Safe indoor temperature limits shape overheating resilience and cooling demand in California homes

Safe indoor temperature limits shape overheating resilience and cooling demand in California homes | 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 Article Safe indoor temperature limits shape overheating resilience and cooling demand in California homes Carlos Duarte Roa, Yan Wang, Harry Jiang, Charlie Huizenga, Aoyu Zou, and 6 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8264868/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract As climate change intensifies, the frequency and severity of extreme heat events are increasing, placing growing stress on building environments and energy systems. Using a physics-based residential building-stock framework, we conducted over 265,000 whole-building simulations (representing over 50,000 California homes across five climate scenarios) to quantify indoor overheating severity and the resulting increases in cooling demand. Indoor heat exposure was evaluated using the indoor overheating degree (IOD), which captures both the duration and magnitude of temperature exceedances above selected comfort limits. Results show that by the 2080s, up to ~ 55% of homes without cooling may exceed safe indoor limits, driving a 20–40% rise in statewide peak cooling electricity loads and potentially surpassing grid adequacy margins. Overheating risk moves from inland where AC is more common to coastal regions where it is not, increasing overall indoor heat exposure while narrowing regional disparities. By linking thermal limits with AC adoption and grid impacts, we quantify where overheating risks and cooling burdens intensify under current and future climates, providing evidence to support thermally safe and grid-aware housing policies. Earth and environmental sciences/Environmental social sciences/Climate-change impacts Physical sciences/Energy science and technology/Energy modelling Scientific community and society/Energy and society/Energy and behaviour Earth and environmental sciences/Natural hazards Climate Change Overheating Risks Air Conditioning Retrofit Cooling Energy Use Building Stock Modeling Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Global temperatures are projected to rise by 1.3-8.0°C above pre-industrial levels by the end of the century 1 and are expected to exacerbate the frequency, intensity, and duration of extreme heat events 2 , posing serious challenges to human health, infrastructure, and energy systems. Prolonged exposure to elevated indoor temperatures has been linked to adverse health outcomes, including cardiovascular 3 and respiratory 4 diseases, and mental health disorders. Vulnerable populations, especially the elderly, children, and low-income households 5 , may face even higher risks. In the western United States, including California, extreme heat events are further amplified by compounding environmental stressors such as drought, wildfires, and air pollution 6 , 7 . Buildings serve as the first line of shielding occupants from excessive heat and cold 8 . When properly designed and operated, buildings can moderate exposure to excessive heat through passive or mechanical means 9 . However, their thermal performance under heat stress is influenced by a combination of various factors, including building and mechanical system characteristics and occupant behavior 10 . While field monitoring campaigns 11 – 13 have provided valuable insights into residential overheating, their limited geographic and temporal scopes restrict generalizability. In contrast, building performance simulations enable systematic assessments of overheating risks across diverse housing conditions and enables one to test different scenarios. Among large-scale modeling methods, building archetype approaches are widely used. Typically, the regional or national housing stock is classified into representative types based on key building characteristics (e.g., vintage, construction, floor area), using data from surveys, census, and Geographic Information System (GIS) databases. In the US they are typically based on the Commercial Buildings Energy Consumption Survey (CBECS) 14 and Residential Energy Consumption Survey (RECS) 15 databases. Simulations are then conducted across thousands of variants by combining core archetypes with key parameters (e.g., orientation and occupant behavior) 16 to predict building energy consumption, peak electricity demand, and indoor environmental conditions. While efficient, this approach has notable limitations in that simplified combinations of building characteristics often fail to capture the true joint distributions and correlations observed in real housing stocks. As a result, simulated building populations may misrepresent the diversity and interdependencies of housing features that drive overheating and energy use. A further critical knowledge gap remains in that the interplay between residential overheating risks and air conditioning (AC) energy demand is rarely evaluated, although this relationship is essential to ensure safe indoor conditions without energy waste. As heat waves intensify, many existing buildings will struggle to passively maintain a comfortable indoor environment 17 – 19 . Consequently, occupants in heat-prone housing units will increasingly rely on AC to sustain thermal comfort, leading to higher energy consumption and straining the electricity grids during peak periods 20 , 21 . Homes lacking mechanical cooling may be forced to install AC units just to maintain safe indoor temperatures during extreme heat events. Particularly in regions like California, where heat waves are intensifying, it is vital to assess the resilience of the residential stock in terms of both occupant well-being and energy system implications. While studies 22 , 23 have acknowledged the rising cooling demand due to climate change, the specific impact of increased AC adoption in response to overheating remains uninvestigated. We addressed these gaps by pursuing two main objectives: (1) to evaluate residential overheating risk across California under current and future climates, and (2) to assess the resulting increases in cooling energy demand due to wider AC adoption. We employ a statistically representative sampling approach that draws directly from the RECS 15 dataset, preserving the multivariate correlations between key attributes such as building size, age, construction type, and cooling system availability. This enables more accurate estimation of both overheating risk and cooling energy demand at scale. Furthermore, we estimate the additional cooling energy use and peak electricity demand associated with AC installations in overheating-prone homes, offering quantitative insights into their impact on the power grid and informing equitable strategies for climate adaptation. Results Unequal cooling access across diverse climates As shown in Fig. 1 (a), California’s residential housing stock is dominated by single-family dwellings (~ 65%), with multi-family buildings accounting for ~ 32% and mobile homes representing only a minor share. The stock is also slightly older than the national average: the median dwelling age is 45 years, compared with 40 years nationally 24 , and most homes were constructed between the 1950s and 1980s. Yet demographic concentration shapes the statewide profile. Nearly a quarter of California’s housing units are located in Los Angeles County, reflecting the dominance of a single metropolitan region in the state’s housing stock (Fig. 1 (b)). Notably, AC availability is highly uneven (Fig. 1 (c)). Nearly half of dwellings (~ 47%) are equipped with central AC, and a smaller fraction (~ 15%) relies on room AC units or heat pumps. In contrast, more than one-third of homes lack any form of mechanical cooling provision. These homes without mechanical cooling are disproportionately concentrated in coastal counties, where historically milder climates suppressed AC adoption. Climatic variation further complicates these patterns. County-level analysis reveals pronounced regional differences in outdoor temperature (Fig. 1 (d)). Average outdoor temperatures exceed 20°C across much of the Central Valley and southern inland counties. In parallel, the outdoor overheating degree above 28°C, calculated as the cumulative exceedance above 28°C normalized by total cooling-season hours (detailed in the Methods section), is highest in the southern deserts and valleys, while coastal and northern regions remain relatively mild. Note that we include 28°C as one of the representative thresholds as it provides a mid-range reference point between milder (26°C) and more severe (30°C) exceedance limits. Results for these additional thresholds are provided in the Supplementary Information. When combined with AC prevalence (Fig. 1 (c)), these patterns indicate broader disparities in thermal resilience across regions. Many inland counties with hotter outdoor conditions exhibit relatively high AC penetration, yet coastal population centers with historically mild climates but limited AC access are increasingly exposed to extreme indoor heat. This combination of aging housing, partial air conditioning adoption, and diverse climate stresses highlights heterogeneous adaptive capacity across California’s residential sector. Statewide overheating risks under the current climate Figure 2 (a) - (c) illustrate the geographic distribution of indoor overheating risk across California’s dwellings without AC under the current climate. Overheating is quantified using the indoor overheating degree (IOD) metric 25 , which accounts for both the frequency and severity of indoor temperature exceedances above a specified threshold (as detailed in the Method). Unlike simple exceedance metrics that count only the number of hours above a limit, IOD integrates both the duration and magnitude of exceedance, consistent with established degree-hour–based metrics used in previous overheating studies 26 . IOD is calculated for three dry-bulb indoor air temperature thresholds (26°C, 28°C, and 30°C), representing a range of plausible comfort or regulatory limits. While these thresholds reflect building-level exceedance criteria rather than direct health indicators, prior studies note that temperatures near 26°C are associated with rising occupant thermal discomfort 27 , 28 , whereas indoor temperatures approaching 30°C have been linked to increased heat-related health risks in vulnerable groups 29 . It is important to emphasize that in this study, the temperature limits are used only to illustrate a modeling-based framework for assessing how different exceedance criteria influence potential increases in cooling demand, rather than to propose or derive specific IOD thresholds. For illustration, an IOD of 1.0°C at a 28°C limit indicates that, on average, indoor temperatures are one degree above 28°C across all cooling-season hours (equivalent to roughly 4,400 hours of exposure at 29°C), or two degrees above 28°C across half the cooling-season hours (about 2,200 hours at 30°C), etc. To support both policy decision-making and climate-based planning, results throughout Fig. 2 (a) - (c) are presented using two spatial aggregations: county boundaries and California Energy Commission (CEC) climate zones (CZs). County-level data aligns with local governance structures responsible for housing, public health, and resilience planning, while climate zones are useful for statewide building code enforcement and energy efficiency standards. At the 26°C threshold (panel a), overheating is widespread and intense across inland regions. CEC CZs 13–15, which cover the Central Valley and southeastern desert areas, exhibit the most extreme values, often exceeding 2.0–3.0°C. These areas also align with counties such as Kern, Fresno, and Imperial, where homes without AC experience persistent indoor thermal stress due to high daytime temperatures and limited nighttime cooling. In contrast, coastal zones such as CZs 1, 3, and 5 (counties along the Bay Area and North Coast) generally remain below 0.5°C, reflecting milder climate regimes. When the threshold is relaxed to 28°C (panel b), statewide overheating intensity declines, but the same inland climate zones and counties remain most vulnerable. Kern, Tulare, and parts of Southern Inland California still show IOD values above 1.0–2.0°C, while coastal counties and zones mostly fall below 0.25°C. At 30°C (panel c), overheating becomes more regionalized. Exceedance is largely confined to the hottest inland counties, including Imperial, Riverside, and the southern Central Valley (CZs 14 and 15), with IOD typically in the 1.0–3.0°C range. In contrast, the majority of counties in the state show minimal or negligible overheating. Taken together, these results show that raising the temperature threshold reduces the absolute magnitude and regional extent of indoor heat exposure, but does not eliminate regional vulnerability. Although the thresholds reflect different levels of heat stress, from milder exposure at 26°C to more severe exposure at 30°C, the same areas consistently rank as most affected. This consistency underscores that thermal risk is driven more by climate and building stock characteristics than by comfort definitions alone. Decision pathways linking comfort limits, AC adoption, and energy impact Figure 2 (d) presents a decision-support framework that links different indoor temperature thresholds to household AC adoption and energy system impacts. The x-axis represents IOD exceedance, which we refer to as the thermal tolerance margin, the extent to which indoor heat levels surpass the chosen temperature threshold. Low IOD values indicate that indoor temperatures only slightly exceed the threshold, while higher values reflect more severe or persistent indoor heat. The y-axis shows the relative change from the as-is baseline in two key outcomes: the number of dwellings that would require mechanical cooling and the projected increase in statewide AC peak electricity load if those dwellings adopted AC. Each curve corresponds to a specific indoor temperature threshold (26°C, 28°C, or 30°C), enabling decision-makers to evaluate how different indoor temperature standards influence both AC adoption and peak cooling demand. The figure shows a strongly nonlinear relationship between thermal tolerance and both AC adoption need and energy system impacts. At a 26°C comfort limit, an IOD of 0°C, meaning indoor temperatures just begin to exceed the threshold, is associated with roughly a 50% increase in dwellings requiring AC and about a 40% rise in statewide peak residential cooling load demand. This reflects the large share of homes currently without AC that would experience conditions above this limit at least occasionally. As the tolerance increases to 0.25°C, these values decline to ~ 41% for new AC needs and ~ 35% for peak load; at 0.5°C, the projected increases fall further to ~ 28% and ~ 26%, respectively. These sharp gradients suggest that even relatively small changes in building performance or policy thresholds can meaningfully shift the scale of required AC adoption and the resulting energy impacts. At the 28°C threshold, a 0°C exceedance results in a ~ 45% increase in AC installations and a ~ 40% increase in peak cooling demand; a 0.25°C tolerance reduces those impacts to ~ 24% and ~ 26%; a 0.5°C tolerance lowers them further to ~ 16% and ~ 19%, all less than the corresponding values at 26°C. At the 30°C threshold, zero tolerance corresponds to ~ 38% increase in homes needing ACs and ~ 32% higher peak load, declining to ~ 10% and ~ 8%, respectively, at a 0.5°C tolerance. Collectively, these results show that increases in cooling demand are much more sensitive to thermal tolerance margins in the lower-to-moderate range, whereas higher tolerance at higher temperature limits yields diminishing proportional responses. The divergence among the curves illustrates the policy implications of the temperature threshold selection. A lower indoor temperature limit (e.g., 26°C) ensures stronger protection for thermal comfort and health but drives a higher adoption of AC and increases peak electricity demand. A higher threshold (e.g., 30°C) reduces AC installation requirements and eases grid pressure but allows greater overheating exposure, which may be unsafe for sensitive groups. Future climate warming exacerbates overheating risks and electrification burdens Figure 3 presents projected changes in indoor thermal exposure and AC adoption needs across California under future climate scenarios, comparing mid-century (2050s) and late-century (2080s) timeframes for two emission pathways: (a) SSP1-2.6 (low-emissions) and (b) SSP5-8.5 (high-emissions). Within each scenario, county-level IOD maps are provided for the 26°C and 28°C thresholds for homes without ACs, along with the distribution of IOD exceedance and projected increases in AC adoption and peak cooling demand as functions of IOD exceedance. For brevity, only county-level maps at two thresholds are shown here, while corresponding climate zone visualizations and results at the 30°C threshold are available in the Supplementary Information. As expected, across all scenarios, future weather warming exacerbates indoor heat exposure, intensifying both the magnitude and spatial extent of overheating. However, the rate and severity of change differ markedly between the two pathways. Spatial intensification of overheating By the 2050s, even under the low-emissions pathway SSP1-2.6 (Fig. 3 a), much of inland California already surpasses IOD values of 2.0–3.0°C at the 26°C threshold, indicating sustained overheating exposure in homes without AC. This pattern is particularly pronounced in counties such as Kern, Imperial, and Riverside, aligning with elevated IOD levels in CEC CZs 13 and 15. At the 28°C threshold, IOD values above 1.0–2.0°C emerge in much of the Central Valley (CZ13). The SSP1-2.6 scenario for the 2080s and the SSP5-8.5 scenario for the 2050s produce spatial patterns and IOD distributions broadly similar to those under SSP1-2.6 for the 2050s and are therefore shown in the Supplementary Information. In contrast, by the 2080s under SSP5-8.5 (Fig. 3 b), indoor temperatures exceed the 26°C threshold across nearly the entire state, with IOD values above 2.0–4.0°C common in counties like Fresno, Kern, and Imperial. Similarly, CZs 13 and 15 show the highest IOD values, often exceeding 2.0–4.0°C at the 26°C threshold. By this point, at the 28°C threshold, widespread exceedance is also evident, with many interior regions surpassing 1.0–2.0°C and only a few coastal counties remaining below 1.0°C. The IOD exceedance distributions shown in the middle row of Fig. 3 include only homes with non-zero IOD values. It is worth noting that the majority of homes, especially under SSP1-2.6, still experience no overheating (IOD = 0) and are therefore not represented in these plots. Among the homes that do overheat, the distribution under SSP1-2.6 (2050s) peaks around 2.0°C at the 26°C threshold, with very few exceeding 4.0°C. At the 28°C threshold, most buildings remain below 1.0°C. In contrast, under SSP5-8.5 in the 2080s, the distribution of non-zero IOD values shifts notably toward higher exceedance. A larger fraction of homes exhibit IOD values above 1.0°C, and the right tail extends well beyond 4.0°C. This pattern is especially pronounced at the 26°C threshold but remains evident even at 28°C, indicating a widespread intensification of indoor heat exposure. This progression illustrates a clear intensification and spatial expansion of thermal stress, with risk growing in both magnitude and geographic footprint over time, especially under high-emissions climate scenarios. A key insight from these projections is the narrowing gap between temperature thresholds. Under the current climate, the choice of 26°C versus 28°C thresholds makes a substantial difference in both the magnitude and spatial extent of overheating. By the 2080s under SSP5-8.5, however, that difference is substantially reduced. Many counties that previously only exceeded 26°C are now exceeding 28°C with non-trivial IOD values. This ‘threshold convergence’ indicates that as indoor temperatures rise across the board, more buildings exceed all defined thresholds with minimal differentiation. This means that small variations in IOD no longer distinguish risk levels, reducing the effectiveness of temperature-based classification in targeting interventions. For example, programs that aim to reduce overheating by raising safe or comfort thresholds (e.g., from 26°C to 28°C with the use of fans 30 , 31 ) may be effective under moderate warming but become less impactful under severe climate scenarios. As such, by the late century, building envelope upgrades, mechanical ventilation, and active cooling will likely be necessary measures to maintain safe indoor conditions, regardless of thermal safety or comfort threshold. Projected AC demand and grid impact under future climates Under the as-is AC prevalence, statewide peak cooling electricity load is projected to increase by 23.8% and 31.5% under the SSP1-2.6 (2050s) and SSP5-8.5 (2080s) scenarios, respectively. The bottom panels of Fig. 3 illustrate the projected increases in new AC adoption and corresponding changes in cooling peak load under each climate scenario. These curves are derived from the same modeling framework used in Fig. 2 (d), applied to future IOD and cooling load projections. Differences across panels reflect how each climate pathway alters the distribution of IOD exceedance and, consequently, the scale of AC adoption required. Compared to the present climate, AC adoption curves under future climate scenarios appear notably flatter, particularly at moderate IOD exceedance levels (< 0.5°C). For example, under SSP1-2.6 (2050s), at a 26°C threshold and 0.5°C IOD exceedance, approximately 40% of homes are projected to require new AC, with a ~ 35% increase in peak cooling electricity demand, significantly higher than the ~ 28% increases seen under the present climate. These values rise only slightly by the 2080s under the low-emissions (SSP1-2.6) pathway, suggesting that sustained mitigation can constrain system-wide demand growth. In contrast, the high-emissions pathway (SSP5-8.5) in the late century shows a marked shift in both curve shape and magnitude. Even at the 28°C threshold and just 0.5°C of IOD exceedance, nearly 40% of uncooled homes would require AC, resulting in a ~ 27% increase in peak load. At higher exceedance levels (e.g., IOD = 1.0°C), projected outcomes plateau, with 30% of naturally cooled dwellings requiring AC while peak loads rise by another 25%. The flattening of these curves reflects a saturation effect. That is, as most buildings already exceed critical comfort limits, additional warming produces proportionally smaller increases in new AC adoption and peak load. Consequently, relaxing overheating standards yields diminishing reductions in projected energy and infrastructure burdens. Notably, the SSP5-8.5 (2080s) curve starts slightly below other scenarios at IOD = 0°C in terms of peak load growth. This counterintuitive result reflects a distributional shift. Under more extreme warming, most buildings leap past the low-exceedance zone, meaning that few remain near the threshold where IOD begins to rise. Consequently, the marginal impact captured at IOD < 0.5°C in this case is limited to mild-climate dwellings, producing a smaller relative increase in AC cooling peak load than in milder climate scenarios, where more buildings cluster near the adaptation threshold. Discussion Regional equity and system resilience Beyond statewide trends, future climate scenarios reveal notable shifts in the spatial inequality of overheating risk. While today's thermal vulnerability is concentrated in a relatively small number of hot inland counties, the late-century projections under the SSP5-8.5 scenario show that many more regions begin to exceed comfort thresholds across moderate IOD exceedance. For instance, the number of counties with an average IOD > 1.0°C at the 28°C threshold increases from fewer than 15 at present to over 40 by 2080. This suggests a rapid expansion of the population living in thermally stressed environments, which will complicate AC adoption targeting and policy prioritization. While IOD provides a quantitative measure of cumulative indoor heat exposure, there is currently no broadly established regulatory threshold specifying what constitutes an unacceptable IOD level. To date, the only explicit degree-hour–based overheating criterion in building standards is found in France’s residential thermal regulation, which defines a maximum annual overheating limit of 2600 degree-hours 32 . Establishing context-specific or health-based IOD thresholds is beyond the scope of this study. Instead, our aim is to demonstrate how a modeling-based framework can support future work that evaluates or develops such thresholds. In addition to expanding geographically, the severity of overheating also increases non-linearly. Under SSP5-8.5 in the 2080s, average IOD values for most inland counties at the 26°C threshold rise to above 2.0°C by the end of the century. This implies that not only are more households affected, but the magnitude of overheating is rising as well. Such compounding trends are likely to overwhelm informal coping strategies (e.g., night ventilation or fans), pushing demand toward more energy-intensive solutions. The analysis also reveals that even moderate exceedance levels (e.g., IOD = 0.25°C) could have large-scale implications due to the vast number of households affected. When scaled across millions of housing units, small per-household increases in discomfort translate into sizable cumulative cooling demand. This underscores the importance of early intervention and highlights the inadequacy of relying solely on retrofit programs targeting only the most overheated units. Instead, adaptation strategies must account for both high-risk and emerging-risk regions, ensuring scalable responses to a widening vulnerability landscape. Moreover, the projected rise in statewide peak cooling load poses a system-wide resilience challenge. Under SSP5-8.5 in the 2080s, the increase in statewide cooling peaks could exceed 3 GW, potentially outstripping the California Independent System Operator’s (CAISO) peak capacity buffer, especially during extreme heat events. These findings reinforce the need to pair cooling adoption with demand response and flexibility programs, energy-efficient cooling technologies, and decarbonized load growth strategies. Ultimately, the trajectory of overheating and cooling adoption in California will be shaped by the intersection of global emission pathways and local policy action. The state’s ability to mitigate grid strain, protect public health, and ensure equitable access to safe indoor temperatures depends not only on reducing greenhouse gas emissions but also on making proactive, data-informed investments in thermal resilience today. Together, the spatial maps and AC response curves in Fig. 3 offer a comprehensive picture of how future climate trajectories will reshape the indoor thermal landscape in the California residential sector. Under a low-emissions pathway, overheating growth is moderate and concentrated in areas already familiar with hot summers. Under a high-emissions pathway by the 2080s, indoor overheating becomes more widespread and intense, requiring rapid and large-scale adaptation of both buildings and infrastructure. These results provide a quantitative basis for designing climate-resilient housing policies, for example: Mandating indoor thermal safety thresholds in future Title 24 building codes. Expanding cooling retrofit and AC incentive programs in high-risk regions. Integrating overheating metrics into local heat-related early warning systems. Coordinating housing, health, and grid planning through shared metrics like IOD. Building retrofit implications and alignment with grid planning The projected expansion in cooling demand implies a potential massive retrofit challenge for California’s residential sector. Assuming a conservative unit-level retrofit cost of $ 6,000–10,000 per central AC installation (see the details for cost estimates in the Supplementary Information), a 50% increase in statewide AC saturation could require $ 20–40 billion in capital investment over the next few decades. These costs would disproportionately impact low-income households and regions with older, inefficient housing stock. These findings underscore the urgency of integrating overheating-driven AC demand into long-range energy planning frameworks. Current Integrated Resource Plans (IRPs) and building decarbonization pathways often assume gradual electrification of cooling loads. However, climate-driven overheating may accelerate demand growth well beyond forecasted rates. Aligning overheating risk maps with CAISO’s Transmission Planning Process (TPP) and CEC’s Title 24 modernization efforts could ensure that system upgrades are targeted where future adaptation needs and equity concerns are greatest. They also suggest that overheating metrics like IOD can serve as effective forward-planning indicators for where cooling demand will accelerate, allowing grid planners to align capacity expansion and decarbonization strategies with building adaptation needs. Limitations and future directions While this simulation-based study provides a spatially and temporally detailed assessment of indoor overheating risk and residential cooling demand across California, several limitations must be acknowledged. First, our modeling framework leverages a statistically representative housing sample (ResStock) and EnergyPlus, a well-established whole-building simulation engine with indoor air temperature predictions empirically validated in both test houses and real buildings 33 – 36 . However, extensive datasets of measured indoor temperatures spanning a wider range of buildings remain limited 37 – 38 , constraining direct validation at larger scales. Indoor temperature records are not yet well-documented in large observational datasets, particularly with the housing metadata needed for direct model comparison. Integrating data from smart thermostats, low-cost sensors, or targeted field campaigns could strengthen future calibration. Meanwhile, the projections assume static HVAC performance, while future improvements in AC efficiency or the adoption of grid-responsive controls (demand response) are not included, which may lead to an overestimation of peak demand impacts. Moreover, while the IOD thresholds offer a valuable behavioral proxy, the current adoption model is based on temperature exceedance, without factoring in socio-economic variables such as income, energy affordability, or housing tenure. Future work could overlay demographic and health vulnerability indicators to identify populations most at risk from both thermal stress and adaptation burden. Finally, the climate inputs used in this study reflect high- and low-emissions trajectories (SSP5-8.5 and SSP1-2.6, respectively) but do not include uncertainty bands, extreme heat events, or probabilistic weather realizations. Urban heat island effects are also not explicitly accounted for, which could lead to an underestimation of overheating risks in dense urban areas. Methods Building Stock Simulations We used the ResStock, a bottom-up, physics-based building stock simulation platform developed by the National Renewable Energy Laboratory (NREL). At its core is EnergyPlus, an open-source whole-building energy model maintained by the U.S. Department of Energy. ResStock incorporates stochastic occupant behavior models to capture the heterogeneous nature of household energy use 39 , and couples these with large public and private datasets, modified Latin hypercube sampling, and high-performance computing resources 40 . The framework enables county-level representation of U.S. housing stock while preserving variability in construction, equipment, and occupancy. Building on ResStock, our modeling framework consisted of four stages (Fig. 4 ). First, we generated over 50,000 EnergyPlus building models to represent California’s residential stock. This corresponds to ~ 10% of the national-scale dataset previously developed by NREL 41 , reflecting California’s share of U.S. dwelling units. Weather and climate scenarios We replaced the default Typical Meteorological Year (TMY) files in ResStock with weather files developed by the CEC, which better capture intra-regional variation (e.g., coastal vs. inland Los Angeles County). Future climate projections were generated with the Future Weather Generator 42 , which downscales General Circulation Model (GCM) outputs from the IPCC Sixth Assessment Report 43 into hourly EPW files. We considered SSP1-2.6 (low-emissions) and SSP5-8.5 (high-emissions) scenarios for the 2050s and 2080s to span a wide range of plausible futures. These two scenarios were selected to capture the broadest range of probable future climates and to represent best- and worst-case bounds in climate-sensitive decision-making. Overheating metric We ran simulations on an hourly basis for the cooling season (May 1–October 31) using a custom Python workflow for parallel execution and automated post-processing. We quantified overheating risk using the indoor overheating degree (IOD) metric, defined as the cumulative exceedance above a given limit (e.g., 28°C), normalized by the total number of cooling-season hours. IOD integrates both duration and magnitude of overheating, providing a measure of cumulative thermal burden rather than a binary exceedance count. From an environmental health perspective, epidemiological evidence shows that risk increases non-linearly with indoor temperature severity, not merely with frequency of exceedances 44 , 45 . From a policy perspective, a few marginal exceedances may be tolerable, but prolonged or severe events pose far greater risks. IOD thus offers a more robust basis for regulatory limits and enables discrimination between scenarios dominated by frequent mild exceedances versus fewer but more severe ones. Cooling demand and retrofit scenario To assess mitigation pathways, we estimated the incremental statewide cooling electricity demand that would result if overheating-prone dwellings (i.e., those exceeding the IOD threshold) were retrofitted with new AC systems. Each retrofit was assumed to deploy equipment with a seasonal energy efficiency ratio (SEER) of 15, consistent with California’s minimum standard for new residential AC units 46 but below the performance of high-efficiency models. This assumption yields a conservative estimate of the additional energy burden associated with large-scale AC adoption as a household adaptation measure to indoor overheating. Model validation ResStock has been extensively validated against utility billing data, end-use consumption, and measured load profiles 47 , 48 , demonstrating its ability to reproduce the magnitude and temporal dynamics of residential energy use with high fidelity. Building on this foundation, we benchmarked our simulated cooling demand against empirical datasets. Predicted average statewide residential cooling electricity consumption was within 9% of the 2019 Residential Appliance Saturation Study (RASS) 49 estimate (~ 800 kWh per household). We also validated peak cooling electricity demand by deriving coincident residential AC peaks from simulated non-coincident totals, using coincidence factors from metering studies, and by benchmarking against CAISO’s observations. The resulting coincident residential peak aligned with legacy statewide estimates (~ 7.5 GW) and with CAISO’s reported system peaks when disaggregated by sectoral shares. Together, these validations confirm that the framework reliably reproduces both the magnitude and timing of residential cooling demand in California, although validation of indoor temperature predictions remains challenging by the scarcity of large-scale indoor measurement datasets. Additional details, including equations and derivations, are provided in the Validation of Simulated Cooling Demand section of the Supplementary Information. References Scafetta N (2024) Impacts and risks of realistic global warming projections for the 21st century. Geosci Front 15:101774 Perkins SE, Alexander LV, Nairn JR (2012) Increasing frequency, intensity and duration of observed global heatwaves and warm spells. Geophys Res Lett 39:2012GL053361 Schwartz J, Samet JM, Patz JA (2004) Hospital Admissions for Heart Disease: The Effects of Temperature and Humidity. Epidemiology 15:755–761 Anderson GB et al (2013) Heat-related Emergency Hospitalizations for Respiratory Diseases in the Medicare Population. Am J Respir Crit Care Med 187:1098–1103 Elmallah S, Crespo Montañés C, Callaway D (2024) Who heats and cools? Access to residential heating and cooling in Northern California and implications for energy transitions. Energy Policy 191:114169 Franklin J, MacDonald GM (2024) Climate change and California sustainability—Challenges and solutions. Proc. Natl. Acad. Sci. U.S.A. 121, e2405458121 Rogers CDW et al (2021) Recent Increases in Exposure to Extreme Humid-Heat Events Disproportionately Affect Populated Regions. Geophysical Research Letters 48, e2021GL094183 Hong T et al (2023) Ten questions concerning thermal resilience of buildings and occupants for climate adaptation. Build Environ 244:110806 Grussa ZD et al (2019) A London residential retrofit case study: Evaluating passive mitigation methods of reducing risk to overheating through the use of solar shading combined with night-time ventilation. Build Serv Eng Res Tech 40:389–408 Lomas KJ, Porritt SM (2017) Overheating in buildings: lessons from research. Building Res Inform 45:1–18 Gamero-Salinas JC, Monge-Barrio A, Sánchez-Ostiz A (2020) Overheating risk assessment of different dwellings during the hottest season of a warm tropical climate. Build Environ 171:106664 Gupta R, Barnfield L, Gregg M (2017) Overheating in care settings: magnitude, causes, preparedness and remedies. Building Res Inform 45:83–101 Vellei M et al (2017) Overheating in vulnerable and non-vulnerable households. Building Res Inform 45:102–118 U.S. Energy Information Administration (2018) Commercial buildings energy consumption survey (CBECS). Retrieved July 10, 2025, from https://www.eia.gov/consumption/commercial/ U.S. Energy Information Administration (2020) Residential Energy Consumption Survey (RECS). Retrieved July 10, 2025, from https://www.eia.gov/consumption/residential/ Mavrogianni A, Wilkinson P, Davies M, Biddulph P, Oikonomou E (2012) Building characteristics as determinants of propensity to high indoor summer temperatures in London dwellings. Build Environ 55:117–130 Baniassadi A, Sailor DJ, Krayenhoff ES, Broadbent AM, Georgescu M (2019) Passive survivability of buildings under changing urban climates across eight US cities. Environ Res Lett 14:074028 Ahmed T, Kumar P, Mottet L (2021) Natural ventilation in warm climates: The challenges of thermal comfort, heatwave resilience and indoor air quality. Renew Sustain Energy Rev 138:110669 Quinn A et al (2014) Predicting indoor heat exposure risk during extreme heat events. Sci Total Environ 490:686–693 Sherman P, Lin H, McElroy M (2022) Projected global demand for air conditioning associated with extreme heat and implications for electricity grids in poorer countries. Energy Build 268:112198 Lundgren K, Kjellstrom T (2013) Sustainability Challenges from Climate Change and Air Conditioning Use in Urban Areas. Sustainability 5:3116–3128 Wang C et al (2023) Impacts of climate change, population growth, and power sector decarbonization on urban building energy use. Nat Commun 14:6434 Amonkar Y, Doss-Gollin J, Farnham DJ, Modi V, Lall U (2023) Differential effects of climate change on average and peak demand for heating and cooling across the contiguous USA. Commun Earth Environ 4:402 U.S. Census Bureau American Housing Survey: 2021 National Public Use File. (U.S. Department of Housing and Urban Development and U.S. Census Bureau, 2022); https://www.census.gov/programs-surveys/ahs.html Hamdy M, Carlucci S, Hoes PJ, Hensen JL (2017) The impact of climate change on the overheating risk in dwellings—A Dutch case study. Build Environ 122:307–323 Borgeson S, Brager G (2011) Comfort standards and variations in exceedance for mixed-mode buildings. Building Res Inform 39(2):118–133 De Dear RJ, Brager GS (2002) Thermal comfort in naturally ventilated buildings: revisions to ASHRAE Standard 55. Energy Build 34(6):549–561 Nicol JF, Humphreys MA (2002) Adaptive thermal comfort and sustainable thermal standards for buildings. Energy Build 34(6):563–572 Kenny GP, Tetzlaff EJ, Journeay WS, Henderson SB, O’Connor FK (2024) Indoor overheating: a review of vulnerabilities, causes, and strategies to prevent adverse human health outcomes during extreme heat events. Temperature 11(3):203–246 Miller D et al (2021) Cooling energy savings and occupant feedback in a two year retrofit evaluation of 99 automated ceiling fans staged with air conditioning. Energy Build 251:111319 Kent MG et al (2023) Energy savings and thermal comfort in a zero energy office building with fans in Singapore. Build Environ 243:110674 Attia, S., Benzidane, C., Rahif, R., Amaripadath, D., Hamdy, M., Holzer, P., … Carlucci,S. (2023). Overheating calculation methods, criteria, and indicators in European regulation for residential buildings. Energy and Buildings, 292, 113170. Anđelković AS, Mujan I, Dakić S (2016) Experimental validation of a EnergyPlus model: Application of a multi-storey naturally ventilated double skin façade. Energy Build 118:27–36 Mateus NM, Pinto A, Da Graça GC (2014) Validation of EnergyPlus thermal simulation of a double skin naturally and mechanically ventilated test cell. Energy Build 75:511–522 Loutzenhiser PG, Manz H, Felsmann C, Strachan PA, Frank TH, Maxwell GM (2007) Empirical validation of models to compute solar irradiance on inclined surfaces for building energy simulation. Sol Energy 81(2):254–267 Haves P, Ravache B, Fergadiotti A, Kohler C, J (2019) September). Accuracy of HVAC load predictions: Validation of EnergyPlus and DOE-2 using an instrumented test facility. Building Simulation 2019, vol 16. IBPSA, pp 4329–4336 Sutton-Klein J, Moody A, Hamilton I, Mindell JS (2021) Associations between indoor temperature, self-rated health and socioeconomic position in a cross-sectional study of adults in England. BMJ open, 11(2), e038500 Ahmed T, Kumar P, Mottet L (2021) Natural ventilation in warm climates: The challenges of thermal comfort, heatwave resilience and indoor air quality. Renew Sustain Energy Rev 138:110669 Chen J, Adhikari R, Wilson E, Robertson J, Fontanini A, Polly B, Olawale O (2022) Stochastic simulation of occupant-driven energy use in a bottom-up residential building stock model. Appl Energy 325:119890 Wilson EJ, Christensen CB, Horowitz SG, Robertson JJ, Maguire JB (2017) Energy Efficiency Potential in the U.S. Single-Family Housing Stock . NREL/TP–5500–68670, 1414819 http://www.osti.gov/servlets/purl/1414819 / 10.2172/1414819 National Renewable Energy Laboratory (2025) ResStock Public dataset. https://resstock.nrel.gov/datasets Rodrigues E, Fernandes MS, Carvalho D (2023) Future weather generator for building performance research: An open-source morphing tool and an application. Build Environ 233:110104 Eyring V et al (2016) Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci Model Dev 9:1937–1958 Sutton-Klein J, Moody A, Hamilton I, Mindell JS (2021) Associations between indoor temperature, self-rated health and socioeconomic position in a cross-sectional study of adults in England. BMJ open, 11(2), e038500 Edwards JR, De Roos AJ, Hampo CC, Huang W, Lincoln E, Hoque S, Schinasi LH (2025) Residential indoor temperatures and health: A scoping review of observational studies. Sci Total Environ 979:179377 California Energy Commission (2022) 2022 Building Energy Efficiency Standards for Residential and Nonresidential Buildings: For the 2022 Building Energy Efficiency Standards Title 24, Part 6, and Associated Administrative Regulations in Part 1. https://www.energy.ca.gov/programs-and-topics/programs/building-energy-efficiency-standards/2022-building-energy-efficiency Wilson EJ et al (2022) End-Use Load Profiles for the U.S. Building Stock: Methodology and Results of Model Calibration, Validation, and Uncertainty Quantification. Golden, CO: National Renewable Energy Laboratory. NREL/TP-5500-80889. https://doi.org/10.2172/1854582 Reyna J et al (2022) State Level Residential Building Stock and Energy Efficiency & Electrification Packages Analysis. Tableau Dashboard. Golden, CO: National Renewable Energy Laboratory. https://doi.org/10.2172/1877069 Palmgren C, Ramirez B, Goldberg M, Williamson C (2021) 2019 California Residential Appliance Saturation Study (RASS): Consultant Report. California Energy Commission. Publication Number: CEC-200-2021-005-PO Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryMaterialNatureCommunications.docx Supplementary Material Cite Share Download PDF Status: Under Review 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-8264868","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":561187973,"identity":"45cf0cd7-d65f-47d5-900c-32d8395e6f82","order_by":0,"name":"Carlos Duarte Roa","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA50lEQVRIiWNgGAWjYFCCBDApZ8DAwMzAUADmGBChJYHBGKLFgAQtiRuI1sLfnmP4ufKHTfp29t7HBh8MbBIb2Ju3SeDTInHmjbHkmYS03J09x40TZxikJTbwHCvDq8VAIsdAsiHhcO6GG2nMh3kMDic2SOSYEdJi/LMh4X+6wf1nzIf/GPxPbJB/Q1CLGdCWAwkGN9iYkxkMDgBt4cGvReLMszLLhrRkw509acyGPQbJxm08acUW+LTwtydvvtlgYydvzn6MWeJHhZ1sP/vhjTfwacEAjm0kKQcBe5J1jIJRMApGwbAHALb4Rw0wIUuSAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-5129-2969","institution":"University of California, Berkeley","correspondingAuthor":true,"prefix":"","firstName":"Carlos","middleName":"Duarte","lastName":"Roa","suffix":""},{"id":561187974,"identity":"764ed2eb-68ba-4464-b02f-7960aba71b2d","order_by":1,"name":"Yan Wang","email":"","orcid":"","institution":"University of California, Berkeley","correspondingAuthor":false,"prefix":"","firstName":"Yan","middleName":"","lastName":"Wang","suffix":""},{"id":561187975,"identity":"aefc4c59-5cc3-4468-9095-2438587c33bc","order_by":2,"name":"Harry Jiang","email":"","orcid":"","institution":"University of California, Berkeley","correspondingAuthor":false,"prefix":"","firstName":"Harry","middleName":"","lastName":"Jiang","suffix":""},{"id":561187976,"identity":"a99a4a53-ffee-4a52-9676-8068d2569c48","order_by":3,"name":"Charlie Huizenga","email":"","orcid":"https://orcid.org/0000-0001-7685-0106","institution":"University of California, Berkeley","correspondingAuthor":false,"prefix":"","firstName":"Charlie","middleName":"","lastName":"Huizenga","suffix":""},{"id":561187977,"identity":"72346bf0-be47-4295-9094-3a193dba2a12","order_by":4,"name":"Aoyu Zou","email":"","orcid":"","institution":"University of California, Berkeley","correspondingAuthor":false,"prefix":"","firstName":"Aoyu","middleName":"","lastName":"Zou","suffix":""},{"id":561187978,"identity":"dc0c060f-d10b-4362-958b-d909d81224f0","order_by":5,"name":"Tong Xiao","email":"","orcid":"https://orcid.org/0000-0002-2130-2281","institution":"University of California, Berkeley","correspondingAuthor":false,"prefix":"","firstName":"Tong","middleName":"","lastName":"Xiao","suffix":""},{"id":561187979,"identity":"b9ed8653-b554-4f3f-9252-b433fad0e5be","order_by":6,"name":"Guoquan Lv","email":"","orcid":"https://orcid.org/0000-0002-4889-1987","institution":"Department of Architecture, Zhejiang University","correspondingAuthor":false,"prefix":"","firstName":"Guoquan","middleName":"","lastName":"Lv","suffix":""},{"id":561187980,"identity":"4526006c-a6f8-4c53-9f0f-ab18118c62db","order_by":7,"name":"Tobias Kramer","email":"","orcid":"","institution":"University of California, Berkeley","correspondingAuthor":false,"prefix":"","firstName":"Tobias","middleName":"","lastName":"Kramer","suffix":""},{"id":561187981,"identity":"2ec7b884-9b2c-4a41-aec3-4d9ade7447b5","order_by":8,"name":"Stefano Schiavon","email":"","orcid":"https://orcid.org/0000-0003-1285-5682","institution":"University of California, Berkeley","correspondingAuthor":false,"prefix":"","firstName":"Stefano","middleName":"","lastName":"Schiavon","suffix":""},{"id":561187982,"identity":"eada2234-62b9-4943-b676-8b2f1751059b","order_by":9,"name":"Paul Raftery","email":"","orcid":"","institution":"Univesity of California, Berkeley","correspondingAuthor":false,"prefix":"","firstName":"Paul","middleName":"","lastName":"Raftery","suffix":""},{"id":561187983,"identity":"6e292ce9-a451-4149-bc33-4be20008df20","order_by":10,"name":"Gail Brager","email":"","orcid":"","institution":"University of California, Berkeley","correspondingAuthor":false,"prefix":"","firstName":"Gail","middleName":"","lastName":"Brager","suffix":""}],"badges":[],"createdAt":"2025-12-03 01:50:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8264868/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8264868/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":98440180,"identity":"dd35202e-198d-4958-a79a-9d555889d28a","added_by":"auto","created_at":"2025-12-17 17:03:28","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5094582,"visible":true,"origin":"","legend":"","description":"","filename":"OverheatingModellingNatureCommunicationscleaned.docx","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/baff2c9b47ce4e15bd4a4461.docx"},{"id":98440058,"identity":"094f0624-1478-45c2-ba3b-6e41d933142e","added_by":"auto","created_at":"2025-12-17 17:03:15","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":11526,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS2595576.json","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/a6bb3021a406ae6944190b47.json"},{"id":98386608,"identity":"1879f915-abff-4954-b04c-95ed6c5abb31","added_by":"auto","created_at":"2025-12-17 08:51:13","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2542538,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterialNatureCommunications.docx","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/a3ac091b97789232cd997ffc.docx"},{"id":98439647,"identity":"46545de3-d93a-4942-b8fa-bc45c07e6db9","added_by":"auto","created_at":"2025-12-17 17:02:24","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":110765,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS25955760enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/dd2b9ba43b02725b27e415d0.xml"},{"id":98386603,"identity":"4cfd7361-3f28-4ca7-9dc7-b23d96f36c88","added_by":"auto","created_at":"2025-12-17 08:51:12","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":489729,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/c5e6d06ec7cbe40110c8c079.png"},{"id":98440455,"identity":"4a190aa4-9057-4827-b1c1-9f538d1d3016","added_by":"auto","created_at":"2025-12-17 17:03:53","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":801352,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/d23da6337622b1eee23e21d5.png"},{"id":98386610,"identity":"f61075ec-2d0e-4c1e-9128-1a94e5366bd5","added_by":"auto","created_at":"2025-12-17 08:51:13","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":727554,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/64c546924d37cfac06fe5063.png"},{"id":98441095,"identity":"d986b6e3-5631-431d-9373-59fd66c066fc","added_by":"auto","created_at":"2025-12-17 17:04:52","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":182939,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/91a2690344b3abc4339d1c37.png"},{"id":98439545,"identity":"a3b57223-6c44-48db-be01-439de6f53e33","added_by":"auto","created_at":"2025-12-17 17:02:04","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":120170,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/ca80ecde93c446aff921ca02.png"},{"id":98386616,"identity":"6f406d5c-24d8-47dc-b1ec-cc0fa153ae44","added_by":"auto","created_at":"2025-12-17 08:51:13","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":116589,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/39f862107144d8c0fad4fe25.png"},{"id":98386605,"identity":"a430cc2f-8a8d-47b9-81d0-9ab990683044","added_by":"auto","created_at":"2025-12-17 08:51:12","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":95831,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/2c7ed8d72628b69c4be7384a.png"},{"id":98439898,"identity":"0ad83e70-c01b-49ac-a971-4cb1b4827271","added_by":"auto","created_at":"2025-12-17 17:03:03","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":38088,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/ee483156fa92072d100c09dc.png"},{"id":98386613,"identity":"da9261f1-83cf-4481-a46a-194343f671e2","added_by":"auto","created_at":"2025-12-17 08:51:13","extension":"xml","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":109693,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS25955760structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/b4758edb28b361bdc879bace.xml"},{"id":98441342,"identity":"799a1fbf-dcbd-43a9-8de3-cc015d4450f7","added_by":"auto","created_at":"2025-12-17 17:05:14","extension":"html","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":119431,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/686b1625d714c1c672c298e3.html"},{"id":98386600,"identity":"a3e4276c-57fb-4179-b20c-7485dd3f507b","added_by":"auto","created_at":"2025-12-17 08:51:12","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":564672,"visible":true,"origin":"","legend":"\u003cp\u003eHousing characteristics, cooling access, and outdoor heat exposure across California. (a) Distribution of residential building characteristics in the building samples, including building type, vintage, and AC types; (b) County-level representation of the building sample, weighted by housing stock distribution; (c) Prevalence of AC across counties, defined as the percentage of residential units with AC systems; (d) Spatial distribution of average outdoor dry-bulb temperature (May–October) across counties and California Energy Commission (CEC) climate zones; (e) Outdoor overheating degree, defined as cumulative temperature exceedance above 28 °C normalized by cooling-season hours for counties and CEC climate zones.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/2602841473e60d4c4f0398b4.png"},{"id":98386601,"identity":"82a1bf01-3ed4-4815-96ce-e56248b95191","added_by":"auto","created_at":"2025-12-17 08:51:12","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":654888,"visible":true,"origin":"","legend":"\u003cp\u003eIndoor overheating exposure and implications for statewide AC adoption under the current climate. (a)–(c) Indoor overheating degree (IOD) across California’s counties and CEC climate zones under present climate conditions, quantified as the average intensity of indoor temperature exceedance above each thermal comfort threshold: 26 °C (a), 28 °C (b), and 30 °C (c). IOD values reflect the intensity of overheating of non-air-conditioned homes during the cooling season (May–October); (d) Estimated statewide increase in new AC adoption and peak cooling electricity demand as a function of IOD exceedance, evaluated for each temperature threshold. For a given limit, exceedance represents the overheating tolerance margin before cooling adoption becomes necessary.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/7183dcd303e9208e713f67d8.png"},{"id":98386604,"identity":"e2e2a10d-b6c1-4c04-b8a6-092abbf191fb","added_by":"auto","created_at":"2025-12-17 08:51:12","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":544082,"visible":true,"origin":"","legend":"\u003cp\u003eProjected indoor overheating exposure and associated air conditioning (AC) adoption and energy impacts under future climate scenarios. Each column compares outcomes under different indoor overheating thresholds (26 °C and 28 °C) and climate projections: (a) SSP1-2.6 for the 2050s and (b) SSP5-8.5 for the 2080s. For each scenario-threshold combination, the top row shows county-level maps of indoor overheating degree (IOD); the middle row presents the distribution of IOD exceedance across statewide buildings without ACs (logarithmic x-axis); the bottom row depicts statewide projections of new AC adoption and peak electricity demand increase as functions of IOD exceedance under multiple thresholds (26 °C, 28 °C, and 30 °C).\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/02689c8f95ea6e0b19681f68.png"},{"id":98386614,"identity":"d4174292-d564-4b20-af7a-483488a00f0f","added_by":"auto","created_at":"2025-12-17 08:51:13","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":247477,"visible":true,"origin":"","legend":"\u003cp\u003eProject workflow diagram showing the process of generating input data files, running energy simulations, and analyzing output data files.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/8d66ee19a6358441b04a17c7.png"},{"id":98445658,"identity":"3beb857b-9286-40bd-8567-4a4879896bd3","added_by":"auto","created_at":"2025-12-17 17:20:41","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2338257,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/99be0fef-ccf1-4025-8176-7f4e78ad3a69.pdf"},{"id":98386599,"identity":"1c963fd5-b3d3-4cab-891a-da7543c042c7","added_by":"auto","created_at":"2025-12-17 08:51:12","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":2542538,"visible":true,"origin":"","legend":"Supplementary Material","description":"","filename":"SupplementaryMaterialNatureCommunications.docx","url":"https://assets-eu.researchsquare.com/files/rs-8264868/v1/8eab52482e4180aea2b08775.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Safe indoor temperature limits shape overheating resilience and cooling demand in California homes","fulltext":[{"header":"Introduction","content":"\u003cp\u003eGlobal temperatures are projected to rise by 1.3-8.0\u0026deg;C above pre-industrial levels by the end of the century\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e and are expected to exacerbate the frequency, intensity, and duration of extreme heat events\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, posing serious challenges to human health, infrastructure, and energy systems. Prolonged exposure to elevated indoor temperatures has been linked to adverse health outcomes, including cardiovascular\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e and respiratory\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e diseases, and mental health disorders. Vulnerable populations, especially the elderly, children, and low-income households\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e, may face even higher risks. In the western United States, including California, extreme heat events are further amplified by compounding environmental stressors such as drought, wildfires, and air pollution\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eBuildings serve as the first line of shielding occupants from excessive heat and cold\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. When properly designed and operated, buildings can moderate exposure to excessive heat through passive or mechanical means\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. However, their thermal performance under heat stress is influenced by a combination of various factors, including building and mechanical system characteristics and occupant behavior\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. While field monitoring campaigns\u003csup\u003e\u003cspan additionalcitationids=\"CR12\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e have provided valuable insights into residential overheating, their limited geographic and temporal scopes restrict generalizability. In contrast, building performance simulations enable systematic assessments of overheating risks across diverse housing conditions and enables one to test different scenarios.\u003c/p\u003e \u003cp\u003eAmong large-scale modeling methods, building archetype approaches are widely used. Typically, the regional or national housing stock is classified into representative types based on key building characteristics (e.g., vintage, construction, floor area), using data from surveys, census, and Geographic Information System (GIS) databases. In the US they are typically based on the Commercial Buildings Energy Consumption Survey (CBECS)\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e and Residential Energy Consumption Survey (RECS)\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e databases. Simulations are then conducted across thousands of variants by combining core archetypes with key parameters (e.g., orientation and occupant behavior)\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e to predict building energy consumption, peak electricity demand, and indoor environmental conditions. While efficient, this approach has notable limitations in that simplified combinations of building characteristics often fail to capture the true joint distributions and correlations observed in real housing stocks. As a result, simulated building populations may misrepresent the diversity and interdependencies of housing features that drive overheating and energy use.\u003c/p\u003e \u003cp\u003eA further critical knowledge gap remains in that the interplay between residential overheating risks and air conditioning (AC) energy demand is rarely evaluated, although this relationship is essential to ensure safe indoor conditions without energy waste. As heat waves intensify, many existing buildings will struggle to passively maintain a comfortable indoor environment\u003csup\u003e\u003cspan additionalcitationids=\"CR18\" citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. Consequently, occupants in heat-prone housing units will increasingly rely on AC to sustain thermal comfort, leading to higher energy consumption and straining the electricity grids during peak periods\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Homes lacking mechanical cooling may be forced to install AC units just to maintain safe indoor temperatures during extreme heat events. Particularly in regions like California, where heat waves are intensifying, it is vital to assess the resilience of the residential stock in terms of both occupant well-being and energy system implications. While studies\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e,\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e have acknowledged the rising cooling demand due to climate change, the specific impact of increased AC adoption in response to overheating remains uninvestigated.\u003c/p\u003e \u003cp\u003eWe addressed these gaps by pursuing two main objectives: (1) to evaluate residential overheating risk across California under current and future climates, and (2) to assess the resulting increases in cooling energy demand due to wider AC adoption. We employ a statistically representative sampling approach that draws directly from the RECS\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e dataset, preserving the multivariate correlations between key attributes such as building size, age, construction type, and cooling system availability. This enables more accurate estimation of both overheating risk and cooling energy demand at scale. Furthermore, we estimate the additional cooling energy use and peak electricity demand associated with AC installations in overheating-prone homes, offering quantitative insights into their impact on the power grid and informing equitable strategies for climate adaptation.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eUnequal cooling access across diverse climates\u003c/h2\u003e \u003cp\u003eAs shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e(a), California\u0026rsquo;s residential housing stock is dominated by single-family dwellings (~\u0026thinsp;65%), with multi-family buildings accounting for ~\u0026thinsp;32% and mobile homes representing only a minor share. The stock is also slightly older than the national average: the median dwelling age is 45 years, compared with 40 years nationally\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e, and most homes were constructed between the 1950s and 1980s. Yet demographic concentration shapes the statewide profile. Nearly a quarter of California\u0026rsquo;s housing units are located in Los Angeles County, reflecting the dominance of a single metropolitan region in the state\u0026rsquo;s housing stock (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e (b)). Notably, AC availability is highly uneven (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e (c)). Nearly half of dwellings (~\u0026thinsp;47%) are equipped with central AC, and a smaller fraction (~\u0026thinsp;15%) relies on room AC units or heat pumps. In contrast, more than one-third of homes lack any form of mechanical cooling provision. These homes without mechanical cooling are disproportionately concentrated in coastal counties, where historically milder climates suppressed AC adoption.\u003c/p\u003e \u003cp\u003eClimatic variation further complicates these patterns. County-level analysis reveals pronounced regional differences in outdoor temperature (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e(d)). Average outdoor temperatures exceed 20\u0026deg;C across much of the Central Valley and southern inland counties. In parallel, the outdoor overheating degree above 28\u0026deg;C, calculated as the cumulative exceedance above 28\u0026deg;C normalized by total cooling-season hours (detailed in the Methods section), is highest in the southern deserts and valleys, while coastal and northern regions remain relatively mild. Note that we include 28\u0026deg;C as one of the representative thresholds as it provides a mid-range reference point between milder (26\u0026deg;C) and more severe (30\u0026deg;C) exceedance limits. Results for these additional thresholds are provided in the Supplementary Information. When combined with AC prevalence (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e (c)), these patterns indicate broader disparities in thermal resilience across regions. Many inland counties with hotter outdoor conditions exhibit relatively high AC penetration, yet coastal population centers with historically mild climates but limited AC access are increasingly exposed to extreme indoor heat. This combination of aging housing, partial air conditioning adoption, and diverse climate stresses highlights heterogeneous adaptive capacity across California\u0026rsquo;s residential sector.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eStatewide overheating risks under the current climate\u003c/h3\u003e\n\u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e (a) - (c) illustrate the geographic distribution of indoor overheating risk across California\u0026rsquo;s dwellings without AC under the current climate. Overheating is quantified using the indoor overheating degree (IOD) metric\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e, which accounts for both the \u003cem\u003efrequency\u003c/em\u003e and \u003cem\u003eseverity\u003c/em\u003e of indoor temperature exceedances above a specified threshold (as detailed in the Method). Unlike simple exceedance metrics that count only the number of hours above a limit, IOD integrates both the duration and magnitude of exceedance, consistent with established degree-hour\u0026ndash;based metrics used in previous overheating studies\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. IOD is calculated for three dry-bulb indoor air temperature thresholds (26\u0026deg;C, 28\u0026deg;C, and 30\u0026deg;C), representing a range of plausible comfort or regulatory limits. While these thresholds reflect building-level exceedance criteria rather than direct health indicators, prior studies note that temperatures near 26\u0026deg;C are associated with rising occupant thermal discomfort\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e, whereas indoor temperatures approaching 30\u0026deg;C have been linked to increased heat-related health risks in vulnerable groups\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. It is important to emphasize that in this study, the temperature limits are used only to illustrate a modeling-based framework for assessing how different exceedance criteria influence potential increases in cooling demand, rather than to propose or derive specific IOD thresholds.\u003c/p\u003e \u003cp\u003eFor illustration, an IOD of 1.0\u0026deg;C at a 28\u0026deg;C limit indicates that, on average, indoor temperatures are one degree above 28\u0026deg;C across all cooling-season hours (equivalent to roughly 4,400 hours of exposure at 29\u0026deg;C), or two degrees above 28\u0026deg;C across half the cooling-season hours (about 2,200 hours at 30\u0026deg;C), etc. To support both policy decision-making and climate-based planning, results throughout Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e (a) - (c) are presented using two spatial aggregations: county boundaries and California Energy Commission (CEC) climate zones (CZs). County-level data aligns with local governance structures responsible for housing, public health, and resilience planning, while climate zones are useful for statewide building code enforcement and energy efficiency standards.\u003c/p\u003e \u003cp\u003eAt the 26\u0026deg;C threshold (panel a), overheating is widespread and intense across inland regions. CEC CZs 13\u0026ndash;15, which cover the Central Valley and southeastern desert areas, exhibit the most extreme values, often exceeding 2.0\u0026ndash;3.0\u0026deg;C. These areas also align with counties such as Kern, Fresno, and Imperial, where homes without AC experience persistent indoor thermal stress due to high daytime temperatures and limited nighttime cooling. In contrast, coastal zones such as CZs 1, 3, and 5 (counties along the Bay Area and North Coast) generally remain below 0.5\u0026deg;C, reflecting milder climate regimes. When the threshold is relaxed to 28\u0026deg;C (panel b), statewide overheating intensity declines, but the same inland climate zones and counties remain most vulnerable. Kern, Tulare, and parts of Southern Inland California still show IOD values above 1.0\u0026ndash;2.0\u0026deg;C, while coastal counties and zones mostly fall below 0.25\u0026deg;C. At 30\u0026deg;C (panel c), overheating becomes more regionalized. Exceedance is largely confined to the hottest inland counties, including Imperial, Riverside, and the southern Central Valley (CZs 14 and 15), with IOD typically in the 1.0\u0026ndash;3.0\u0026deg;C range. In contrast, the majority of counties in the state show minimal or negligible overheating.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTaken together, these results show that raising the temperature threshold reduces the absolute magnitude and regional extent of indoor heat exposure, but does not eliminate regional vulnerability. Although the thresholds reflect different levels of heat stress, from milder exposure at 26\u0026deg;C to more severe exposure at 30\u0026deg;C, the same areas consistently rank as most affected. This consistency underscores that thermal risk is driven more by climate and building stock characteristics than by comfort definitions alone.\u003c/p\u003e\n\u003ch3\u003eDecision pathways linking comfort limits, AC adoption, and energy impact\u003c/h3\u003e\n\u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e(d) presents a decision-support framework that links different indoor temperature thresholds to household AC adoption and energy system impacts. The x-axis represents IOD exceedance, which we refer to as the thermal tolerance margin, the extent to which indoor heat levels surpass the chosen temperature threshold. Low IOD values indicate that indoor temperatures only slightly exceed the threshold, while higher values reflect more severe or persistent indoor heat. The y-axis shows the relative change from the as-is baseline in two key outcomes: the number of dwellings that would require mechanical cooling and the projected increase in statewide AC peak electricity load if those dwellings adopted AC. Each curve corresponds to a specific indoor temperature threshold (26\u0026deg;C, 28\u0026deg;C, or 30\u0026deg;C), enabling decision-makers to evaluate how different indoor temperature standards influence both AC adoption and peak cooling demand.\u003c/p\u003e \u003cp\u003eThe figure shows a strongly nonlinear relationship between thermal tolerance and both AC adoption need and energy system impacts. At a 26\u0026deg;C comfort limit, an IOD of 0\u0026deg;C, meaning indoor temperatures just begin to exceed the threshold, is associated with roughly a 50% increase in dwellings requiring AC and about a 40% rise in statewide peak residential cooling load demand. This reflects the large share of homes currently without AC that would experience conditions above this limit at least occasionally. As the tolerance increases to 0.25\u0026deg;C, these values decline to ~\u0026thinsp;41% for new AC needs and ~\u0026thinsp;35% for peak load; at 0.5\u0026deg;C, the projected increases fall further to ~\u0026thinsp;28% and ~\u0026thinsp;26%, respectively. These sharp gradients suggest that even relatively small changes in building performance or policy thresholds can meaningfully shift the scale of required AC adoption and the resulting energy impacts.\u003c/p\u003e \u003cp\u003eAt the 28\u0026deg;C threshold, a 0\u0026deg;C exceedance results in a\u0026thinsp;~\u0026thinsp;45% increase in AC installations and a\u0026thinsp;~\u0026thinsp;40% increase in peak cooling demand; a 0.25\u0026deg;C tolerance reduces those impacts to ~\u0026thinsp;24% and ~\u0026thinsp;26%; a 0.5\u0026deg;C tolerance lowers them further to ~\u0026thinsp;16% and ~\u0026thinsp;19%, all less than the corresponding values at 26\u0026deg;C. At the 30\u0026deg;C threshold, zero tolerance corresponds to ~\u0026thinsp;38% increase in homes needing ACs and ~\u0026thinsp;32% higher peak load, declining to ~\u0026thinsp;10% and ~\u0026thinsp;8%, respectively, at a 0.5\u0026deg;C tolerance. Collectively, these results show that increases in cooling demand are much more sensitive to thermal tolerance margins in the lower-to-moderate range, whereas higher tolerance at higher temperature limits yields diminishing proportional responses.\u003c/p\u003e \u003cp\u003eThe divergence among the curves illustrates the policy implications of the temperature threshold selection. A lower indoor temperature limit (e.g., 26\u0026deg;C) ensures stronger protection for thermal comfort and health but drives a higher adoption of AC and increases peak electricity demand. A higher threshold (e.g., 30\u0026deg;C) reduces AC installation requirements and eases grid pressure but allows greater overheating exposure, which may be unsafe for sensitive groups.\u003c/p\u003e\n\u003ch3\u003eFuture climate warming exacerbates overheating risks and electrification burdens\u003c/h3\u003e\n\u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e presents projected changes in indoor thermal exposure and AC adoption needs across California under future climate scenarios, comparing mid-century (2050s) and late-century (2080s) timeframes for two emission pathways: (a) SSP1-2.6 (low-emissions) and (b) SSP5-8.5 (high-emissions). Within each scenario, county-level IOD maps are provided for the 26\u0026deg;C and 28\u0026deg;C thresholds for homes without ACs, along with the distribution of IOD exceedance and projected increases in AC adoption and peak cooling demand as functions of IOD exceedance. For brevity, only county-level maps at two thresholds are shown here, while corresponding climate zone visualizations and results at the 30\u0026deg;C threshold are available in the Supplementary Information. As expected, across all scenarios, future weather warming exacerbates indoor heat exposure, intensifying both the magnitude and spatial extent of overheating. However, the rate and severity of change differ markedly between the two pathways.\u003c/p\u003e\n\u003ch3\u003eSpatial intensification of overheating\u003c/h3\u003e\n\u003cp\u003eBy the 2050s, even under the low-emissions pathway SSP1-2.6 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea), much of inland California already surpasses IOD values of 2.0\u0026ndash;3.0\u0026deg;C at the 26\u0026deg;C threshold, indicating sustained overheating exposure in homes without AC. This pattern is particularly pronounced in counties such as Kern, Imperial, and Riverside, aligning with elevated IOD levels in CEC CZs 13 and 15. At the 28\u0026deg;C threshold, IOD values above 1.0\u0026ndash;2.0\u0026deg;C emerge in much of the Central Valley (CZ13). The SSP1-2.6 scenario for the 2080s and the SSP5-8.5 scenario for the 2050s produce spatial patterns and IOD distributions broadly similar to those under SSP1-2.6 for the 2050s and are therefore shown in the Supplementary Information. In contrast, by the 2080s under SSP5-8.5 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb), indoor temperatures exceed the 26\u0026deg;C threshold across nearly the entire state, with IOD values above 2.0\u0026ndash;4.0\u0026deg;C common in counties like Fresno, Kern, and Imperial. Similarly, CZs 13 and 15 show the highest IOD values, often exceeding 2.0\u0026ndash;4.0\u0026deg;C at the 26\u0026deg;C threshold. By this point, at the 28\u0026deg;C threshold, widespread exceedance is also evident, with many interior regions surpassing 1.0\u0026ndash;2.0\u0026deg;C and only a few coastal counties remaining below 1.0\u0026deg;C.\u003c/p\u003e \u003cp\u003eThe IOD exceedance distributions shown in the middle row of Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e include only homes with non-zero IOD values. It is worth noting that the majority of homes, especially under SSP1-2.6, still experience no overheating (IOD\u0026thinsp;=\u0026thinsp;0) and are therefore not represented in these plots. Among the homes that do overheat, the distribution under SSP1-2.6 (2050s) peaks around 2.0\u0026deg;C at the 26\u0026deg;C threshold, with very few exceeding 4.0\u0026deg;C. At the 28\u0026deg;C threshold, most buildings remain below 1.0\u0026deg;C. In contrast, under SSP5-8.5 in the 2080s, the distribution of non-zero IOD values shifts notably toward higher exceedance. A larger fraction of homes exhibit IOD values above 1.0\u0026deg;C, and the right tail extends well beyond 4.0\u0026deg;C. This pattern is especially pronounced at the 26\u0026deg;C threshold but remains evident even at 28\u0026deg;C, indicating a widespread intensification of indoor heat exposure.\u003c/p\u003e \u003cp\u003eThis progression illustrates a clear intensification and spatial expansion of thermal stress, with risk growing in both magnitude and geographic footprint over time, especially under high-emissions climate scenarios. A key insight from these projections is the narrowing gap between temperature thresholds. Under the current climate, the choice of 26\u0026deg;C versus 28\u0026deg;C thresholds makes a substantial difference in both the magnitude and spatial extent of overheating. By the 2080s under SSP5-8.5, however, that difference is substantially reduced. Many counties that previously only exceeded 26\u0026deg;C are now exceeding 28\u0026deg;C with non-trivial IOD values. This \u0026lsquo;threshold convergence\u0026rsquo; indicates that as indoor temperatures rise across the board, more buildings exceed all defined thresholds with minimal differentiation. This means that small variations in IOD no longer distinguish risk levels, reducing the effectiveness of temperature-based classification in targeting interventions. For example, programs that aim to reduce overheating by raising safe or comfort thresholds (e.g., from 26\u0026deg;C to 28\u0026deg;C with the use of fans\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e) may be effective under moderate warming but become less impactful under severe climate scenarios. As such, by the late century, building envelope upgrades, mechanical ventilation, and active cooling will likely be necessary measures to maintain safe indoor conditions, regardless of thermal safety or comfort threshold.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eProjected AC demand and grid impact under future climates\u003c/h2\u003e \u003cp\u003eUnder the as-is AC prevalence, statewide peak cooling electricity load is projected to increase by 23.8% and 31.5% under the SSP1-2.6 (2050s) and SSP5-8.5 (2080s) scenarios, respectively. The bottom panels of Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e illustrate the projected increases in new AC adoption and corresponding changes in cooling peak load under each climate scenario. These curves are derived from the same modeling framework used in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e(d), applied to future IOD and cooling load projections. Differences across panels reflect how each climate pathway alters the distribution of IOD exceedance and, consequently, the scale of AC adoption required.\u003c/p\u003e \u003cp\u003eCompared to the present climate, AC adoption curves under future climate scenarios appear notably flatter, particularly at moderate IOD exceedance levels (\u0026lt;\u0026thinsp;0.5\u0026deg;C). For example, under SSP1-2.6 (2050s), at a 26\u0026deg;C threshold and 0.5\u0026deg;C IOD exceedance, approximately 40% of homes are projected to require new AC, with a\u0026thinsp;~\u0026thinsp;35% increase in peak cooling electricity demand, significantly higher than the ~\u0026thinsp;28% increases seen under the present climate.\u003c/p\u003e \u003cp\u003eThese values rise only slightly by the 2080s under the low-emissions (SSP1-2.6) pathway, suggesting that sustained mitigation can constrain system-wide demand growth. In contrast, the high-emissions pathway (SSP5-8.5) in the late century shows a marked shift in both curve shape and magnitude. Even at the 28\u0026deg;C threshold and just 0.5\u0026deg;C of IOD exceedance, nearly 40% of uncooled homes would require AC, resulting in a\u0026thinsp;~\u0026thinsp;27% increase in peak load. At higher exceedance levels (e.g., IOD\u0026thinsp;=\u0026thinsp;1.0\u0026deg;C), projected outcomes plateau, with 30% of naturally cooled dwellings requiring AC while peak loads rise by another 25%. The flattening of these curves reflects a saturation effect. That is, as most buildings already exceed critical comfort limits, additional warming produces proportionally smaller increases in new AC adoption and peak load. Consequently, relaxing overheating standards yields diminishing reductions in projected energy and infrastructure burdens.\u003c/p\u003e \u003cp\u003eNotably, the SSP5-8.5 (2080s) curve starts slightly below other scenarios at IOD\u0026thinsp;=\u0026thinsp;0\u0026deg;C in terms of peak load growth. This counterintuitive result reflects a distributional shift. Under more extreme warming, most buildings leap past the low-exceedance zone, meaning that few remain near the threshold where IOD begins to rise. Consequently, the marginal impact captured at IOD\u0026thinsp;\u0026lt;\u0026thinsp;0.5\u0026deg;C in this case is limited to mild-climate dwellings, producing a smaller relative increase in AC cooling peak load than in milder climate scenarios, where more buildings cluster near the adaptation threshold.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eRegional equity and system resilience\u003c/h2\u003e \u003cp\u003eBeyond statewide trends, future climate scenarios reveal notable shifts in the spatial inequality of overheating risk. While today's thermal vulnerability is concentrated in a relatively small number of hot inland counties, the late-century projections under the SSP5-8.5 scenario show that many more regions begin to exceed comfort thresholds across moderate IOD exceedance. For instance, the number of counties with an average IOD\u0026thinsp;\u0026gt;\u0026thinsp;1.0\u0026deg;C at the 28\u0026deg;C threshold increases from fewer than 15 at present to over 40 by 2080. This suggests a rapid expansion of the population living in thermally stressed environments, which will complicate AC adoption targeting and policy prioritization.\u003c/p\u003e \u003cp\u003eWhile IOD provides a quantitative measure of cumulative indoor heat exposure, there is currently no broadly established regulatory threshold specifying what constitutes an unacceptable IOD level. To date, the only explicit degree-hour\u0026ndash;based overheating criterion in building standards is found in France\u0026rsquo;s residential thermal regulation, which defines a maximum annual overheating limit of 2600 degree-hours\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. Establishing context-specific or health-based IOD thresholds is beyond the scope of this study. Instead, our aim is to demonstrate how a modeling-based framework can support future work that evaluates or develops such thresholds.\u003c/p\u003e \u003cp\u003eIn addition to expanding geographically, the severity of overheating also increases non-linearly. Under SSP5-8.5 in the 2080s, average IOD values for most inland counties at the 26\u0026deg;C threshold rise to above 2.0\u0026deg;C by the end of the century. This implies that not only are more households affected, but the magnitude of overheating is rising as well. Such compounding trends are likely to overwhelm informal coping strategies (e.g., night ventilation or fans), pushing demand toward more energy-intensive solutions.\u003c/p\u003e \u003cp\u003eThe analysis also reveals that even moderate exceedance levels (e.g., IOD\u0026thinsp;=\u0026thinsp;0.25\u0026deg;C) could have large-scale implications due to the vast number of households affected. When scaled across millions of housing units, small per-household increases in discomfort translate into sizable cumulative cooling demand. This underscores the importance of early intervention and highlights the inadequacy of relying solely on retrofit programs targeting only the most overheated units. Instead, adaptation strategies must account for both high-risk and emerging-risk regions, ensuring scalable responses to a widening vulnerability landscape.\u003c/p\u003e \u003cp\u003eMoreover, the projected rise in statewide peak cooling load poses a system-wide resilience challenge. Under SSP5-8.5 in the 2080s, the increase in statewide cooling peaks could exceed 3 GW, potentially outstripping the California Independent System Operator\u0026rsquo;s (CAISO) peak capacity buffer, especially during extreme heat events. These findings reinforce the need to pair cooling adoption with demand response and flexibility programs, energy-efficient cooling technologies, and decarbonized load growth strategies. Ultimately, the trajectory of overheating and cooling adoption in California will be shaped by the intersection of global emission pathways and local policy action. The state\u0026rsquo;s ability to mitigate grid strain, protect public health, and ensure equitable access to safe indoor temperatures depends not only on reducing greenhouse gas emissions but also on making proactive, data-informed investments in thermal resilience today.\u003c/p\u003e \u003cp\u003eTogether, the spatial maps and AC response curves in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e offer a comprehensive picture of how future climate trajectories will reshape the indoor thermal landscape in the California residential sector. Under a low-emissions pathway, overheating growth is moderate and concentrated in areas already familiar with hot summers. Under a high-emissions pathway by the 2080s, indoor overheating becomes more widespread and intense, requiring rapid and large-scale adaptation of both buildings and infrastructure. These results provide a quantitative basis for designing climate-resilient housing policies, for example:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eMandating indoor thermal safety thresholds in future Title 24 building codes.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eExpanding cooling retrofit and AC incentive programs in high-risk regions.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eIntegrating overheating metrics into local heat-related early warning systems.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eCoordinating housing, health, and grid planning through shared metrics like IOD.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eBuilding retrofit implications and alignment with grid planning\u003c/h2\u003e \u003cp\u003eThe projected expansion in cooling demand implies a potential massive retrofit challenge for California\u0026rsquo;s residential sector. Assuming a conservative unit-level retrofit cost of \u003cspan\u003e$\u003c/span\u003e6,000\u0026ndash;10,000 per central AC installation (see the details for cost estimates in the Supplementary Information), a 50% increase in statewide AC saturation could require \u003cspan\u003e$\u003c/span\u003e20\u0026ndash;40\u0026nbsp;billion in capital investment over the next few decades. These costs would disproportionately impact low-income households and regions with older, inefficient housing stock.\u003c/p\u003e \u003cp\u003eThese findings underscore the urgency of integrating overheating-driven AC demand into long-range energy planning frameworks. Current Integrated Resource Plans (IRPs) and building decarbonization pathways often assume gradual electrification of cooling loads. However, climate-driven overheating may accelerate demand growth well beyond forecasted rates. Aligning overheating risk maps with CAISO\u0026rsquo;s Transmission Planning Process (TPP) and CEC\u0026rsquo;s Title 24 modernization efforts could ensure that system upgrades are targeted where future adaptation needs and equity concerns are greatest. They also suggest that overheating metrics like IOD can serve as effective forward-planning indicators for where cooling demand will accelerate, allowing grid planners to align capacity expansion and decarbonization strategies with building adaptation needs.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eLimitations and future directions\u003c/h2\u003e \u003cp\u003eWhile this simulation-based study provides a spatially and temporally detailed assessment of indoor overheating risk and residential cooling demand across California, several limitations must be acknowledged. First, our modeling framework leverages a statistically representative housing sample (ResStock) and EnergyPlus, a well-established whole-building simulation engine with indoor air temperature predictions empirically validated in both test houses and real buildings\u003csup\u003e\u003cspan additionalcitationids=\"CR34 CR35\" citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. However, extensive datasets of measured indoor temperatures spanning a wider range of buildings remain limited\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e, constraining direct validation at larger scales. Indoor temperature records are not yet well-documented in large observational datasets, particularly with the housing metadata needed for direct model comparison. Integrating data from smart thermostats, low-cost sensors, or targeted field campaigns could strengthen future calibration. Meanwhile, the projections assume static HVAC performance, while future improvements in AC efficiency or the adoption of grid-responsive controls (demand response) are not included, which may lead to an overestimation of peak demand impacts. Moreover, while the IOD thresholds offer a valuable behavioral proxy, the current adoption model is based on temperature exceedance, without factoring in socio-economic variables such as income, energy affordability, or housing tenure. Future work could overlay demographic and health vulnerability indicators to identify populations most at risk from both thermal stress and adaptation burden. Finally, the climate inputs used in this study reflect high- and low-emissions trajectories (SSP5-8.5 and SSP1-2.6, respectively) but do not include uncertainty bands, extreme heat events, or probabilistic weather realizations. Urban heat island effects are also not explicitly accounted for, which could lead to an underestimation of overheating risks in dense urban areas.\u003c/p\u003e \u003c/div\u003e "},{"header":"Methods","content":"\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eBuilding Stock Simulations\u003c/h2\u003e \u003cp\u003eWe used the ResStock, a bottom-up, physics-based building stock simulation platform developed by the National Renewable Energy Laboratory (NREL). At its core is EnergyPlus, an open-source whole-building energy model maintained by the U.S. Department of Energy. ResStock incorporates stochastic occupant behavior models to capture the heterogeneous nature of household energy use\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e, and couples these with large public and private datasets, modified Latin hypercube sampling, and high-performance computing resources\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. The framework enables county-level representation of U.S. housing stock while preserving variability in construction, equipment, and occupancy. Building on ResStock, our modeling framework consisted of four stages (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). First, we generated over 50,000 EnergyPlus building models to represent California\u0026rsquo;s residential stock. This corresponds to ~\u0026thinsp;10% of the national-scale dataset previously developed by NREL\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e, reflecting California\u0026rsquo;s share of U.S. dwelling units.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eWeather and climate scenarios\u003c/h2\u003e \u003cp\u003eWe replaced the default Typical Meteorological Year (TMY) files in ResStock with weather files developed by the CEC, which better capture intra-regional variation (e.g., coastal vs. inland Los Angeles County). Future climate projections were generated with the \u003cem\u003eFuture Weather Generator\u003c/em\u003e\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e, which downscales General Circulation Model (GCM) outputs from the IPCC Sixth Assessment Report\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e into hourly EPW files. We considered SSP1-2.6 (low-emissions) and SSP5-8.5 (high-emissions) scenarios for the 2050s and 2080s to span a wide range of plausible futures. These two scenarios were selected to capture the broadest range of probable future climates and to represent best- and worst-case bounds in climate-sensitive decision-making.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eOverheating metric\u003c/h2\u003e \u003cp\u003eWe ran simulations on an hourly basis for the cooling season (May 1\u0026ndash;October 31) using a custom Python workflow for parallel execution and automated post-processing. We quantified overheating risk using the indoor overheating degree (IOD) metric, defined as the cumulative exceedance above a given limit (e.g., 28\u0026deg;C), normalized by the total number of cooling-season hours. IOD integrates both duration and magnitude of overheating, providing a measure of cumulative thermal burden rather than a binary exceedance count. From an environmental health perspective, epidemiological evidence shows that risk increases non-linearly with indoor temperature severity, not merely with frequency of exceedances\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e,\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. From a policy perspective, a few marginal exceedances may be tolerable, but prolonged or severe events pose far greater risks. IOD thus offers a more robust basis for regulatory limits and enables discrimination between scenarios dominated by frequent mild exceedances versus fewer but more severe ones.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eCooling demand and retrofit scenario\u003c/h2\u003e \u003cp\u003eTo assess mitigation pathways, we estimated the incremental statewide cooling electricity demand that would result if overheating-prone dwellings (i.e., those exceeding the IOD threshold) were retrofitted with new AC systems. Each retrofit was assumed to deploy equipment with a seasonal energy efficiency ratio (SEER) of 15, consistent with California\u0026rsquo;s minimum standard for new residential AC units\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e but below the performance of high-efficiency models. This assumption yields a conservative estimate of the additional energy burden associated with large-scale AC adoption as a household adaptation measure to indoor overheating.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eModel validation\u003c/h2\u003e \u003cp\u003eResStock has been extensively validated against utility billing data, end-use consumption, and measured load profiles\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e,\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e, demonstrating its ability to reproduce the magnitude and temporal dynamics of residential energy use with high fidelity. Building on this foundation, we benchmarked our simulated cooling demand against empirical datasets. Predicted average statewide residential cooling electricity consumption was within 9% of the 2019 Residential Appliance Saturation Study (RASS)\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e estimate (~\u0026thinsp;800 kWh per household). We also validated peak cooling electricity demand by deriving coincident residential AC peaks from simulated non-coincident totals, using coincidence factors from metering studies, and by benchmarking against CAISO\u0026rsquo;s observations. The resulting coincident residential peak aligned with legacy statewide estimates (~\u0026thinsp;7.5 GW) and with CAISO\u0026rsquo;s reported system peaks when disaggregated by sectoral shares. Together, these validations confirm that the framework reliably reproduces both the magnitude and timing of residential cooling demand in California, although validation of indoor temperature predictions remains challenging by the scarcity of large-scale indoor measurement datasets. Additional details, including equations and derivations, are provided in the \u003cem\u003eValidation of Simulated Cooling Demand\u003c/em\u003e section of the Supplementary Information.\u003c/p\u003e \u003c/div\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eScafetta N (2024) Impacts and risks of realistic global warming projections for the 21st century. Geosci Front 15:101774\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePerkins SE, Alexander LV, Nairn JR (2012) Increasing frequency, intensity and duration of observed global heatwaves and warm spells. Geophys Res Lett 39:2012GL053361\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchwartz J, Samet JM, Patz JA (2004) Hospital Admissions for Heart Disease: The Effects of Temperature and Humidity. Epidemiology 15:755\u0026ndash;761\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnderson GB et al (2013) Heat-related Emergency Hospitalizations for Respiratory Diseases in the Medicare Population. Am J Respir Crit Care Med 187:1098\u0026ndash;1103\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eElmallah S, Crespo Monta\u0026ntilde;\u0026eacute;s C, Callaway D (2024) Who heats and cools? Access to residential heating and cooling in Northern California and implications for energy transitions. Energy Policy 191:114169\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFranklin J, MacDonald GM (2024) Climate change and California sustainability\u0026mdash;Challenges and solutions. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A.\u003c/em\u003e 121, e2405458121\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRogers CDW et al (2021) Recent Increases in Exposure to Extreme Humid-Heat Events Disproportionately Affect Populated Regions. \u003cem\u003eGeophysical Research Letters\u003c/em\u003e 48, e2021GL094183\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHong T et al (2023) Ten questions concerning thermal resilience of buildings and occupants for climate adaptation. Build Environ 244:110806\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrussa ZD et al (2019) A London residential retrofit case study: Evaluating passive mitigation methods of reducing risk to overheating through the use of solar shading combined with night-time ventilation. Build Serv Eng Res Tech 40:389\u0026ndash;408\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLomas KJ, Porritt SM (2017) Overheating in buildings: lessons from research. Building Res Inform 45:1\u0026ndash;18\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGamero-Salinas JC, Monge-Barrio A, S\u0026aacute;nchez-Ostiz A (2020) Overheating risk assessment of different dwellings during the hottest season of a warm tropical climate. Build Environ 171:106664\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGupta R, Barnfield L, Gregg M (2017) Overheating in care settings: magnitude, causes, preparedness and remedies. Building Res Inform 45:83\u0026ndash;101\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVellei M et al (2017) Overheating in vulnerable and non-vulnerable households. Building Res Inform 45:102\u0026ndash;118\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eU.S. Energy Information Administration (2018) Commercial buildings energy consumption survey (CBECS). Retrieved July 10, 2025, from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.eia.gov/consumption/commercial/\u003c/span\u003e\u003cspan address=\"https://www.eia.gov/consumption/commercial/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eU.S. Energy Information Administration (2020) Residential Energy Consumption Survey (RECS). Retrieved July 10, 2025, from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.eia.gov/consumption/residential/\u003c/span\u003e\u003cspan address=\"https://www.eia.gov/consumption/residential/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMavrogianni A, Wilkinson P, Davies M, Biddulph P, Oikonomou E (2012) Building characteristics as determinants of propensity to high indoor summer temperatures in London dwellings. Build Environ 55:117\u0026ndash;130\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaniassadi A, Sailor DJ, Krayenhoff ES, Broadbent AM, Georgescu M (2019) Passive survivability of buildings under changing urban climates across eight US cities. Environ Res Lett 14:074028\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAhmed T, Kumar P, Mottet L (2021) Natural ventilation in warm climates: The challenges of thermal comfort, heatwave resilience and indoor air quality. Renew Sustain Energy Rev 138:110669\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQuinn A et al (2014) Predicting indoor heat exposure risk during extreme heat events. Sci Total Environ 490:686\u0026ndash;693\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSherman P, Lin H, McElroy M (2022) Projected global demand for air conditioning associated with extreme heat and implications for electricity grids in poorer countries. Energy Build 268:112198\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLundgren K, Kjellstrom T (2013) Sustainability Challenges from Climate Change and Air Conditioning Use in Urban Areas. Sustainability 5:3116\u0026ndash;3128\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang C et al (2023) Impacts of climate change, population growth, and power sector decarbonization on urban building energy use. Nat Commun 14:6434\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAmonkar Y, Doss-Gollin J, Farnham DJ, Modi V, Lall U (2023) Differential effects of climate change on average and peak demand for heating and cooling across the contiguous USA. Commun Earth Environ 4:402\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eU.S. Census Bureau American Housing Survey: 2021 National Public Use File. (U.S. Department of Housing and Urban Development and U.S. Census Bureau, 2022); \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.census.gov/programs-surveys/ahs.html\u003c/span\u003e\u003cspan address=\"https://www.census.gov/programs-surveys/ahs.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHamdy M, Carlucci S, Hoes PJ, Hensen JL (2017) The impact of climate change on the overheating risk in dwellings\u0026mdash;A Dutch case study. Build Environ 122:307\u0026ndash;323\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBorgeson S, Brager G (2011) Comfort standards and variations in exceedance for mixed-mode buildings. Building Res Inform 39(2):118\u0026ndash;133\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDe Dear RJ, Brager GS (2002) Thermal comfort in naturally ventilated buildings: revisions to ASHRAE Standard 55. Energy Build 34(6):549\u0026ndash;561\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNicol JF, Humphreys MA (2002) Adaptive thermal comfort and sustainable thermal standards for buildings. Energy Build 34(6):563\u0026ndash;572\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKenny GP, Tetzlaff EJ, Journeay WS, Henderson SB, O\u0026rsquo;Connor FK (2024) Indoor overheating: a review of vulnerabilities, causes, and strategies to prevent adverse human health outcomes during extreme heat events. Temperature 11(3):203\u0026ndash;246\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiller D et al (2021) Cooling energy savings and occupant feedback in a two year retrofit evaluation of 99 automated ceiling fans staged with air conditioning. Energy Build 251:111319\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKent MG et al (2023) Energy savings and thermal comfort in a zero energy office building with fans in Singapore. Build Environ 243:110674\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAttia, S., Benzidane, C., Rahif, R., Amaripadath, D., Hamdy, M., Holzer, P., \u0026hellip; Carlucci,S. (2023). Overheating calculation methods, criteria, and indicators in European regulation for residential buildings. Energy and Buildings, 292, 113170.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnđelković AS, Mujan I, Dakić S (2016) Experimental validation of a EnergyPlus model: Application of a multi-storey naturally ventilated double skin fa\u0026ccedil;ade. Energy Build 118:27\u0026ndash;36\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMateus NM, Pinto A, Da Gra\u0026ccedil;a GC (2014) Validation of EnergyPlus thermal simulation of a double skin naturally and mechanically ventilated test cell. Energy Build 75:511\u0026ndash;522\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLoutzenhiser PG, Manz H, Felsmann C, Strachan PA, Frank TH, Maxwell GM (2007) Empirical validation of models to compute solar irradiance on inclined surfaces for building energy simulation. Sol Energy 81(2):254\u0026ndash;267\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHaves P, Ravache B, Fergadiotti A, Kohler C, J (2019) September). Accuracy of HVAC load predictions: Validation of EnergyPlus and DOE-2 using an instrumented test facility. Building Simulation 2019, vol 16. IBPSA, pp 4329\u0026ndash;4336\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSutton-Klein J, Moody A, Hamilton I, Mindell JS (2021) Associations between indoor temperature, self-rated health and socioeconomic position in a cross-sectional study of adults in England. BMJ open, 11(2), e038500\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAhmed T, Kumar P, Mottet L (2021) Natural ventilation in warm climates: The challenges of thermal comfort, heatwave resilience and indoor air quality. Renew Sustain Energy Rev 138:110669\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen J, Adhikari R, Wilson E, Robertson J, Fontanini A, Polly B, Olawale O (2022) Stochastic simulation of occupant-driven energy use in a bottom-up residential building stock model. Appl Energy 325:119890\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilson EJ, Christensen CB, Horowitz SG, Robertson JJ, Maguire JB (2017) \u003cem\u003eEnergy Efficiency Potential in the U.S. Single-Family Housing Stock\u003c/em\u003e. NREL/TP\u0026ndash;5500\u0026ndash;68670, 1414819 \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.osti.gov/servlets/purl/1414819\u003c/span\u003e\u003cspan address=\"http://www.osti.gov/servlets/purl/1414819\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e/ \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2172/1414819\u003c/span\u003e\u003cspan address=\"10.2172/1414819\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNational Renewable Energy Laboratory (2025) ResStock Public dataset. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://resstock.nrel.gov/datasets\u003c/span\u003e\u003cspan address=\"https://resstock.nrel.gov/datasets\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRodrigues E, Fernandes MS, Carvalho D (2023) Future weather generator for building performance research: An open-source morphing tool and an application. Build Environ 233:110104\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEyring V et al (2016) Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci Model Dev 9:1937\u0026ndash;1958\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSutton-Klein J, Moody A, Hamilton I, Mindell JS (2021) Associations between indoor temperature, self-rated health and socioeconomic position in a cross-sectional study of adults in England. BMJ open, 11(2), e038500\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEdwards JR, De Roos AJ, Hampo CC, Huang W, Lincoln E, Hoque S, Schinasi LH (2025) Residential indoor temperatures and health: A scoping review of observational studies. Sci Total Environ 979:179377\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCalifornia Energy Commission (2022) 2022 Building Energy Efficiency Standards for Residential and Nonresidential Buildings: For the 2022 Building Energy Efficiency Standards Title 24, Part 6, and Associated Administrative Regulations in Part 1. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.energy.ca.gov/programs-and-topics/programs/building-energy-efficiency-standards/2022-building-energy-efficiency\u003c/span\u003e\u003cspan address=\"https://www.energy.ca.gov/programs-and-topics/programs/building-energy-efficiency-standards/2022-building-energy-efficiency\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilson EJ et al (2022) End-Use Load Profiles for the U.S. Building Stock: Methodology and Results of Model Calibration, Validation, and Uncertainty Quantification. Golden, CO: National Renewable Energy Laboratory. NREL/TP-5500-80889. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2172/1854582\u003c/span\u003e\u003cspan address=\"10.2172/1854582\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReyna J et al (2022) State Level Residential Building Stock and Energy Efficiency \u0026amp; Electrification Packages Analysis. Tableau Dashboard. Golden, CO: National Renewable Energy Laboratory. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2172/1877069\u003c/span\u003e\u003cspan address=\"10.2172/1877069\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePalmgren C, Ramirez B, Goldberg M, Williamson C (2021) 2019 California Residential Appliance Saturation Study (RASS): Consultant Report. California Energy Commission. Publication Number: CEC-200-2021-005-PO\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"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":"Climate Change, Overheating Risks, Air Conditioning Retrofit, Cooling Energy Use, Building Stock Modeling","lastPublishedDoi":"10.21203/rs.3.rs-8264868/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8264868/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eAs climate change intensifies, the frequency and severity of extreme heat events are increasing, placing growing stress on building environments and energy systems. Using a physics-based residential building-stock framework, we conducted over 265,000 whole-building simulations (representing over 50,000 California homes across five climate scenarios) to quantify indoor overheating severity and the resulting increases in cooling demand. Indoor heat exposure was evaluated using the indoor overheating degree (IOD), which captures both the duration and magnitude of temperature exceedances above selected comfort limits. Results show that by the 2080s, up to ~\u0026thinsp;55% of homes without cooling may exceed safe indoor limits, driving a 20\u0026ndash;40% rise in statewide peak cooling electricity loads and potentially surpassing grid adequacy margins. Overheating risk moves from inland where AC is more common to coastal regions where it is not, increasing overall indoor heat exposure while narrowing regional disparities. By linking thermal limits with AC adoption and grid impacts, we quantify where overheating risks and cooling burdens intensify under current and future climates, providing evidence to support thermally safe and grid-aware housing policies.\u003c/p\u003e","manuscriptTitle":"Safe indoor temperature limits shape overheating resilience and cooling demand in California homes","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-17 08:51:08","doi":"10.21203/rs.3.rs-8264868/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"ae8ec2a1-ad2e-428b-b5f9-1b13a90512bc","owner":[],"postedDate":"December 17th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":59749342,"name":"Earth and environmental sciences/Environmental social sciences/Climate-change impacts"},{"id":59749343,"name":"Physical sciences/Energy science and technology/Energy modelling"},{"id":59749344,"name":"Scientific community and society/Energy and society/Energy and behaviour"},{"id":59749345,"name":"Earth and environmental sciences/Natural hazards"}],"tags":[],"updatedAt":"2026-01-27T12:57:04+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-17 08:51:08","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8264868","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8264868","identity":"rs-8264868","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}