S. E. & Lewis, S. C. Increasing trends in regional heatwaves. Nat Commun 11 , 3357 (2020). Teng, H., Leung, R., Branstator, G., Lu, J. & Ding, Q. Warming Pattern over the Northern Hemisphere Midlatitudes in Boreal Summer 1979–2020. JOURNAL OF CLIMATE 35 , (2022). Donat, M. G., Pitman, A. J. & Seneviratne, S. I. Regional warming of hot extremes accelerated by surface energy fluxes. Geophysical Research Letters 44 , 7011–7019 (2017). Seneviratne, S. I. et al. Impact of soil moisture‐climate feedbacks on CMIP5 projections: First results from the GLACE‐CMIP5 experiment. Geophys. Res. Lett. 40 , 5212–5217 (2013). Douville, H., Colin, J., Krug, E., Cattiaux, J. & Thao, S. Midlatitude daily summer temperatures reshaped by soil moisture under climate change. Geophysical Research Letters 43 , 812–818 (2016). Diro, G. T., Sushama, L. & Huziy, O. Snow-atmosphere coupling and its impact on temperature variability and extremes over North America. Clim Dyn 50 , 2993–3007 (2018). Coumou, D., Lehmann, J. & Beckmann, J. The weakening summer circulation in the Northern Hemisphere mid-latitudes. Science 348 , 324–327 (2015). Horton, D. E. et al. Contribution of changes in atmospheric circulation patterns to extreme temperature trends. Nature 522 , 465–469 (2015). Rousi, E., Kornhuber, K., Beobide-Arsuaga, G., Luo, F. & Coumou, D. Accelerated western European heatwave trends linked to more-persistent double jets over Eurasia. Nat Commun 13 , 3851 (2022). Arblaster, J. M. & Alexander, L. V. The impact of the El Niño‐Southern Oscillation on maximum temperature extremes. Geophys. Res. Lett. 39 , 2012GL053409 (2012). Miralles, D. G., Teuling, A. J., van Heerwaarden, C. C. & Vilà-Guerau de Arellano, J. Mega-heatwave temperatures due to combined soil desiccation and atmospheric heat accumulation. Nature Geosci 7 , 345–349 (2014). Luo, M. & Lau, N.-C. Summer heat extremes in northern continents linked to developing ENSO events. Environ. Res. Lett. 15 , 074042 (2020). Koster, R. D., Chang, Y., Wang, H. & Schubert, S. D. Impacts of Local Soil Moisture Anomalies on the Atmospheric Circulation and on Remote Surface Meteorological Fields during Boreal Summer: A Comprehensive Analysis over North America. JOURNAL OF CLIMATE 29 , (2016). Seneviratne, S. I. et al. Investigating soil moisture–climate interactions in a changing climate: A review. Earth-Science Reviews 99 , 125–161 (2010). Cheng, S., Huang, J., Ji, F. & Lin, L. Uncertainties of soil moisture in historical simulations and future projections. JGR Atmospheres 122 , 2239–2253 (2017). Qu, X. & Hall, A. On the persistent spread in snow-albedo feedback. Clim Dyn 42 , 69–81 (2014). Shepherd, T. G. Atmospheric circulation as a source of uncertainty in climate change projections. Nature Geosci 7 , 703–708 (2014). Hersbach, H. et al. The ERA5 global reanalysis. Quart J Royal Meteoro Soc 146 , 1999–2049 (2020). Schiemann, R. et al. The Resolution Sensitivity of Northern Hemisphere Blocking in Four 25-km Atmospheric Global Circulation Models. J. Climate 30 , 337–358 (2017). Almazroui, M. et al. Projected Changes in Climate Extremes Using CMIP6 Simulations Over SREX Regions. Earth Syst Environ 5 , 481–497 (2021). Stouffer, R. J., Manabe, S. & Bryan, K. Interhemispheric asymmetry in climate response to a gradual increase of atmospheric CO2. Nature 342 , 660–662 (1989). Vogel, M. M., Hauser, M. & Seneviratne, S. I. Projected changes in hot, dry and wet extreme events’ clusters in CMIP6 multi-model ensemble. Environ. Res. Lett. 15 , 094021 (2020). Bathiany, S., Dakos, V., Scheffer, M. & Lenton, T. M. Climate models predict increasing temperature variability in poor countries. Sci. Adv. 4 , eaar5809 (2018). Perkins, S. E. A review on the scientific understanding of heatwaves—Their measurement, driving mechanisms, and changes at the global scale. Atmospheric Research 164–165 , 242–267 (2015). Berg, A. et al. Impact of Soil Moisture–Atmosphere Interactions on Surface Temperature Distribution. Journal of Climate 27 , 7976–7993 (2014). Seneviratne, S. I., Lüthi, D., Litschi, M. & Schär, C. Land–atmosphere coupling and climate change in Europe. Nature 443 , 205–209 (2006). Lavell, A. et al. Climate Change: New Dimensions in Disaster Risk, Exposure, Vulnerability, and Resilience. In: Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation[Field, C.B., V. Barros, T.F. Stocker, D. Qin, D.J. Dokken, K.L. Ebi, M.D. Mastrandrea, K.J. Mach, G.-K. Plattner, S.K. Allen, M. Tignor, and P.M. Midgley (Eds.)]. A Special Report of Working Groups I and II of the Intergovernmental Panel on Climate Change (IPCC). (Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2012). Ruff, T. W. & Neelin, J. D. Long tails in regional surface temperature probability distributions with implications for extremes under global warming. Geophysical Research Letters 39 , 2011GL050610 (2012). Schär, C. et al. The role of increasing temperature variability in European summer heatwaves. Nature 427 , 332–336 (2004). Mueller, B. & Seneviratne, S. I. Systematic land climate and evapotranspiration biases in CMIP5 simulations. Geophysical Research Letters 41 , 128–134 (2014). Chang, P. et al. An Unprecedented Set of High‐Resolution Earth System Simulations for Understanding Multiscale Interactions in Climate Variability and Change. J Adv Model Earth Syst 12 , (2020). Haarsma, R. J. et al. High Resolution Model Intercomparison Project (HighResMIP v1.0) for CMIP6. Geosci. Model Dev. 9 , 4185–4208 (2016). Fischer, E. M. & Schär, C. Future changes in daily summer temperature variability: driving processes and role for temperature extremes. Clim Dyn 33 , 917–935 (2009). Chattopadhyay, A., Nabizadeh, E. & Hassanzadeh, P. Analog Forecasting of Extreme‐Causing Weather Patterns Using Deep Learning. J Adv Model Earth Syst 12 , (2020). Additional Declarations There is NO Competing Interest. Supplementary Files SIsubmit.pdf Article Supplementary Cite Share Download PDF Status: Published Journal Publication published 26 Nov, 2025 Read the published version in Communications Earth & Environment → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5393056","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Physical Sciences - Article","associatedPublications":[],"authors":[{"id":395049159,"identity":"459344c6-b4a8-4015-a6b3-50e107faeb59","order_by":0,"name":"Xiaohui Ma","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA0ElEQVRIiWNgGAWjYJCCAyDExt4A5jA2EK+F5wAJWiC6JBKI1KLbfsbwwI+KO4l9ko+PSfxgsJHdcID52QN8WszOpCUc7DnzLLFNOi1NsochzXjDATZzA7xaDiQfOMDbdhioJcdMmoHhcOKGAzxsEni1nH/YcPAvSIvkGZCW/0RouZF84DDYFgkekJYDxGh5lnBY5sxh4zaetGTLHoNk45mH2cwIOCzH+OObisOy89sPH7zxo8JOtu948zO8WtAAKKiYSVA/CkbBKBgFowA7AADpp08mWnPvlwAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0001-9937-3859","institution":"Ocean University of China","correspondingAuthor":true,"prefix":"","firstName":"Xiaohui","middleName":"","lastName":"Ma","suffix":""},{"id":395049160,"identity":"1fbc5c2e-613a-409a-9277-a85ed3642bd9","order_by":1,"name":"Zhili Tang","email":"","orcid":"","institution":"Ocean University of China","correspondingAuthor":false,"prefix":"","firstName":"Zhili","middleName":"","lastName":"Tang","suffix":""},{"id":395049161,"identity":"40cb60f3-4a68-4d34-a0ed-8005ab020094","order_by":2,"name":"Shenghui Zhou","email":"","orcid":"","institution":"Laoshan Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Shenghui","middleName":"","lastName":"Zhou","suffix":""},{"id":395049162,"identity":"3e663357-2f59-49b5-bfa9-18f2538194b3","order_by":3,"name":"Lixin Wu","email":"","orcid":"https://orcid.org/0000-0002-4694-5531","institution":"Ocean University of China","correspondingAuthor":false,"prefix":"","firstName":"Lixin","middleName":"","lastName":"Wu","suffix":""},{"id":395049163,"identity":"60e5d4c7-51da-4662-b28f-c1429bc43234","order_by":4,"name":"Wenju Cai","email":"","orcid":"https://orcid.org/0000-0001-6520-0829","institution":"Center for Southern Hemisphere Oceans Research","correspondingAuthor":false,"prefix":"","firstName":"Wenju","middleName":"","lastName":"Cai","suffix":""},{"id":395049164,"identity":"53cf1afe-a304-45d3-adc1-cb64cd364187","order_by":5,"name":"Zhao Jing","email":"","orcid":"https://orcid.org/0000-0002-8430-9149","institution":"Ocean University of China","correspondingAuthor":false,"prefix":"","firstName":"Zhao","middleName":"","lastName":"Jing","suffix":""},{"id":395049165,"identity":"f772fcc3-c033-408e-96b1-2cc90c167780","order_by":6,"name":"Zhaohui Chen","email":"","orcid":"https://orcid.org/0000-0002-0830-2332","institution":"Ocean University of China","correspondingAuthor":false,"prefix":"","firstName":"Zhaohui","middleName":"","lastName":"Chen","suffix":""},{"id":395049166,"identity":"a4855247-4bf1-4d1f-97c0-0fe9dea75999","order_by":7,"name":"Bolan Gan","email":"","orcid":"https://orcid.org/0000-0001-7620-485X","institution":"Ocean University of China","correspondingAuthor":false,"prefix":"","firstName":"Bolan","middleName":"","lastName":"Gan","suffix":""}],"badges":[],"createdAt":"2024-11-05 07:25:07","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5393056/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5393056/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s43247-025-02929-3","type":"published","date":"2025-11-26T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":76024141,"identity":"9bc59ea4-d89b-4eea-ab97-6d59f2bd909c","added_by":"auto","created_at":"2025-02-11 14:11:19","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":568255,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eResponses of hot extremes to anthropogenic warming in the high-resolution Community Earth System Model (HR-CESM).\u003c/strong\u003e Differences in summer season (JJA in the Northern Hemisphere and DJF in the Southern Hemisphere) averages of total duration (days/summer, \u003cstrong\u003ea\u003c/strong\u003e), frequency (events/summer, \u003cstrong\u003eb\u003c/strong\u003e), intensity (°C, \u003cstrong\u003ec\u003c/strong\u003e) and cumulative heat (°C/summer, \u003cstrong\u003ed\u003c/strong\u003e) for hot extreme events between historical (1981-2000) and future simulations (2081-2100) in HR-CESM. Differences above the 95% confidence level based on a two-sided Student’s test are shaded by gray dots. (\u003cstrong\u003ee\u003c/strong\u003e) Global averages (70°S-70°N, areas without hot extremes are assigned a value of 0) of hot extreme metrics in historical (blue bars) and future simulations (red bars), with corresponding values labeled above the bars. The red numbers at the top indicate the fold increases between future and historical values. (\u003cstrong\u003ef\u003c/strong\u003e) Time series of cumulative heat averaged for the Northern Hemisphere (15°N-70°N, red line), Southern Hemisphere (15°S-70°S, blue line), along with summer surface temperature changes relative to 1979-1984 (5-year moving average) for the Northern Hemisphere (JJA, 15°N-70°N, green line) and Southern Hemisphere (DJF, 15°S-70°S, orange line). Shaded regions in (\u003cstrong\u003ef\u003c/strong\u003e) outline the historical and future periods analyzed, with mean summer temperatures for both hemispheres labeled for the respective periods.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-5393056/v1/67fffbf5946744a2a069a42a.png"},{"id":76024437,"identity":"38c40c9b-ff07-439f-89ba-0671026be39f","added_by":"auto","created_at":"2025-02-11 14:19:25","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":544379,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRole of temperature variance in driving the spatial heterogeneity of future hot extremes.\u003c/strong\u003eSummer season averaged temperature variance in historical simulations (1981-2000, \u003cstrong\u003ea\u003c/strong\u003e) and the difference of that between future (2081-2100) and historical simulations (1981-2000) in the high-resolution Community Earth System Model (HR-CESM) (\u003cstrong\u003eb\u003c/strong\u003e). Regions with a negative correlation between hot extreme change and historical temperature variance are shaded by plus markers in (\u003cstrong\u003ea\u003c/strong\u003e). (\u003cstrong\u003ec\u003c/strong\u003e) Scatter plot between anomalous hot extreme total duration changes (future minus historical) and anomalous historical temperature variance (historical) in the regions highlighted by plus markers in (\u003cstrong\u003ea\u003c/strong\u003e). (\u003cstrong\u003ed\u003c/strong\u003e) Scatter plot between anomalous hot extreme total duration changes (future minus historical) and anomalous temperature variance changes (future minus historical) in the regions highlighted by plus markers in (\u003cstrong\u003ea\u003c/strong\u003e). The anomalous values plotted in (\u003cstrong\u003ec\u003c/strong\u003e) and (\u003cstrong\u003ed\u003c/strong\u003e) are computed as the absolute values minus the global mean. (\u003cstrong\u003ee\u003c/strong\u003e) Changes in T90p (the proportion of temperatures exceeding the 90th percentile) relative to different standard deviation levels (from 1σ to\u003cem\u003e \u003c/em\u003e4σ) and mean temperature shifts based on idealized Gaussian distributions. (\u003cstrong\u003ef\u003c/strong\u003e) Temperature probability distribution functions (PDFs) for high-response regions identified by K-means clustering in historical (blue solid line) and future (red solid line) simulations in HR-CESM. The red dashed line represents the shifted PDF corresponding to the mean temperature change. Bar plots show region-averaged total duration and temperature variance for historical (blue) and future (red) periods, with corresponding values and fold increases labeled above the bars. (\u003cstrong\u003eg\u003c/strong\u003e) Same as (\u003cstrong\u003ef\u003c/strong\u003e) but for low-response regions.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-5393056/v1/a2d70ab3f877974387dde5d6.png"},{"id":76024435,"identity":"75c5914a-043e-45d2-a3b0-69251a09c51c","added_by":"auto","created_at":"2025-02-11 14:19:19","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":681875,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePhysical processes contributing to the spatial heterogeneity of baseline temperature variability.\u003c/strong\u003e Summer season averaged variance of latent heat flux (W\u003csup\u003e2\u003c/sup\u003e/m\u003csup\u003e4\u003c/sup\u003e\u003cstrong\u003e, a\u003c/strong\u003e), sensible heat flux (W\u003csup\u003e2\u003c/sup\u003e/m\u003csup\u003e4\u003c/sup\u003e\u003cstrong\u003e, b\u003c/strong\u003e), downwelling radiation (W\u003csup\u003e2\u003c/sup\u003e/m\u003csup\u003e4\u003c/sup\u003e\u003cstrong\u003e, c\u003c/strong\u003e), and soil moisture (g\u003csup\u003e2\u003c/sup\u003e/m\u003csup\u003e4\u003c/sup\u003e\u003cstrong\u003e, d\u003c/strong\u003e) in historical simulations (1981-2000) in the high-resolution Community Earth System Model (HR-CESM). (\u003cstrong\u003ee\u003c/strong\u003e) Scatter plot of historical temperature variance in high- and low-response regions identified by K-means clustering (colored dots) as a function of both latent heat flux variance (horizontal axis) and downwelling radiation variance (vertical axis). (\u003cstrong\u003ef\u003c/strong\u003e) and (\u003cstrong\u003eg\u003c/strong\u003e), same as \u003cstrong\u003e(e\u003c/strong\u003e) but with sensible heat flux and soil moisture variance on the vertical axis. Note that latent heat flux and soil moisture variance in (\u003cstrong\u003eg\u003c/strong\u003e) are normalized to eliminate scale differences between the two variables.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-5393056/v1/256d7e3075be77064f4260af.png"},{"id":76024145,"identity":"d655e02f-657f-4ee0-922f-ac0aa3ccd4e3","added_by":"auto","created_at":"2025-02-11 14:11:19","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":421563,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePersistence in the spatial distribution of land-atmosphere coupling.\u003c/strong\u003e Summer season averaged land-atmosphere coupling, defined as the correlation between detrended surface temperature and evaporation, in historical (1981-2000,\u003cstrong\u003e a\u003c/strong\u003e) and future simulations (2081-2100, \u003cstrong\u003eb\u003c/strong\u003e) in the high-resolution Community Earth System Model (HR-CESM), where negative (positive) correlations represent active (inactive) coupling. (\u003cstrong\u003ec\u003c/strong\u003e) Coupling sign changes between future and historical simulations, with unchanged signs in color and reversed signs marked by ‘x’. (\u003cstrong\u003ed\u003c/strong\u003e) Scatter plot between historical and future land-atmosphere coupling in HR-CESM, with 94.8% of global areas showing persistent signs and a correlation coefficient of 0.86 above the 99% confidence level. (\u003cstrong\u003ee\u003c/strong\u003e) Scatter plot for Coupled Model Intercomparison Project Phase 6 (CMIP6) simulations, with 99.3% of global areas showing persistent signs and a coefficient of 0.97 above the 99% confidence level.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-5393056/v1/433c86e5535a835ac313ec3b.png"},{"id":96885675,"identity":"c06cce2b-b454-4287-8452-19a32b9842dd","added_by":"auto","created_at":"2025-11-27 08:11:04","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2935474,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5393056/v1/71852b35-f745-4d5b-af63-0d7e4d37f69b.pdf"},{"id":76024144,"identity":"81b46430-f709-4aa8-aa58-39a015c222af","added_by":"auto","created_at":"2025-02-11 14:11:19","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1946636,"visible":true,"origin":"","legend":"Article Supplementary","description":"","filename":"SIsubmit.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5393056/v1/e6b35e842bde0e2cf8a42fa8.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Baseline Temperature Variability Shaping Geographical Distribution of Future Hot Extremes under Anthropogenic Warming","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eHot extreme events, characterized by prolonged periods of excessively high temperatures exceeding the average, have devastating impacts on human health, agriculture, and the natural environment\u003csup\u003e1\u0026ndash;4\u003c/sup\u003e. One of the most severe events, the 2003 European heatwave, led to approximately 70,000 deaths and more than \u0026euro;15\u0026nbsp;billion in economic loss\u003csup\u003e5,6\u003c/sup\u003e. Similarly, the 2010 Russian heatwave resulted in over 55,000 deaths and a 25% reduction in crop production\u003csup\u003e7\u003c/sup\u003e. Under climate change, these events are expected to increase in frequency, intensity and duration\u003csup\u003e8\u003c/sup\u003e. Importantly, the geographical response of hot extremes to anthropogenic warming is highly heterogeneous, with some regions experiencing far more intense extremes than the global average, leading to disproportionately socioeconomic impacts\u003csup\u003e9\u0026ndash;11\u003c/sup\u003e. Understanding the drivers behind the spatial heterogeneity is essential for systematically improving regional climate projections for adaptation strategies.\u003c/p\u003e \u003cp\u003eWhile there is broad consensus regarding the increasing occurrence, magnitude of hot extremes under anthropogenic warming\u003csup\u003e12\u0026ndash;15\u003c/sup\u003e, the physical processes driving their spatial heterogeneity are regionally dependent and mechanistically diverse\u003csup\u003e16\u003c/sup\u003e. One key driver widely discussed is a soil moisture-temperature feedback, where decreasing soil moisture leads to rising mean temperature due to reduced evaporation, and vice versa\u003csup\u003e10,17,18\u003c/sup\u003e. The feedback plays a crucial role in amplifying regional hot extremes, particularly in midlatitude regions. In high-latitude regions, snow/ice-temperature feedback modifies surface albedo, further affecting warming\u003csup\u003e19\u003c/sup\u003e. Additionally, atmospheric dynamical processes, such as blocking events and large-scale circulation patterns, significantly contribute to the regional variability of hot extremes. For example, intense European heatwaves have been linked to weakened atmospheric circulation and increased atmospheric blocking\u003csup\u003e20\u0026ndash;22\u003c/sup\u003e. Sea surface temperatures (SSTs) may also impact adjacent land hot extremes by altering atmospheric circulation and heat advection\u003csup\u003e23\u0026ndash;25\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThese drivers are intricately connected\u0026mdash;soil moisture can modulate atmospheric circulation, exacerbating hot extremes\u003csup\u003e26\u003c/sup\u003e; while changes in atmospheric circulation can feedback onto soil moisture and SST\u003csup\u003e27\u003c/sup\u003e, creating a complex interaction system. Identifying the dominant factors and unraveling how these drivers contribute to the geographical distribution of projected hot extremes remain a challenge. Moreover, large uncertainties in the future evolution of soil moisture, snow/ice and atmospheric circulation in current climate models\u003csup\u003e8,28\u0026ndash;30\u003c/sup\u003e add difficulty to accurately projecting how hot extremes will manifest regionally under anthropogenic warming.\u003c/p\u003e \u003cp\u003eThe issue of whether there is a systematic predictive indicator for the projection of regional variability in future hot extremes is critical yet unresolved. In this study, we analyze outputs from an unprecedented set of eddy-resolving, high-resolution simulations from Community Earth System Model (HR-CESM) with ~\u0026thinsp;0.25\u0026deg; atmospheric and ~\u0026thinsp;0.1\u0026deg; oceanic resolution and 10 climate models in Coupled Model Intercomparison Project Phase 6 (CMIP6, see \u0026ldquo;High-Resolution CESM and CMIP6 simulations\u0026rdquo; in Methods; Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e), to investigate the dominant factors and associated physical processes driving the spatial heterogeneity of future hot extremes in warming climate.\u003c/p\u003e"},{"header":"2. Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Simulated and Projected Hot Extremes in HR-CESM and CMIP6\u003c/h2\u003e \u003cp\u003eWe first evaluate the models\u0026rsquo; ability to simulate the observed characteristics of hot extreme events, including occurrence frequency, intensity, total duration, and cumulative heat (a hot extreme event is defined as at least three consecutive days with daily temperatures exceeding the 90th percentile local threshold during the summer season, see \u0026ldquo;Definitions of Hot Extremes and Related Metrics\u0026rdquo; in Methods for detailed definitions). The simulations are compared against ERA5 reanalysis data (the fifth-generation reanalysis from the European Centre for Medium-Range Weather Forecasts\u003csup\u003e31\u003c/sup\u003e) for the period 1979 to 2019. Both HR-CESM and CMIP6 faithfully reproduce the spatial distribution of hot extreme total duration (our primary measure for hot extreme events, see \u0026ldquo;Definitions of Hot Extremes and Related Metrics\u0026rdquo; in Methods), with only slight differences in magnitude (Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003ea-c). The same holds true for the occurrence frequency, intensity and cumulative heat distributions as well. In addition, we evaluate the models\u0026rsquo; ability to represent observed trends in hot extremes over the historical period (Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003ed-f). While both HR-CESM and CMIP6 capture regions with significant increases in hot extreme duration, they tend to slightly overestimate the trends. Overall, the models show high fidelity in simulating key characteristics of hot extremes. However, we note that increasing the resolution to eddy-resolving scales in HR-CESM provides limited improvement in their climatological simulation compared with CMIP6, a finding consistent with previous research\u003csup\u003e32\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eWe next examine the projected changes of hot extremes in HR-CESM under the high emission scenario of Representative Concentration Pathway 8.5 (RCP8.5). By the end of the 21st century (2081\u0026ndash;2100), substantial global increases in the total duration, occurrence frequency, intensity, and cumulative heat of hot extremes are projected compared to the historical period (1981\u0026ndash;2000, Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea-d). The global averaged total duration of hot extremes is expected to rise from 2.57 days per summer to 66.46 days per summer, a more than 25-fold increase (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ee). The increase in duration, together with a higher intensity, drives a substantial rise in cumulative heat, with future levels projected to exceed current values by over 70 times (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ee). In addition, there is a near linear-relationship between the rise in hot extremes and summer mean temperature increase from 1979 to 2100, along with a hemispheric asymmetry in the response (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ef). The more pronounced increase in hot extremes in the Northern Hemisphere is attributed to enhanced warming, driven by the greater land coverage in this region. These projected features align with previous findings\u003csup\u003e33\u0026ndash;35\u003c/sup\u003e, confirming the robustness of hot extreme projections under warming conditions by HR-CESM.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe projected changes in hot extreme events exhibit pronounced spatial heterogeneity (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea-d). Regions such as northwestern and eastern coastal North America, the Amazon region in northern South America, southern Europe, the Middle East, northern and central Africa, and South Asia experience larger increases in total duration (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). In contrast, regions like the central U.S., northeastern Canada, southern South America, northern Eurasia, the Sahel, and Australia show a suppressed rise. The contrast in regional variability aligns closely with changes in intensity and cumulative heat, while the frequency of events shows an opposite trend. The relatively smaller increase in frequency in regions with prolonged duration response is attributed to more persistent hot extremes in a warming climate.\u003c/p\u003e \u003cp\u003eWe verify the hot extreme projections using CMIP6, across different warming periods and thresholds. The projected rising trend and hemispheric asymmetry for 2031\u0026ndash;2050 are evident in CMIP6, albeit with a lower amplitude (Fig. S2 a-f). By mid-century, the global averaged hot extreme duration is projected to be approximately 11 times the current value (Fig. S2e), representing 40% of the increase expected in HR-CESM by 2100. Importantly, the spatial variability of projected hot extremes in CMIP6 closely mirrors that of HR-CESM (Fig. S2a-d), displaying similar hot and cold spot patterns, though the increase is more subdued in southern Europe.\u003c/p\u003e \u003cp\u003eThe above projections are based on a fixed threshold referenced to the historical period (T90\u003csub\u003ehist\u003c/sub\u003e). We also validate the results using a varying threshold referenced to future warming periods (T90\u003csub\u003efuture\u003c/sub\u003e, see \u0026ldquo;Definitions of Hot Extremes and Related Metrics\u0026rdquo; in Methods). Compared to the fixed threshold, the varying threshold results in an overall shorter but still statistically significant increase in hot extreme total duration (Fig. S3a). The spatial distribution of hot extreme increase remains similar, except for a less pronounced response in South Asia (Fig. S3a). To ensure consistent identifications of hot and cold spots across HR-CESM, CMIP6, and different thresholds, southern Europe and South Asia are excluded from the hot spot regions in the subsequent analysis when comparing temperature probability density functions (PDFs) between these regions. The exclusion is also supported by the K-means clustering results (see \u0026ldquo;K-means Clustering Analysis\u0026rdquo; in Methods) as discussed later.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Drivers of Spatial Heterogeneity in Future Hot Extremes\u003c/h2\u003e \u003cp\u003eChanges in temperature mean and variability are widely recognized as key factors influencing projections of hot extremes under warming\u003csup\u003e10,36,37\u003c/sup\u003e. However, the spatial distribution of projected hot extremes and mean temperature changes show little similarity, with only a weak and insignificant positive correlation under the RCP8.5 scenario in HR-CESM (Fig. S4a,b). Similarly, no clear correlation is found between the spatial distribution of projected hot extremes and temperature variability changes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb, d). Instead, 83% of projected changes in global hot extremes are negatively correlated with historical temperature variability (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). A linear regression shows a significant negative correlation, with a coefficient of -0.73 above the 99% confidence level (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec). This negative correlation also holds across over 71% of the global area when applying a varying threshold (Fig. S3). These findings are further validated in CMIP6 (Fig. S4c,d, S5), revealing a consistent negative correlation between the increase in hot extremes and baseline temperature variability across 93% of the global area. Together, these results suggest the robust role of baseline temperature variability in shaping the future geographical distribution of hot extremes.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo better assess the relative contributions of temperature mean and variability in driving regional differences in hot extreme responses, we analyzed temperature PDFs for both high- and low-response regions. To minimize uncertainties in region selection, we apply the K-means clustering approach (see \u0026ldquo;K-means Clustering Analysis\u0026rdquo; in Methods), which produces high- and low-response regions that closely match the hot and cold spots consistently identified across HR-CESM, CMIP6, and various threshold definitions (Fig. S6a), confirming the robustness of the classification.\u003c/p\u003e \u003cp\u003eThe temperature PDFs of these two groups are distinct, with high-response regions exhibiting much narrower distribution and lower historical temperature variability than low-response regions (blue lines in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef, g). Specifically, the standard deviation of historical temperatures (the square root of variance) in high-response regions is only one-quarter that of the low-response regions (blue bars in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef, g). However, the projected total duration of future hot extremes in high-response regions is approximately twice that of low-response regions (red bars in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef, g). Under warming conditions, both temperature PDFs shift rightward, leading to a significant increase in the proportion of hot extreme days (red lines in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef, g). Projections show that T90p, representing the proportion of temperatures exceeding the 90th percentile, will rise to 79.8% in high-response regions in the future, compared to 45.3% in low-response regions. Interestingly, the mean temperature shifts are comparable in both regions, with a slightly greater warming observed in low-response regions (5.6\u0026deg;C increase in high-response regions and 5.9\u0026deg;C in low-response regions). This suggests that the greater increase in hot extremes in high-response regions is not attributed to differences in mean temperature shifts.\u003c/p\u003e \u003cp\u003eThe impact of baseline temperature variability on the increase in T90p is further quantified using idealized Gaussian temperature distributions (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee). For the same 6\u0026deg;C warming, T90p for temperature profiles with low baseline variability (one standard deviation) is expected to rise to nearly 100%, while for profiles with high baseline variability (four standard deviations), T90p increases to only 41.3%. The difference in response magnitude aligns with model projections, confirming that baseline temperature variability is the primary driver of the regional differences in future hot extreme intensification.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Physical Processes Influencing Baseline Temperature Variability\u003c/h2\u003e \u003cp\u003eTo identify the physical processes contributing to the spatial variability of baseline temperature variability, we examine latent heat flux (LH), sensible heat flux (SH), and downwelling radiation (RD), all of which influence the surface energy balance that drives land surface temperature and, consequently, temperature variability\u003csup\u003e10,36\u003c/sup\u003e. Spatial comparison shows that across most of the globe, baseline temperature variance closely aligns with LH variance, except in the northern high latitudes (north of 50 \u0026deg;N), where it tends to be more influenced by SH and RD (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea-c). After excluding this region, scatter plots clearly reveal that the temperature variance is primarily driven by LH variance, with a smaller contribution from SH and a minimal impact from RD (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee, f).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eSoil moisture has been widely reported to influence the partition between LH and SH through land-atmospheric coupling\u003csup\u003e17,27,38\u003c/sup\u003e, emphasizing its potential role in driving temperature variability. This is corroborated by the strong spatial resemblance between temperature and soil moisture variance (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed). A closer examination of the scatter plots reveals two distinct regimes (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eg). In the low-temperature variance regime, the scatter cloud is nearly vertical, indicating the predominant role of soil moisture. In the high-temperature variance regime, the scatter cloud aligns at a 45\u0026deg; angle, suggesting joint impacts from soil moisture and LH. The difference is likely linked to local dry or wet conditions, which affect the sensitivity of evapotranspiration responses to soil moisture, as suggested by previous studies\u003csup\u003e27,38\u003c/sup\u003e. These findings, consistent with prior research, support the critical role of soil moisture in modulating temperature variability.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Century-Scale Persistence in Land-Atmosphere Coupling\u003c/h2\u003e \u003cp\u003eThe role of baseline temperature variability in driving the spatial distribution of future hot extremes over the next century raises an important question: what is the key driver behind this \"century memory\"? The above analysis reveals a strong link between soil moisture and temperature variability, primarily operating through a negative soil moisture-evaporation-temperature feedback according to previous studies\u003csup\u003e27\u003c/sup\u003e. This feedback is often referred to as land-atmosphere coupling and can be effectively captured by the correlation between the detrended evaporation and surface temperature\u003csup\u003e39\u003c/sup\u003e. Regions characterized by both low hot extreme response and high baseline temperature variability coincide with areas of active land-atmosphere coupling (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). Notably, the coupling remains remarkably stable across historical and future simulations (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea, b), with nearly 95% of global areas showing persistent patterns (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec). The spatial match of active and inactive land-atmosphere coupling areas between past and future simulations shows a high correlation of 0.86 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed), underscoring the persistence of this distribution over a century timescale. This enduring pattern is also corroborated by multiple CMIP6 models, indicating that it is not model-specific (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee). The results suggest that the spatial distribution of baseline temperature variability is intrinsically tied to land-atmosphere coupling, which operates over a long time and anchors the spatial heterogeneity of future hot extremes.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"3. Discussion","content":"\u003cp\u003eUsing eddy-resolving HR-CESM simulations and multiple model results from CMIP6, we identify baseline temperature variability as a key factor shaping the uneven distribution of future hot extremes, a result validated across various climate models, warming levels, and thresholds. The baseline temperature variability is anchored by persistent land-atmosphere coupling, which operates over century timescales and sustains the spatial heterogeneity of future hot extremes. Moreover, the coupling primarily influences the temperature variability through soil moisture and latent heat flux, consistent with the current understanding of soil moisture-temperature feedback dynamics, and further supporting the robustness of the results.\u003c/p\u003e \u003cp\u003eExisting studies have emphasized the role of changes in temperature mean and variability in determining hot extreme projections under warming\u003csup\u003e10,36,37,40\u0026ndash;42\u003c/sup\u003e, underscoring the need for accurate representations of these factors for both historical and future periods to ensure reliable estimates of future hot extremes. However, the task remains challenging due to substantial uncertainties in projected changes in temperature mean and variability in future periods. Our findings show that regional variations in projected hot extremes are only weakly connected to changes in temperature mean or variability under warming. Instead, baseline temperature variability primarily determines the global pattern of hot extreme projections, highlighting it as a potential predictor for assessing regional variations in future hot extreme responses. These findings provide a simplified framework for assessing the geographic distribution of future hot extremes based on historical state without integrating into the future, offering valuable insights for developing more targeted adaptation strategies for regions most vulnerable to extreme heat under climate change. However, the results also stress the importance of more precise and comprehensive observations to quantify historical temperature variability, as well as improved climate models\u0026rsquo; capability of accurately simulating this variability, to enhance projections of regional variations in future hot extremes. The findings may be subject to uncertainties due to their reliance on the century-scale persistence of land-atmosphere coupling, while accurately representing and projecting the coupling remains a major challenge for current climate models\u003csup\u003e16,28,43\u003c/sup\u003e.\u003c/p\u003e "},{"header":"Methods","content":"\u003cp\u003e \u003cb\u003eHigh-Resolution CESM and CMIP6 Simulations\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThe eddy-resolving high-resolution Community Earth System Model (HR-CESM) simulations with ~\u0026thinsp;0.25\u0026deg; atmosphere and ~\u0026thinsp;0.1\u0026deg; ocean components developed by the National Center for Atmosphere Research (NCAR) are analyzed\u003csup\u003e44\u003c/sup\u003e. The simulations consist of a 500-year preindustrial control run and a 250-year simulation covering historical (1850\u0026ndash;2005) and future (2006\u0026ndash;2100) climates. Historical forcing is applied from 1850 to 2005, followed by the RCP8.5 scenario (a high greenhouse gas concentration pathway) forcing from 2006 onwards. Daily outputs are used to assess hot extreme events and related variables, comparing two periods: a historical period from 1981 to 2000, and a future period from 2081 to 2100.\u003c/p\u003e \u003cp\u003eTen climate model simulations from the Coupled Model Intercomparison Project Phase 6 (CMIP6) are examined, including five from the High-Resolution Model Intercomparison Project (HighResMIP\u003csup\u003e45\u003c/sup\u003e) and five pairing low-resolution simulations. The oceanic resolution of these models ranges from 0.1\u0026deg; to 1\u0026deg;, while atmospheric resolution ranges from 0.25\u0026deg; to 1\u0026deg;. The selected models and their specific resolutions are detailed in Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e. Each model includes a 100-year historical and future (under RCP8.5 scenario) simulation from 1950\u0026ndash;2050. Hot extreme events during the period overlapping with ERA5 and HR-CESM are analyzed and compared.\u003c/p\u003e \u003cp\u003e \u003cb\u003eDefinition of Hot Extremes and Related Metrics\u003c/b\u003e \u003c/p\u003e \u003cp\u003eFollowing previous studies\u003csup\u003e14,46\u003c/sup\u003e, a hot extreme event is defined as at least three consecutive days with daily temperatures exceeding the 90th percentile threshold during summer (JJA in the Northern Hemisphere and DJF in the Southern Hemisphere). The 90th percentile is calculated for each calendar day at each grid point using a 15-day moving window centered around the targeted day over the reference periods\u003csup\u003e46\u003c/sup\u003e. Two thresholds are computed: one reference to the historical period (1979\u0026ndash;2019, T90\u003csub\u003ehist\u003c/sub\u003e) and another to future warming periods (2060\u0026ndash;2100, T90\u003csub\u003efuture\u003c/sub\u003e). Correspondingly, hot extreme projections under anthropogenic warming are evaluated in two ways. For the fixed threshold, hot extremes in both historical and future periods are identified using T90\u003csub\u003ehist\u003c/sub\u003e. For the varying threshold, hot extremes are identified using T90\u003csub\u003ehist\u003c/sub\u003e for the historical and T90\u003csub\u003efuture\u003c/sub\u003e for the future period. Projected changes in hot extremes using the fixed threshold account for contribution from mean temperature increases between the historical and future periods, while the varying threshold excludes this effect.\u003c/p\u003e \u003cp\u003eThe occurrence frequency is defined as the number of hot extreme events per summer. The intensity of each event is measured by the maximum temperature anomaly above the 90th percentile threshold during the event and then averaged over the summer season. Total duration refers to the cumulative number of days during which hot extreme events occur per summer. Cumulative heat is the sum of temperature anomalies exceeding the 90th percentile threshold across all hot extreme events\u003csup\u003e14\u003c/sup\u003e. Notably, total duration is an effective indicator of projected hot extreme changes, capturing both occurrence frequency and mean duration response (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). It is also the dominant contributor to cumulative heat increase (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed), a finding consistent with prior research\u003csup\u003e14\u003c/sup\u003e. Therefore, we use total duration as the primary measure of hot extreme responses in our analyses.\u003c/p\u003e \u003cp\u003e \u003cb\u003eK-means Clustering Analysis\u003c/b\u003e \u003c/p\u003e \u003cp\u003eK-means clustering is applied to identify regions with high or low hot extreme responses based on projected changes in total duration and historical temperature variance\u003csup\u003e47\u003c/sup\u003e. Both variables are normalized (ranging from \u0026minus;\u0026thinsp;2.5 to 2.5 standard deviation) and divided into five categories, labeled from \u0026minus;\u0026thinsp;2 to 2, with each category spanning one standard deviation. This ensures comparability between the variables and prevents one from dominating the clustering due to scale differences. The K-means algorithm is then applied to the processed data and repeated 1,000 times with random initial cluster centroids, yielding an optimized clustering of 5. The two clusters with the largest and opposite variations in hot extreme changes and temperature variance are selected, which gives the high- and low-response regions shown in Fig. S6. As illustrated in the spatial map of hot extreme change and temperature variance (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea \u003cb\u003eand\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea), high-response regions correspond to areas with low historical temperature variance, and vice versa. These regions also closely match the hot and cold spots consistently identified across HR-CESM, CMIP6, and different threshold definitions, confirming the robustness of the classification. Note that in all identified regions, historical temperature variance is significantly lower in high-response regions compared to low-response regions. However, due to differing geographical locations, the temperature PDFs in these regions exhibit distinct mean temperatures, which can obscure the overall PDF shape when combined. To enhance clarity, the temperature PDFs for North Africa and Northern Eurasia (regions with distinct mean temperatures) are plotted separately (Fig. S6b, c), while the remaining regions are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef, g.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eERA5 reanalysis can be downloaded from https://doi.org/10.24381/cds.bd0915c6. The CESM simulations can be achieved through https://ihesp.github.io/archive/products/ds_archive/Sunway_Runs.html. The CMIP6 data can be downloaded from https://pcmdi.llnl.gov/CMIP6/.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCode Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePython and Matlab codes to reproduce the analyses are available upon request from the corresponding author.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research is supported by the National Natural Science Foundation of China (42376025), Science and Technology Innovation Program of Laoshan Laboratory (LSKJ202300302, LSKJ202202503), Shandong Provincial Natural Science Foundation (ZR2022YQ29), Taishan Scholar Funds (tsqn202103028). We thank Laoshan Laboratory in Qingdao and the National Supercomputing Center in Jinan for providing the high resolution CESM simulations and high performance computing resources that contributed to the research results reported in this paper.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eZ. T. performed most of the analyses under X. M.\u0026rsquo;s instruction. S. Z. detected the hot extremes and assisted with data processing. X. M. conceived the central idea, \u0026nbsp;designed the study and wrote the manuscript. L. W. supervised the project. W. C. \u0026nbsp;contributed to discussions on physical mechanisms influencing temperature variability. Z. J., Z. C. and B. G.\u0026nbsp;contributed\u0026nbsp;to interpreting the results and improving the manuscript.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eLobell, D. B. \u0026amp; Field, C. B. Global scale climate\u0026ndash;crop yield relationships and the impacts of recent warming. \u003cem\u003eEnviron. Res. Lett.\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 014002 (2007).\u003c/li\u003e\n\u003cli\u003eCampbell, S., Remenyi, T. A., White, C. J. \u0026amp; Johnston, F. H. Heatwave and health impact research: A global review. \u003cem\u003eHealth \u0026amp; Place\u003c/em\u003e \u003cstrong\u003e53\u003c/strong\u003e, 210\u0026ndash;218 (2018).\u003c/li\u003e\n\u003cli\u003eDomeisen, D. I. V. \u003cem\u003eet al.\u003c/em\u003e Prediction and projection of heatwaves. \u003cem\u003eNat Rev Earth Environ\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 36\u0026ndash;50 (2022).\u003c/li\u003e\n\u003cli\u003eYin, J. \u003cem\u003eet al.\u003c/em\u003e Future socio-ecosystem productivity threatened by compound drought\u0026ndash;heatwave events. \u003cem\u003eNat Sustain\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 259\u0026ndash;272 (2023).\u003c/li\u003e\n\u003cli\u003eRobine, J.-M. \u003cem\u003eet al.\u003c/em\u003e Death toll exceeded 70,000 in Europe during the summer of 2003. \u003cem\u003eComptes Rendus Biologies\u003c/em\u003e \u003cstrong\u003e331\u003c/strong\u003e, 171\u0026ndash;178 (2008).\u003c/li\u003e\n\u003cli\u003eGarc\u0026iacute;a-Le\u0026oacute;n, D. \u003cem\u003eet al.\u003c/em\u003e Current and projected regional economic impacts of heatwaves in Europe. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 5807 (2021).\u003c/li\u003e\n\u003cli\u003eWMO. \u003cem\u003eAtlas of Mortality and Economic Losses from Weather, Climate and Water Extremes\u003c/em\u003e. vol. 1123 (2014).\u003c/li\u003e\n\u003cli\u003eSeneviratne, S. I. \u003cem\u003eet al.\u003c/em\u003e Weather and Climate Extreme Events in a Changing Climate. \u003cem\u003eClimate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change\u003c/em\u003e 1513\u0026ndash;1766 (2021) doi:10.1017/9781009157896.013.\u003c/li\u003e\n\u003cli\u003eSeneviratne, S. I., Donat, M. G., Pitman, A. J., Knutti, R. \u0026amp; Wilby, R. L. Allowable CO2 emissions based on regional and impact-related climate targets. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e529\u003c/strong\u003e, 477\u0026ndash;483 (2016).\u003c/li\u003e\n\u003cli\u003eVogel, M. M. \u003cem\u003eet al.\u003c/em\u003e Regional amplification of projected changes in extreme temperatures strongly controlled by soil moisture‐temperature feedbacks. \u003cem\u003eGeophys. Res. Lett.\u003c/em\u003e \u003cstrong\u003e44\u003c/strong\u003e, 1511\u0026ndash;1519 (2017).\u003c/li\u003e\n\u003cli\u003eTurnau, R., Robinson, W. A., Lackmann, G. M. \u0026amp; Michaelis, A. C. Model Projections of Increased Severity of Heat Waves in Eastern Europe. \u003cem\u003eGeophysical Research Letters\u003c/em\u003e \u003cstrong\u003e49\u003c/strong\u003e, (2022).\u003c/li\u003e\n\u003cli\u003ePerkins, S. E. \u0026amp; Alexander, L. V. On the Measurement of Heat Waves. \u003cem\u003eJournal of Climate\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e, 4500\u0026ndash;4517 (2013).\u003c/li\u003e\n\u003cli\u003eRusso, S. \u003cem\u003eet al.\u003c/em\u003e Magnitude of extreme heat waves in present climate and their projection in a warming world. \u003cem\u003eJGR Atmospheres\u003c/em\u003e \u003cstrong\u003e119\u003c/strong\u003e, (2014).\u003c/li\u003e\n\u003cli\u003ePerkins-Kirkpatrick, S. E. \u0026amp; Lewis, S. C. Increasing trends in regional heatwaves. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 3357 (2020).\u003c/li\u003e\n\u003cli\u003eTeng, H., Leung, R., Branstator, G., Lu, J. \u0026amp; Ding, Q. Warming Pattern over the Northern Hemisphere Midlatitudes in Boreal Summer 1979\u0026ndash;2020. \u003cem\u003eJOURNAL OF CLIMATE\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, (2022).\u003c/li\u003e\n\u003cli\u003eDonat, M. G., Pitman, A. J. \u0026amp; Seneviratne, S. I. Regional warming of hot extremes accelerated by surface energy fluxes. \u003cem\u003eGeophysical Research Letters\u003c/em\u003e \u003cstrong\u003e44\u003c/strong\u003e, 7011\u0026ndash;7019 (2017).\u003c/li\u003e\n\u003cli\u003eSeneviratne, S. I. \u003cem\u003eet al.\u003c/em\u003e Impact of soil moisture‐climate feedbacks on CMIP5 projections: First results from the GLACE‐CMIP5 experiment. \u003cem\u003eGeophys. Res. Lett.\u003c/em\u003e \u003cstrong\u003e40\u003c/strong\u003e, 5212\u0026ndash;5217 (2013).\u003c/li\u003e\n\u003cli\u003eDouville, H., Colin, J., Krug, E., Cattiaux, J. \u0026amp; Thao, S. Midlatitude daily summer temperatures reshaped by soil moisture under climate change. \u003cem\u003eGeophysical Research Letters\u003c/em\u003e \u003cstrong\u003e43\u003c/strong\u003e, 812\u0026ndash;818 (2016).\u003c/li\u003e\n\u003cli\u003eDiro, G. T., Sushama, L. \u0026amp; Huziy, O. Snow-atmosphere coupling and its impact on temperature variability and extremes over North America. \u003cem\u003eClim Dyn\u003c/em\u003e \u003cstrong\u003e50\u003c/strong\u003e, 2993\u0026ndash;3007 (2018).\u003c/li\u003e\n\u003cli\u003eCoumou, D., Lehmann, J. \u0026amp; Beckmann, J. The weakening summer circulation in the Northern Hemisphere mid-latitudes. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e348\u003c/strong\u003e, 324\u0026ndash;327 (2015).\u003c/li\u003e\n\u003cli\u003eHorton, D. E. \u003cem\u003eet al.\u003c/em\u003e Contribution of changes in atmospheric circulation patterns to extreme temperature trends. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e522\u003c/strong\u003e, 465\u0026ndash;469 (2015).\u003c/li\u003e\n\u003cli\u003eRousi, E., Kornhuber, K., Beobide-Arsuaga, G., Luo, F. \u0026amp; Coumou, D. Accelerated western European heatwave trends linked to more-persistent double jets over Eurasia. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 3851 (2022).\u003c/li\u003e\n\u003cli\u003eArblaster, J. M. \u0026amp; Alexander, L. V. The impact of the El Ni\u0026ntilde;o‐Southern Oscillation on maximum temperature extremes. \u003cem\u003eGeophys. Res. Lett.\u003c/em\u003e \u003cstrong\u003e39\u003c/strong\u003e, 2012GL053409 (2012).\u003c/li\u003e\n\u003cli\u003eMiralles, D. G., Teuling, A. J., van Heerwaarden, C. C. \u0026amp; Vil\u0026agrave;-Guerau de Arellano, J. Mega-heatwave temperatures due to combined soil desiccation and atmospheric heat accumulation. \u003cem\u003eNature Geosci\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 345\u0026ndash;349 (2014).\u003c/li\u003e\n\u003cli\u003eLuo, M. \u0026amp; Lau, N.-C. Summer heat extremes in northern continents linked to developing ENSO events. \u003cem\u003eEnviron. Res. Lett.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 074042 (2020).\u003c/li\u003e\n\u003cli\u003eKoster, R. D., Chang, Y., Wang, H. \u0026amp; Schubert, S. D. Impacts of Local Soil Moisture Anomalies on the Atmospheric Circulation and on Remote Surface Meteorological Fields during Boreal Summer: A Comprehensive Analysis over North America. \u003cem\u003eJOURNAL OF CLIMATE\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, (2016).\u003c/li\u003e\n\u003cli\u003eSeneviratne, S. I. \u003cem\u003eet al.\u003c/em\u003e Investigating soil moisture\u0026ndash;climate interactions in a changing climate: A review. \u003cem\u003eEarth-Science Reviews\u003c/em\u003e \u003cstrong\u003e99\u003c/strong\u003e, 125\u0026ndash;161 (2010).\u003c/li\u003e\n\u003cli\u003eCheng, S., Huang, J., Ji, F. \u0026amp; Lin, L. Uncertainties of soil moisture in historical simulations and future projections. \u003cem\u003eJGR Atmospheres\u003c/em\u003e \u003cstrong\u003e122\u003c/strong\u003e, 2239\u0026ndash;2253 (2017).\u003c/li\u003e\n\u003cli\u003eQu, X. \u0026amp; Hall, A. On the persistent spread in snow-albedo feedback. \u003cem\u003eClim Dyn\u003c/em\u003e \u003cstrong\u003e42\u003c/strong\u003e, 69\u0026ndash;81 (2014).\u003c/li\u003e\n\u003cli\u003eShepherd, T. G. Atmospheric circulation as a source of uncertainty in climate change projections. \u003cem\u003eNature Geosci\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 703\u0026ndash;708 (2014).\u003c/li\u003e\n\u003cli\u003eHersbach, H. \u003cem\u003eet al.\u003c/em\u003e The ERA5 global reanalysis. \u003cem\u003eQuart J Royal Meteoro Soc\u003c/em\u003e \u003cstrong\u003e146\u003c/strong\u003e, 1999\u0026ndash;2049 (2020).\u003c/li\u003e\n\u003cli\u003eSchiemann, R. \u003cem\u003eet al.\u003c/em\u003e The Resolution Sensitivity of Northern Hemisphere Blocking in Four 25-km Atmospheric Global Circulation Models. \u003cem\u003eJ. Climate\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 337\u0026ndash;358 (2017).\u003c/li\u003e\n\u003cli\u003eAlmazroui, M. \u003cem\u003eet al.\u003c/em\u003e Projected Changes in Climate Extremes Using CMIP6 Simulations Over SREX Regions. \u003cem\u003eEarth Syst Environ\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 481\u0026ndash;497 (2021).\u003c/li\u003e\n\u003cli\u003eStouffer, R. J., Manabe, S. \u0026amp; Bryan, K. Interhemispheric asymmetry in climate response to a gradual increase of atmospheric CO2. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e342\u003c/strong\u003e, 660\u0026ndash;662 (1989).\u003c/li\u003e\n\u003cli\u003eVogel, M. M., Hauser, M. \u0026amp; Seneviratne, S. I. Projected changes in hot, dry and wet extreme events\u0026rsquo; clusters in CMIP6 multi-model ensemble. \u003cem\u003eEnviron. Res. Lett.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 094021 (2020).\u003c/li\u003e\n\u003cli\u003eBathiany, S., Dakos, V., Scheffer, M. \u0026amp; Lenton, T. M. Climate models predict increasing temperature variability in poor countries. \u003cem\u003eSci. Adv.\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, eaar5809 (2018).\u003c/li\u003e\n\u003cli\u003ePerkins, S. E. A review on the scientific understanding of heatwaves\u0026mdash;Their measurement, driving mechanisms, and changes at the global scale. \u003cem\u003eAtmospheric Research\u003c/em\u003e \u003cstrong\u003e164\u0026ndash;165\u003c/strong\u003e, 242\u0026ndash;267 (2015).\u003c/li\u003e\n\u003cli\u003eBerg, A. \u003cem\u003eet al.\u003c/em\u003e Impact of Soil Moisture\u0026ndash;Atmosphere Interactions on Surface Temperature Distribution. \u003cem\u003eJournal of Climate\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 7976\u0026ndash;7993 (2014).\u003c/li\u003e\n\u003cli\u003eSeneviratne, S. I., L\u0026uuml;thi, D., Litschi, M. \u0026amp; Sch\u0026auml;r, C. Land\u0026ndash;atmosphere coupling and climate change in Europe. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e443\u003c/strong\u003e, 205\u0026ndash;209 (2006).\u003c/li\u003e\n\u003cli\u003eLavell, A. \u003cem\u003eet al.\u003c/em\u003e \u003cem\u003eClimate Change: New Dimensions in Disaster Risk, Exposure, Vulnerability, and Resilience. In: Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation[Field, C.B., V. Barros, T.F. Stocker, D. Qin, D.J. Dokken, K.L. Ebi, M.D. Mastrandrea, K.J. Mach, G.-K. Plattner, S.K. Allen, M. Tignor, and P.M. Midgley (Eds.)]. A Special Report of Working Groups I and II of the Intergovernmental Panel on Climate Change (IPCC).\u003c/em\u003e (Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2012).\u003c/li\u003e\n\u003cli\u003eRuff, T. W. \u0026amp; Neelin, J. D. Long tails in regional surface temperature probability distributions with implications for extremes under global warming. \u003cem\u003eGeophysical Research Letters\u003c/em\u003e \u003cstrong\u003e39\u003c/strong\u003e, 2011GL050610 (2012).\u003c/li\u003e\n\u003cli\u003eSch\u0026auml;r, C. \u003cem\u003eet al.\u003c/em\u003e The role of increasing temperature variability in European summer heatwaves. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e427\u003c/strong\u003e, 332\u0026ndash;336 (2004).\u003c/li\u003e\n\u003cli\u003eMueller, B. \u0026amp; Seneviratne, S. I. Systematic land climate and evapotranspiration biases in CMIP5 simulations. \u003cem\u003eGeophysical Research Letters\u003c/em\u003e \u003cstrong\u003e41\u003c/strong\u003e, 128\u0026ndash;134 (2014).\u003c/li\u003e\n\u003cli\u003eChang, P. \u003cem\u003eet al.\u003c/em\u003e An Unprecedented Set of High‐Resolution Earth System Simulations for Understanding Multiscale Interactions in Climate Variability and Change. \u003cem\u003eJ Adv Model Earth Syst\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, (2020).\u003c/li\u003e\n\u003cli\u003eHaarsma, R. J. \u003cem\u003eet al.\u003c/em\u003e High Resolution Model Intercomparison Project (HighResMIP v1.0) for CMIP6. \u003cem\u003eGeosci. Model Dev.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 4185\u0026ndash;4208 (2016).\u003c/li\u003e\n\u003cli\u003eFischer, E. M. \u0026amp; Sch\u0026auml;r, C. Future changes in daily summer temperature variability: driving processes and role for temperature extremes. \u003cem\u003eClim Dyn\u003c/em\u003e \u003cstrong\u003e33\u003c/strong\u003e, 917\u0026ndash;935 (2009).\u003c/li\u003e\n\u003cli\u003eChattopadhyay, A., Nabizadeh, E. \u0026amp; Hassanzadeh, P. Analog Forecasting of Extreme‐Causing Weather Patterns Using Deep Learning. \u003cem\u003eJ Adv Model Earth Syst\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, (2020).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5393056/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5393056/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eHot extreme events are among the most devastating disasters affecting human health and the natural environment. While there is broad consensus on an increasing severity of these events under anthropogenic warming, their geographical distribution exhibits substantial spatial heterogeneity, and its driving factors remain uncertain. Here, utilizing an eddy-resolving high-resolution climate model alongside multiple simulations from Coupled Model Intercomparison Project Phase 6, we find baseline temperature variability as a key factor shaping the global distribution of projected hot extremes, with over 80% of the global increase in hot extremes anticorrelated with baseline temperature variability. We further demonstrate that the baseline temperature variability is anchored by persistent land-atmosphere coupling, which endures over century timescales and sustains the spatial heterogeneity of future hot extremes. Our findings suggest that baseline temperature variability could serve as a potential indicator for future hot extreme distribution, offering valuable insights for developing targeted adaptation strategies and improving regional resilience.\u003c/p\u003e","manuscriptTitle":"Baseline Temperature Variability Shaping Geographical Distribution of Future Hot Extremes under Anthropogenic Warming","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-02-11 14:11:14","doi":"10.21203/rs.3.rs-5393056/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"communications-earth-and-environment","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"commsenv","sideBox":"Learn more about [Communications Earth and Environment](https://www.nature.com/commsenv/)","snPcode":"","submissionUrl":"","title":"Communications Earth \u0026 Environment","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Communications Series","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"e78d9684-8805-47b2-86c0-ee38a67601f2","owner":[],"postedDate":"February 11th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":42096734,"name":"Earth and environmental sciences/Climate sciences/Climate change"},{"id":42096735,"name":"Earth and environmental sciences/Climate sciences/Climate change/Projection and prediction"}],"tags":[],"updatedAt":"2025-11-27T08:10:58+00:00","versionOfRecord":{"articleIdentity":"rs-5393056","link":"https://doi.org/10.1038/s43247-025-02929-3","journal":{"identity":"communications-earth-and-environment","isVorOnly":false,"title":"Communications Earth \u0026 Environment"},"publishedOn":"2025-11-26 05:00:00","publishedOnDateReadable":"November 26th, 2025"},"versionCreatedAt":"2025-02-11 14:11:14","video":"","vorDoi":"10.1038/s43247-025-02929-3","vorDoiUrl":"https://doi.org/10.1038/s43247-025-02929-3","workflowStages":[]},"version":"v1","identity":"rs-5393056","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5393056","identity":"rs-5393056","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}