Projected Increase in Heatwaves under 1.5 and 2.0 °C Warming Levels Will Increase the Socio-Economic Exposure across China by the Late 21st Century

Abstract

The impending challenge posed by escalating heatwave events due to projected global warming scenarios of 1.5 and 2.0 °C underscores the critical need for a comprehensive understanding of their impact on human health and socio-economic realms. This study delves into the anticipated implications of elevated global temperatures, specifically the 1.5 and 2.0 °C warming scenarios under the SSP2-4.5 and SSP5-8.5 pathways, on population and GDP exposure to heatwaves in China. We also evaluated the aggregated impacts of climate, population, and GDP and their interactions on future socio-economic exposure across China. We leveraged data sourced from the climatic output of Coupled Model Intercomparison Project Phase 6 (CMIP6) for heatwave analysis and integrated population and GDP projections under divergent Shared Socio-economic Pathways (SSPs), including SSP2-4.5 (low emission) and SSP5-8.5 (high-emission). Results indicate a drastic surge in the number of heatwave days under both warming scenarios, particularly in regions like Xinjiang (XJ), North China (NC), and South China (SC) subregions, with a notable disparity in the elevation of heatwave days among different levels. There is an alarming surge in population exposure, escalating approximately 7.94–8.70 times under the 1.5 °C warming scenario and markedly increasing by 14.48–14.75 times by the 2100s relative to the baseline (1985–2014) under the more extreme 2.0 °C warming level. Likewise, the study unveils a substantial elevation in GDP exposure, ranging from 40.65 to 47.21 times under the 1.5 °C warming level and surging dramatically by 110.85–113.99 times under the 2.0 °C warming level. Further analyses disclose that the climate effect predominantly influences changes in population exposure, constituting 72.55–79.10% of the total change. Meanwhile, the interaction effect notably shapes GDP exposure alterations, contributing 77.70–85.99% to the total change. The comprehensive investigation into alterations in population and GDP exposure under varying warming scenarios, coupled with the quantification of each contributing factor, holds paramount importance in mitigating the detrimental repercussions of heatwaves on both human life and socio-economic landscapes.

1. Introduction

Climate change has begun profoundly affecting both natural ecosystems and human societies. Projections derived from climate models strongly indicate a persistent trajectory of warming temperatures in the future, accompanied by a surge in extreme events, amplifying their substantial impacts [1]. Specifically, since the onset of the 21st century, there has been a relentless succession of heatwave events, marked by a consistent escalation in their intensity, duration, and geographical extent [2,3], causing significant implications on human life, socio-economic development, and the natural environmental systems [4,5]. The Paris Agreement [6] explicitly outlines the imperative of constraining global average warming to within 2.0 °C above pre-industrial levels, while urging concerted endeavors to cap warming at 1.5 °C. These ambitious targets stand as pivotal measures to mitigate the risks and consequences of climate change [7]. Numerous studies have scrutinized the climate change implications of global warming of 1.5 and 2.0 °C at the global and regional scales, and their findings revealed that global warming will exert profound and multifaceted impacts, particularly manifesting in heightened occurrences of extreme climatic events and significant socio-economic repercussions [7,8]. The projected extreme temperatures with escalation in frequency and intensity across China under a 1.5 °C global warming scenario are evident from previous studies [9]. The number of heatwave days above a certain threshold showed a nonlinear increase as the temperature rose [10]. Global climate change will be further exacerbated in the future, with more frequent heatwaves, larger impact areas, and longer durations [11]. This evolving scenario intertwines with socio-economic growth, amplifying the risk posed by heatwaves, posing a severe threat to human health and sustainable socio-economic development [8,12]. The exposure to climate risk is expected to expand associated with population growth and increasing GDPs under the projected socio-economic development scenarios [13,14]. Climate risk is a function of hazard, exposure, and vulnerability, while disaster risk management focuses on reducing exposure and vulnerability to extreme weather events [15]. Analyzing heatwave exposure forms the basis for comprehending their impacts on human health and socio-economic domains. The exposure degree depends not only on climate change but also on the number and size of disaster-bearing bodies. For example, higher population density regions have a larger probability of climate risks with a higher number of affected people [16]. Recent advancements in climate modeling have led to an increased focus on exploring future exposure to heatwaves. Predicting shifts in population and GDP exposure to heatwaves and delineating the contributions of each influencing factor are crucial steps toward assessing future heatwave risks. This comprehension holds a pivotal role in developing actionable science solutions to mitigate climate degradation and its associated risks.

Future climate radiative forcing scenarios and socio-economic development scenarios are the basis for studying future climate and risk projections. Climate scenarios are an important basis for climate change research, and the rationalization of socio-economic development scenarios is a key aspect of climate change impact assessment [17]. Relevant studies have been conducted to analyze heatwave exposure based on model data under climate scenarios. For example, Jones et al. [16] constructed a population exposure indicator as the product of the number of high-temperature days and the number of the population using data from SRES A2 and investigated the future changes in population exposure to extreme heat in the United States and the associated influence factors. Liu et al. [18] predicted global population exposure to heatwave events using combined SSP-RCP scenarios, assessing the influences of climatic and demographic factors. Ma et al. [19] evaluated global population exposure and contributions of climate and population factors under different SSP scenarios, utilizing CMIP6 daily temperature data and projected population figures. Further studies transitioned from population to socio-economic exposure [20,21] and analyzed global socio-economic exposure to heatwaves based on climate model data, future population, and GDP projections, fostering advancements in heatwave-related exposure studies. Despite prior efforts to evaluate heatwaves and their impact on socio-economic factors, critical knowledge gaps persist. Firstly, only a few studies have considered the 1.5 and 2.0 °C global warming scenarios in their estimates of future heatwave exposure, and there is a lack of exploration of changes in exposure and analysis of the factors influencing this scenario. This lack of consideration limits our understanding of how exposure might change and the factors driving these alterations. Additionally, previous methods for calculating population and GDP exposure to heatwaves often simplified assessments. They relied on the total number of heatwave days multiplied by population and GDP, overlooking the nuanced variations in exposure levels across different heatwave intensities. This oversimplification negates the intricate differences in population and GDP exposure levels concerning varying heatwave intensities. Hence, there is an urgent need for a comprehensive analysis that delves into the exposure of population size and GDP production to different heatwave intensities under the projected global warming scenarios of 1.5 and 2.0 °C. Addressing these gaps not only fills critical voids in current knowledge but also offers nuanced insights essential for informed decision-making in mitigating future climate risks.

China is a crucial focal point in global climate change, marked by its sensitivity and substantial susceptibility to its impacts. Heatwaves, a prevalent phenomenon, wield severe economic and social repercussions across China, influenced by demographic shifts and socio-economic growth. Projections indicate a sustained escalation in extreme heatwave occurrences throughout China under varying warming scenarios, compelling a dire need for a comprehensive regional assessment of heatwave exposure at the 1.5 and 2.0 °C global warming levels.

China stands within the realm of global climate change, marked by its sensitivity and substantial susceptibility to its impacts [22]. The heatwaves have severe economic and social impacts on China with demographic changes and socio-economic development [23]. The number of extreme heatwave days in China will continue to increase under different warming scenarios [24], and a quantitative assessment of China’s regional heatwave exposure under global warming scenarios of 1.5 and 2.0 °C is strongly required. Current research [25,26,27] on regional heatwave exposure in China focuses on the occurrence of heatwave events and assesses the extent of population exposure to heatwaves in historical times. Some studies [28,29] have examined the exposure of population to high temperatures using the results of global model simulations, which show that the exposure of population to high temperatures in both the medium- and high-emission scenarios increases to some extent in the future and is mainly influenced by climate change factors. However, most of the data used for China’s population scenarios are global-scale population projections that do not take into account the uniqueness of regional economic development and China’s different population policies, which may lead to systematic bias in the regional projection results, potentially impacting the accuracy of population exposure forecasts. Furthermore, estimations regarding GDP exposure to heatwaves in China remain limited. Addressing this gap necessitates the use of more regionally specific population and GDP data. By doing so, it becomes feasible to precisely measure how population and GDP are exposed to heatwaves across China’s various warming scenarios, considering both spatial and temporal distribution disparities.

This study utilized the outputs of global climate models (GCMs) to evaluate the projected implications of heatwaves on population and GDP in China under the 1.5 and 2.0 °C warming levels. Additionally, this research seeks to elucidate the substantial influences of climatic, demographic, and GDP-related factors on these shifts in population and GDP exposure. This study unfolds through three key components: (1) investigating the spatial and temporal variations in the number of heatwave days under the 1.5 and 2.0 °C warming levels; (2) quantifying the spatial and temporal distribution patterns and changes in population and GDP exposure to regional heatwaves in China under different emission scenarios (mainly SSP2 and SSP5) and warming levels (1.5 and 2.0 °C); and (3) exploring the aggregated impacts of climate, population, and GDP and their interactions on future socio-economic exposure. Through these comprehensive analyses, this study aims to provide valuable insights into the complex interplay among climatic, demographic, and economic factors, offering a deeper understanding of the dynamics driving population and GDP exposure to heatwaves across China and various subregions.

2. Materials and Methods

2.1. Study Area

China possesses an extensive territory with a wide range of climate types, significant geographic variability in human social factors, and pronounced regional divergence in the number of heatwave days [30]. To better explain the characteristics of the spatial distribution of population and GDP exposure to regional heatwaves in China, a geographic partitioning method used in our previous study was adopted [31]. According to this division method, China is divided into seven subregions, as shown in Figure 1a. They are Xinjiang (XJ), Qinghai–Tibetan Plateau (QTP), Northwest (NW), Northeast (NE), North China (NC), Southwest (SW), and South China (SC). Among these seven subregions, the population is largest in NE, NC, and SC (Figure 1b). NC has the highest GDP among these subregions (Figure 1c). The abbreviations for the full name can be found in

Table A1. Key abbreviations and corresponding full names presented in this study.

Classification	Abbreviation	Full Name
Place name	XJ	Xinjiang
	QTP	Qinghai–Tibetan Plateau
	NW	Northwest
	NE	Northeast
	NC	Northern China
	SW	Southwest
	SC	Southern China
Program/Mission	CMIP	Coupled Model Intercomparison Project
Factor	GDP	Gross Domestic Product
Heatwave parameters	HWI	Heatwave index
	LHWD	The number of light heatwave days
	MHWD	The number of moderate heatwave days
	SHWD	The number of severe heatwave days

in the section of Appendix A.

2.2. Datasets

2.2.1. Maximum Temperature and Relative Humidity

Daily maximum temperature (T_max) and relative humidity (RH) are taken from five CMIP6 GCMs, including IPSL-CM6A-LR, MPI-ESM1-2-HR, MRI-ESM2-0, UKESM1-0-LL, and GFDL-ESM4, for the baseline period (1984–2014) and the future (2015–2100). We used data for two emission scenarios, including SSP2-4.5 (low) and SSP5-8.5 (high). Using these meteorological parameters, the heatwave index was calculated for the 1.5 and 2.0 °C warming levels under the SSP2-4.5 and SSP5-8.5 scenarios. To unify the data resolution, the T_max and RH from all models were resampled to 0.25° × 0.25°. The detailed information of these five models can be seen in

Table 1. Information on the five models used in this study and the time series adopted to calculate the heatwave index.

Model	Nationality	Institution	Resolution	Time Series
				Variable	Baseline Period	Future Period
IPSL-CM6A-LR	France	IPSL	1.3° × 2.5°	Daily maximum near-surface air temperature and near-surface relative humidity	1984 – 2014	2015 – 2100
MPI-ESM1-2-HR	Germany	DKRZ	0.9° × 0.9°
MRI-ESM2-0	Japanese	MRI	1.12° × 1.12°
UKESM1-0-LL	Britain	NCAS	1.3° × 1.9°
GFDL-ESM4	USA	NOAA-GFDL	1° × 1.25°

.

2.2.2. Population and Gross Domestic Product (GDP)

The population and GDP data are derived from 1980–2100 global 0.5° × 0.5° grid point data based on SSP paths published by the Center for Global Environmental Research (CGER) [32]. They have a temporal resolution of 10 years, in which the 1980–2010 data are downscaled by real population and GDP data. Herein, the data for 1990, 2000, and 2010 were used to represent three time periods of the baseline: 1984–1994, 1995–2004, and 2005–2014, respectively. The projected population and GDP data are derived from the SSP demographic and economic gridded dataset developed by Jiang et al. [33]. They have a temporal resolution of 1 year and a spatial resolution of 0.5° × 0.5°. Based on the 1.5 and 2.0 °C global warming levels, this study selected population and GDP data for the SSP2 and SSP5 scenarios under the universal two-child policy. To maintain spatial consistency with the heatwave index data, the population and GDP data were resampled to a grid resolution of 0.25° × 0.25°.

2.3. Methods

2.3.1. Definition of Heatwave and Heatwave Days

There is no uniform definition and grading criteria for heatwaves in the world, and the number of heatwave days varies widely for different definitions [34]. The previous studies on heatwave exposure assessment have defined heatwaves by absolute or percentile thresholds [35,36] and the heatwave index (HWI) [37,38]. Traditional projections regarding heatwaves consider only one variable, temperature, and likely underestimate the potential impacts of future climate change [39]. Considering the combined effect of temperature–humidity can better explain the heat load of the human body in high-temperature environments and thus better characterize the intensity of heatwave events [40]. We used the combined temperature and humidity heatwave index to define heatwaves, considering the climatic backgrounds of different regions in China [41]. A previously published study has described specific calculation steps in detail [30]. The impact of heatwaves on socio-economic and human health is categorized into three levels, i.e., light, moderate, and severe (

Table 2. Ranges used for classification of the heatwave index.

Levels	Light	Moderate	Severe
Classification	2.8 ≤ HWI < 6.5	6.5 ≤ HWI < 10.5	HWI ≥ 10.5

). The number of annual heatwave days was calculated for mapping population and GDP exposure. The number of days in each year in which the heatwave index at the grid points meets the corresponding thresholds, e.g., the number of light heatwave days (LHWDs), can be obtained from Equation (1):

$$LHWD={\displaystyle \sum _{i=1}^n(2.8\le HW{I}_i<6.5)}$$

(1)

where n is the number of days for the specific year, i.e., 365 or 366.

2.3.2. Determination of Global Warming Time

According to the definition of the United Nations Framework Convention on Climate Change (UNFCCC), a temperature rise year is a year in which the global average annual temperature rises to a specific temperature relative to the pre-industrial period [42,43]. However, there is much discussion on the specific definition of temperature rise, including post-stabilization temperatures, peak temperatures, and temperatures in the transient state [44]. Meanwhile, since the IPCC Assessment Report (AR) 5 climate modeling group did not take into account modern climate temperature changes relative to pre-industrial conditions when calculating warming relative to 1986–2005, the future changes relative to the 1986–2005 model were used in both its Working Group (WG)Ⅰ and WG Ⅲ reports in addition to the superimposed changes based on the HadCRUT4 observations to the 0.61 °C warming of the modern climate relative to the pre-industrial period [45,46]. However, in practice, it has been found that the modern climate of each model does not have the same warming and observations relative to the pre-industrial period, and significant observations were lacking in the early industrial stage [46]. Therefore, five climate models, including IPSL-CM6A-LR, MPI-ESM1-2-HR, MRI-ESM2-0, UKESM1-0-LL, and GFDL-ESM4, were used, and warming levels were defined by using the temperature change of each model itself relative to the pre-industrial temperature. This is also consistent with the IPCC AR5 WGIII definition of warming.

Specifically, regarding the 1.5 and 2.0 °C temperature rises, the following calculation scheme was used in this study: (1) considering the starting year of 1861 for most of the model historical test simulations, a total of 40 years from 1861 to 1900 for each model was used as the pre-industrial time period, (2) at the same time, the range of warming for each model relative to the pre-industrial period was calculated by using the average of every 31 years thereafter as the climatic period for the central year of each corresponding time period, and (3) finally, to eliminate as much uncertainty as possible in each model, the results calculated for each model are averaged across the multimodal ensemble. According to the steps mentioned above, we obtained the detailed periods of each model for the 1.5 and 2.0 °C warming scenarios under SSP2-4.5 and SSP5-8.5, and then their ensemble average was calculated, which can be found in

Table 3. The detailed periods of the five models’ ensemble average for the 1.5 and 2.0 °C warming levels under SSP2-4.5 and SSP5-8.5.

SSPs	Warming Levels	Central Year	Period
SSP2-4.5	1.5 °C	2032	2017–2047
	2.0 °C	2051	2036–2066
SSP5-8.5	1.5 °C	2029	2015–2045
	2.0 °C	2042	2027–2057

.

2.3.3. Population and GDP Exposure to Heatwave

The population exposure is calculated by multiplying the days of heatwave events (e.g., the number of days when WHI ≥ 2.8) and the number of people residing in a specific pixel [16]. The degree of population exposure was quantified by person-days. GDP exposure is calculated using the same methodology, and its magnitude is measured quantitatively by purchasing power parity-days (ppp-days). The temporal resolution of the population and GDP data in the baseline period is 10 years, so the population and GDP data in 1990, 2000, and 2010 are multiplied by the annual average number of heatwave days in 1984–1994, 1995–2004, and 2005–2014, respectively, to obtain the population and GDP exposures in the three periods of the baseline period, and then averaged to obtain the population and GDP exposures, respectively. Similarly, the population and GDP exposures for the future SSP2-4.5 and SSP5-8.5 scenarios for global warming levels of 1.5 and 2.0 °C are obtained by averaging population and GDP exposures over the 31-year period corresponding to the warming periods. The average population and GDP exposure were calculated for each grid cell using Equations (2) and (3):

$$E_P={\displaystyle \frac{{\displaystyle \sum _{i=1}^{n}HW{D}_i\times P_i}}{n}}$$

(2)

$$E_G={\displaystyle \frac{{\displaystyle \sum _{i=1}^{n}HW{D}_i\times G_i}}{n}}$$

(3)

where E_P and E_G are the mean population (person-days) and GDP exposure (ppp-days) respectively, i represents the ith year of the study period, HWD_i denotes the annual number of heatwave days, and P_i and G_i are the population and GDP for the corresponding periods, respectively. The population and GDP derived from the future SSP2-4.5 and SSP5-8.5 scenarios were first normalized and subsequently multiplied by the corresponding annual number of heatwave days, and then averaged in an equally weighted way. It can be obtained by Equation (4):

$$E={\displaystyle \frac{{\displaystyle \sum _{i=1}^{31}\left(0.5\times P_{Ni}+0.5\times G_{Ni}\right)\times HW{D}_i}}{31}}$$

(4)

where E represents the overall population and GDP exposure, P_Ni and G_Ni denote normalized population and GDP, respectively, and HWD is heatwave days.

2.3.4. Analysis of the Relative Contribution of the Exposure Factors

The changes in heatwave-related exposure are affected by the number of heatwave days and the changes in population and GDP. To estimate the relative contributions of climate, population, and GDP factors to socio-economic exposure, the changes in population and GDP exposure were decomposed into three components by referring to previous studies [16,18], i.e., the climate effect, the population, GDP effect, and the interaction effect. Precisely, the climate effect was calculated by keeping the population or GDP constant and using the number of HWDs in each global warming scenario of 1.5 and 2.0 °C multiplied by the population or GDP in the baseline period. Similarly, the population or GDP effect was calculated by holding the climate constant while using the number of HWDs in the baseline period multiplied by the population or GDP in each global warming scenario of 1.5 and 2.0 °C. Finally, the interaction effect was calculated by the total exposure change minus the summation of climate and population or GDP effect changes. Therefore, the changes in population and GDP exposure can be expressed by Equations (4) and (5), respectively:

$$\Delta E_P=HW{D}_B\times \Delta P+P_B\times \Delta HWD+\Delta P\times \Delta HWD$$

(5)

$$\Delta E_G=HW{D}_B\times \Delta G+G_B\times \Delta HWD+\Delta G\times \Delta HWD$$

(6)

where ΔE_P and ΔE_G are the total change in exposure of population and GDP, respectively; HWD_B, P_B, and G_B are the value of HWDs, population, and GDP in the baseline period, respectively; and ΔHWD, ΔP, and ΔG are the changes in HWDs, population, and GDP between baseline and future periods, respectively. Here, we refer to HWD_B × ΔP and HWD_B × ΔG as the population and GDP effect, respectively, to P_B × ΔHWD and G_B × ΔHWD as the climate effect, and to ΔP × ΔHWD and ΔG × ΔHWD as the interaction effect. Therefore, the formula for the contribution of each factor to the change in population and GDP exposure can be expressed as Equations (7) and (8), respectively:

$$C{R}_{c\mathrm{li}}={\displaystyle \frac{P_B\times \Delta C}{\Delta E_{C\&P}}},C{R}_{pop}={\displaystyle \frac{C_B\times \Delta P}{\Delta E_{C\&P}}},C{R}_{\mathrm{int}}={\displaystyle \frac{\Delta C\times \Delta P}{\Delta E_{C\&P}}}$$

(7)

$$C{R}_{cli}={\displaystyle \frac{G_B\times \Delta C}{\Delta E_{C\&G}}},{\mathrm{CR}}_{GDP}={\displaystyle \frac{C_B\times \Delta G}{\Delta E_{C\&G}}},C{R}_{\mathrm{int}}={\displaystyle \frac{\Delta C\times \Delta G}{\Delta E_{C\&G}}}$$

(8)

where CR_cli refers to the contribution rate of the climate effect, CR_pop and CR_GDP represent the contribution rate of the population and GDP effect, respectively, and CR_int is the contribution rate of the interactive effect. ΔC, which is actually ΔHWD, stands for climate effect, while ΔE_C&P and ΔE_C&G denote the total effect caused by climate and population (C&P) and climate and GDP (C&G), respectively.

3. Results

3.1. Spatiotemporal Changes in Heatwave Days

The analysis of total heatwave days revealed a pronounced increasing trend during both the baseline and future development scenarios, notably more significant in the projected future scenarios (Figure 2a). Particularly, the SSP5-8.5 scenario showcased a significantly higher increase rate (8 days/10 years) than SSP2-4.5 (5 days/10 years). The annual average count of heatwave days across all heatwave levels exhibited a substantial rise under both the 1.5 and 2.0 °C warming scenarios, notably more pronounced in the 2.0 °C scenario (Figure 2b). Moreover, differing heatwave classes displayed varying magnitudes of change and associated uncertainties in the count of heatwave days, revealing a greater relative increase compared to the baseline period with higher heatwave classes. Specifically, projection outcomes indicated that during the 1.5 and 2.0 °C warming levels under SSP2-4.5, the annual average total heatwave days escalated to 25 and 36 days, marking a surge of 163% and 268%, respectively, compared to the baseline period. Meanwhile, under SSP5-8.5, these values surged by 183% and 283%, respectively. During the 1.5 °C warming, the increase in light, moderate, and severe heatwave days was about one, two, and five times that of the baseline period. Contrastingly, during 2.0 °C warming, these increases surged to about one, three, and nine times, respectively, relative to the baseline period.

Distinct regional variations in the spatial distribution of heatwave days were observed, accompanied by consistent patterns in changes under different warming scenarios (Figure 3). During the baseline period, XJ exhibited the highest number of heatwave days among all subregions, with most areas experiencing over 10 days and the eastern part witnessing over 20 days—significantly surpassing other regions. The NC and SC followed relatively high heatwave day counts, especially in the northern sections, exceeding 10 days in northern SC and 5 days in northern NC. Under the 1.5 and 2.0 °C warming levels, most regions—except QTP—witnessed increased heatwave days, with a notably higher surge under the 2.0 °C scenario. The increase was most pronounced in XJ, NC, and SC, notably in XJ. These regions experienced more than a twofold increase in most areas under the 1.5 °C warming scenario and over a threefold increase under the 2.0 °C warming scenario, compared to the baseline period. The different levels of heatwaves revealed an overall increasing trend under different warming scenarios across most regions, excluding areas without heatwaves or any noticeable changes (Figure 4). XJ, NC, and SC displayed larger increases, consistent with the spatial distribution changes observed in total heatwave days. However, the XJ region demonstrated distinct characteristics in the change of different heatwave levels. The southeastern part experienced a significant increase in moderate and severe heatwave days under both the 1.5 and 2.0 °C warming levels compared to the baseline period, while the count of light heatwave days decreased.

3.2. Spatiotemporal Changes in Exposure to Heatwave

3.2.1. Population Exposure

Comparing different warming scenarios to the baseline period revealed a marked increase in population exposure across all heatwave levels (Figure 5). During the baseline period, the total heatwave population exposure was 3 billion person-days. The population exposure during the 1.5 and 2.0 °C warming periods in the SSP2-4.5 and SSP5-8.5 scenarios was about 8 and 15 times and 9 and 15 times higher relative to the baseline period, respectively. Notably, population exposure during the 2.0 °C period exceeded that of the 1.5 °C warming scenario due to increased heatwave days. The trend in population exposure across different heatwave levels mirrored the total heatwave exposure, showing greater increases under the 2.0 °C warming scenario. For instance, under the 1.5 °C warming level, the population exposure to light, moderate, and severe heatwaves spiked approximately 7, 18–21, and 42–53 times, respectively, compared to the baseline period. Light heatwave population exposure accounted for the largest share of total population exposure, though the proportions varied across warming scenarios, accounting for about 75–77% in the 1.5 °C scenario and approximately 57% in the 2.0 °C scenario. These findings imply a significantly greater relative increase in population exposure to moderate and severe heatwaves under a 2.0 °C warming level. For instance, population exposure to moderate and severe heatwaves was 22% and 2% of the overall population for the 1.5 °C warming level and rose to 33% and 9% for the 2.0 °C warming level, respectively. This difference mainly arises from a more pronounced increase in moderate and severe heatwave days within the 2.0 °C warming level.

The projections under different scenarios underscore a noteworthy escalation in population exposure across the majority of regions during both the 1.5 and 2.0 °C global warming periods compared to the baseline, with a more pronounced increase observed during the 2.0 °C warming phase—excluding the QTP (Figure 6). Spatially, the patterns of population exposure, influenced by population distribution, exhibit variations somewhat different from the spatial distribution of the number of heatwave days. Regions with notably high population exposure values primarily encompass NC, SC, and the northern part of SW, witnessing substantial increases in their population exposure. During the baseline period, population exposure levels surpassed 1 million person-days in north-central NC, while certain urban areas in SC exhibited exposure exceeding 5 million person-days. Under the 1.5 °C warming scenario, most of NC experiences an increase exceeding 10 million person-days, with a significant rise of 5 million person-days in sizable portions of SC and northern SW. However, under the 2.0 °C warming scenario, the escalation in NC surpasses 20 million person-days, with parts of SC and the northern SW witnessing increases surpassing 10 million person-days. Differential warming scenarios showcase varying degrees of population exposure escalation across different heatwave levels, particularly notable in NC and SC (Figure 7). NC remains the region with the most substantial increase, particularly pronounced under the 2.0 °C warming scenario. Additionally, under the 2.0 °C warming scenario, there is a discernible increase in population exposure to severe heatwaves in the western XJ region.

3.2.2. GDP Exposure

Figure 8 illustrates a substantial increase in GDP exposure across various warming scenarios compared to the baseline period, particularly notable under the 2.0 °C warming scenario. During the baseline period, the total GDP exposure is accounted as 1.7 × 10⁴ billion ppp-days. Under the SSP2-4.5 and SSP5-8.5 scenarios for global warming at 1.5 and 2.0 °C, GDP exposure surged to approximately 41 and 111 times and 47 and 114 times higher than the baseline period, respectively. Notably, GDP exposure depicts an increasing trend under different warming levels, consistent with the total change, with a more pronounced surge during the 2.0 °C warming scenario. For instance, during moderate heatwaves in the 1.5 and 2.0 °C warmings under the SSP2-4.5 scenario, GDP exposure surged approximately 89 and 388 times higher than the baseline period. Moreover, higher heatwave levels exhibit a more prominent relative surge in GDP exposure compared to the baseline period. For instance, during the 1.5 °C warming scenario under SSP5-5.8, GDP exposure to light, moderate, and severe heatwaves escalated to 39, 114, and 310 times higher than the baseline period, respectively. The percentage of each level of heatwave GDP exposure to the total GDP exposure under different warming scenarios is consistent with that reflected in the population exposure. Initially dominated by light heatwave GDP exposure, the total GDP exposure shifts its composition as warming scenarios progress. Under the 1.5 and 2.0 °C warming scenarios, the proportion of light heatwave GDP exposure decreases, while the percentage of moderate and severe heatwave GDP exposure continues to rise.

From the baseline period, the GDP exposure shows good consistency with the GDP distribution, and the high-value exposure areas are distributed in NC and around the megacities in SC (Figure 9a). The results of future changes under different scenarios show that, except for the Tibetan Plateau, most of the remaining regions in China will have higher GDP exposure under the 1.5 and 2.0 °C scenarios. In particular, under the 2.0 °C scenario, the GDP exposure is higher than that of 1.5 °C in different regions, and the increase range will be further expanded (Figure 9b–e). Compared to the baseline period, the regions with larger increases in GDP exposure under different warming scenarios are mainly located in NC, SC, and the northern part of SW. The largest increase is in NC, with most regions increasing by more than 20 billion ppp-days under the 1.5 °C warming scenario, more than 100 times the base period; under the 2.0 °C warming scenario, most regions in NC increase by between 500 and 1000 billion ppp-days, and in the northern regions of NC, the increase exceeds 1000 billion ppp-days. Regarding the different levels of heatwave severity (Figure 10), the spatial changes in GDP exposure under different warming scenarios are consistent with the changes in overall GDP exposure compared to the baseline period. Overall, the increase in GDP exposure for different heatwave levels is greater under the 2.0 °C warming scenario, and in terms of spatial distribution, the areas of high values and the most pronounced increases are primarily located in NC.

3.2.3. Combined Population and GDP Exposure (CPGE)

Figure 11 illustrates that the hotspots of high values of CPGE under different warmings are primarily located in NC. High-level exposures (grades 3–5) encompassed 6% of the study area during the 1.5 °C warming scenario and approximately 10% during the 2.0 °C warming scenario. On the national scale, the SSP2-4.5 and SSP5-8.5 scenarios had heatwave exposures of 2311 and 2583 during the 1.5 °C warming, increasing by two times during the 2.0 °C warming. Compared with the warming of 1.5 °C, the degree of exposure and the extent of the high-value area were significantly larger in the warming of 2.0 °C scenario, with the increased high-value area primarily in NC.

For all levels of heatwave severity, both across the entire study area and within subregions, exposure significantly escalated under a 2.0 °C warming scenario in contrast to a 1.5 °C warming scenario, with higher levels experiencing relatively greater increases (Figure 12). Between the periods of 1.5 C and 2.0 C warming, the exposure to light, moderate, and severe heatwaves increased by factors of 1.5, 3, and 9, respectively, as a national average. In terms of the contribution of each level to the total heatwave exposure, light heatwave exposure constituted the majority, comprising 75% and 57% in the 1.5 and 2.0 °C warming scenarios, respectively. Notably, there is a visible shift in the 2.0 °C warming scenario, where light heatwave exposure’s contribution to total exposure decreases, while that of moderate and severe heatwave exposure increases. Across various subregions, there were notable disparities in exposure; however, NC and SC predominantly contributed to the total heatwave exposure across all levels. In particular, NC and SC contributed approximately 90% of the national heatwave exposure under varying warming scenarios for light and moderate heatwaves. Nevertheless, concerning exposure to severe heatwaves, XJ also contributes to the total exposure to a certain extent, accounting for approximately 17% during the 1.5 °C warming scenario and 6% during the 2.0 °C warming scenario.

3.3. Analysis of the Importance of Factors Affecting Exposure at Different Scales

At the national scale, changes in population exposure under different warmings are dominated by the climate effect, contributing about 75% to the total population exposure, with the population effect contributing the least (Figure 13a,

Table 4. Contributions of impact factors to changes in total population and GDP exposure at the national scale in China during the 1.5 and 2.0 °C warming periods under the SSP2-4.5 and SSP5-8.5 scenarios.

Period	Population Exposure			GDP Exposure
	Climate Effect	Population Effect	Interaction Effect	Climate Effect	GDP Effect	Interaction Effect
SSP2-4.5/1.5 °C	72.55	5.06	22.39	9.77	12.53	77.70
SSP2-4.5/2.0 °C	75.25	2.50	22.25	7.14	7.38	85.48
SSP5-8.5/1.5 °C	76.07	3.98	19.95	9.73	10.57	79.70
SSP5-8.5/2.0 °C	79.10	2.09	18.81	7.16	6.85	85.99

). In comparison to the 1.5 °C warming scenario, the 2.0 °C warming scenario witnesses an upsurge in the contribution of the climate effect, accompanied by a reduction in the contribution of the population effect, while the interaction effect remains relatively unchanged. These changes imply the sustained dominance of the climate effect in shaping population exposure, emphasizing an intensified influence on total population exposure under the 2.0 °C scenario. For each subregion, except for QTP and SW, where the interaction effect slightly surpasses the climate effect in contributing to changes in population exposure, the principal factors affecting alterations in population exposure under diverse warming scenarios in the other regions mirror those at the national level—primarily influenced by the climate effect (Figure 13b–e). Notably, in NE, the climate effect is even more dominant in all warming scenarios, while the interaction effect exhibited negative contributions. Meanwhile, aside from the reduced contributions of the climate effect in the NW and NE and heightened population effect contributions in NE under the SSP5-8.5 scenario, most regions exhibit an increasing trend in climate effect but a declining trend in population effect under the 2.0 °C scenario compared to the 1.5 °C warming scenario.

At the national scale (Figure 14a, Table 4), the change in GDP exposure is dominated by the interaction effect under different warming scenarios, with its contribution being about 79% under the 1.5 °C warming scenario and about 86% under the 2.0 °C warming scenario. Under the warming 1.5 °C scenario, the contribution of the GDP effect is slightly higher than that of the climate effect, whereas under the warming 2.0 °C scenario, the contribution of the GDP effect is lower than the climate effect. Overall, the interaction effect is the dominant factor in the change in GDP exposure under different warming scenarios, and the contribution increases under the 2.0 °C warming scenario compared to the 1.5 warming scenario. For the subregions (Figure 14b–e), the interaction effect is the dominant factor in the change in GDP exposure under different warming scenarios, and the relationship between the contribution rates of the GDP effect and climate effect varies to some extent across subregions and different warming scenarios, which is consistent with the national region. For both XJ and NC, the contribution of the GDP effect is larger than that of the climate effect under different warming scenarios, more pronounced in XJ, especially at 1.5 °C of warming, while in NE, the situation is reversed. Besides that, the GDP effect generally contributes more than the climate effect in the 1.5 °C warming scenario, and the climate effect generally contributes more than the GDP effect in the 2.0 °C warming scenario. Compared with the warming of 1.5 °C, the contribution of the impact factors under the warming of 2.0 °C scenario changes consistently across subregions, where the contribution of the interaction effect increases and the contribution of the GDP effect and climate effect decreases, except for the QTP.

4. Discussion

Within the framework of climate change and in response to increasingly frequent and intense heatwave events [47], this study provides a comprehensive analysis, quantifying population and GDP exposure associated with different levels of heatwaves across China and its subregions. Our findings underscore that during 1.5 and 2.0 °C warmings, population and GDP exposure to regional heatwaves in China significantly surpass baseline levels and are even more pronounced. These observations underscore the critical importance of curbing greenhouse gas emissions and controlling rising temperatures to mitigate the adverse impacts of heatwaves on population and economic exposure. Moreover, the amplification of population and GDP exposure multipliers under different warming scenarios, especially at higher heatwave levels, implies a heightened susceptibility to extreme heatwave events in future warming scenarios. The dominance of climate and interaction effects in shaping population and GDP exposure aligns with earlier research [28,29]. The magnitude of the relative contributions of these factors is contingent on the changes in climate, population, and GDP effects [48]. The change in the interaction effect divided by the change in the climate effect is equal to the relative change in population (ΔP/P_B) and GDP (ΔG/G_B), respectively. When P and G exceed their values in the baseline period, the contribution of the interaction effect to total change exceeds the contribution of the climate effect.

Similarly, the change in the interaction effect divided by the change in the population effect or GDP effect equals the relative change in the number of heatwave days (ΔC/C_B). When ΔC exceeds C_B by one time in the baseline period, the interaction effect contributes more to total change than the population or GDP effect. Notably, for population exposure, while the alteration in the number of heatwave days during both the 1.5 and 2.0 °C warming periods surpasses that of the baseline period, the change in population does not exceed this baseline threshold (Figure A1). As a result, the significance of the climate, interaction, and population effects decreases in order of their impact. Regarding GDP exposure, both the number of heatwave days and GDP are larger in the 1.5 and 2.0 C scenarios than in the baseline period (Figure A1). Consequently, the contribution of the interaction effect surpasses that of the climate effect and GDP effect. The hierarchy of effects descends as follows: interaction effect, GDP effect, and climate effect in 1.5 °C warming. Conversely, during the 2.0 °C warming, where the relative change in the number of heatwave days takes precedence, the hierarchy changes to interaction effect, climate effect, and GDP effect.

While this study aimed for scientific rigor and accuracy, certain limitations and uncertainties persist. Firstly, future regional climate change studies would benefit from bolstering the application of dynamical downscaling ensemble projection methods, particularly concerning extreme climate events. A critical step forward involves establishing a dynamic simulation model for socio-economic development influenced by policies, integrating regional historical characteristics, resource conditions, and policy impacts comprehensively within the prediction and simulation process. It is essential to note that the inclusion of the population under fortified protection in the calculation of population exposure introduces an error, overestimating the actual population exposure. Furthermore, the analysis of the population effect on changes in population exposure was solely based on alterations in population size, disregarding the effects of population redistribution [49]. Future population shifts due to urbanization and evolving land use patterns could lead to intricate changes in population exposure, potentially exposing more individuals to high heatwave days and exacerbating population exposure.

While this study enriches our understanding by grading heatwaves based on the heatwave index definition, providing a more nuanced analysis of heatwave exposure at varying intensity levels and projecting future changes in population and GDP exposure to moderate and severe heatwaves, it remains incomplete due to the complexity of heatwave exposure prediction. Heatwave exposure is influenced by numerous factors, and future studies should delve deeper into aspects like heatwave duration to more accurately assess their potential impacts on population and GDP exposure. Moreover, the notion of heatwave exposure encompasses multiple facets of disaster impact, each with its complexities. While previous studies have predominantly focused on changes in population exposure, our study examines both population and GDP exposure under distinct warming scenarios, amalgamating their exposure. Yet, these variables alone cannot fully encapsulate societal development. To enhance the accuracy of heatwave exposure predictions, future research should endeavor to integrate various disaster-bearing factors, considering vulnerability and socio-economic disparities among different groups.

Notably, prior discussions on future heatwave exposure often segmented the 21st century into pre-, mid-, and post-periods, offering limited insights into global warming concerning the 1.5 and 2.0 °C scenarios. The present study quantifies the spatial and temporal variations of heatwave exposure under the global warming scenarios of 1.5 and 2.0 °C. However, while our comparative analysis primarily centers on the baseline period, a more comprehensive examination of exposure changes resulting from an additional 0.5 °C is warranted. Future research should explore the ramifications of this increment in greater detail.

5. Conclusions

This investigation meticulously projected population and GDP exposure to heatwaves amid global warming scenarios of 1.5 and 2.0 °C under the SSP2-4.5 and SSP5-8.5 frameworks. Through an analysis of the climate effect, population effect, and GDP effect at both national and subregional scales, several key conclusions emerge: (1) Across China, during the 1.5 and 2.0 °C warming scenarios, annual average heatwave days surged by approximately 170% and 278%, respectively, compared to the baseline period. Higher heatwave levels experienced a more pronounced relative increase, barring the QTP, with XJ, NC, and SC manifesting the most substantial upticks in heatwave days. (2) Population and GDP exposures to regional heatwaves in China increased significantly under different warming scenarios, with population exposures under the 1.5 and 2.0 °C scenarios about 8 and 15 times higher than those in the baseline period, while GDP exposures were about 44 and 112 times higher than those in the baseline period. (3) For diverse heatwave intensities, there was a shift in percentage contributions to exposure. Light heatwave conditions saw a decrease in population and GDP exposure percentages, whereas moderate and severe conditions witnessed an increase under both warming scenarios. Regions showing heightened exposure, especially prominent in NC and SC, experienced greater elevations in the 2.0 °C scenario compared to the 1.5 °C scenario. (4) Spatially and temporally, changes in population and GDP exposure align closely, concentrating mainly in NC. The 2.0 °C warming scenario indicated a substantial two times higher exposure compared to the 1.5 °C scenario, with the higher heatwave level showing amplified multipliers. (5) At the national and subregional scales, the climate effect was the primary contributor (76%) to population exposure changes, particularly dominant in the 2.0 °C scenario, accompanied by a decreased contribution of the population effect. Contrastingly, the interaction effect predominantly drove GDP exposure changes (82%), significantly amplified in the 2.0 °C warming scenario. The interplay between GDP and climate effects varied across warming scenarios, with the GDP effect generally outweighing the climate effect in the 1.5 °C scenario but the reverse in the 2.0 °C scenario regarding GDP exposure alterations.

The findings of this study will guide future research on the mechanisms inherent in changes in heatwave exposure in analyses that reduce the risk of heatwave exposure by controlling for climate, population, and GDP changes.