Spatio-temporal Variation Characteristics of Integrated Temperatures of ≥10 ℃ on the Loess Plateau under Global Climate Warming

AN Bin¹,²,³, CHEN Wenjing⁴, XIAO Weiwei²,³
¹School of Chemistry & Environment, Ankang University, Ankang 725000, Shaanxi, China
²Shaannan Eco-economy Research Center, Ankang 725000, Shaanxi, China
³Research Center for Rural Revitalization in Southern Shaanxi, Ankang 725000, Shaanxi, China
⁴School of Economics & Management, Ankang University, Ankang 725000, Shaanxi, China

Abstract

The Loess Plateau (LP) in China is highly sensitive to climate change, making it an ideal region for understanding temperature dynamics under global warming. This study analyzed the spatio-temporal variations of integrated temperature indicators for ≥10 ℃ and ≥0 ℃ thresholds—including the first date (FD), ending date (ED), duration (DD), and active integrated temperature (AIT)—using daily average temperature data from 55 meteorological stations on the LP spanning 1960 to 2019. Methods such as linear fitting, mutation tests, and dominance analysis were employed. The results indicate that, from 1960 to 2019, the indicators for both thresholds changed synchronously, with an advancing FD, a delayed ED, a prolonged DD, and an increasing AIT (P<0.01). Notably, most interdecadal shifts occurred in the 1990s, with abrupt changes concentrated from the late 1990s to the early 2000s. The spatial distribution of mean values for both thresholds was similar, showing that FD advanced, ED delayed, DD prolonged, and AIT increased from northwest to southeast. However, the spatial trends differed: the magnitude of the ED delay followed an east-west pattern with alternating "higher-lower-higher" phases, while the increase in AIT was higher in the east and lower in the west. For the ≥10 ℃ threshold, changes in FD, ED, and DD were influenced jointly by latitude and altitude, whereas changes in the other indicators were mainly driven by altitude, with contribution rates between 65.59% and 72.17%. The contribution of FD changes to DD changes was 65.1% for ≥10 ℃ and 68.4% for ≥0 ℃. Compared with 1960–1989, most indicators—except DD and AIT for ≥0 ℃—showed significant shifts (in terms of earlier or delayed timing, extended duration, or increased magnitude) during 1990–2019, with more pronounced changes at the ≥10 ℃ threshold. Furthermore, the contribution of FD change to DD change decreased by 2.3% for ≥10 ℃ and increased by 15.2% for ≥0 ℃, each exhibiting opposite spatial distribution patterns. Spatially, the variation in contribution rates exhibited a "higher-lower-higher" pattern along the south-to-west axis and a contrast with higher values in the southeast and lower in the northwest. Overall, the integrated temperature indicators for both thresholds on the LP show significant responses to climate warming, with distinct regional and temporal characteristics.

Keywords: integrated temperature; dominance analysis; contribution rate; spatio-temporal variation; the Loess Plateau

1. Data and Methods

1.1 Data Sources and Study Area

The daily average temperature data used in this study were obtained from the China Ground Climate Data Daily Value Dataset (V3.0), which has undergone quality control checks including station extreme values, climate boundary values, and internal consistency verification, ensuring high accuracy. Based on the principle of less than 10% missing data within a year and maximum observation length, we established average temperature series for 55 representative meteorological stations across the Loess Plateau (Fig. 1 [FIGURE:1]). Following the research of Liu et al. [31], the Loess Plateau was divided into three eco-geographical sub-regions: the East Asian Monsoon Eco-region (26 stations), the Northwest Arid Eco-region (20 stations), and the Qinghai-Tibet Plateau Eco-region (9 stations).

1.2 Integrated Temperature Indicators

The integrated temperature indicators analyzed in this study include the first date (FD), ending date (ED), duration (DD), and active integrated temperature (AIT) for ≥10 ℃ and ≥0 ℃ thresholds. The computational principles for these indicators follow established methods in the literature [9-10,14-16]. To facilitate comparison of spatio-temporal variation characteristics, the study period was divided into two sub-periods: 1960–1989 and 1990–2019.

1.3 Research Methods

Linear fitting was employed to calculate the variation trends of integrated temperature indicators on the Loess Plateau, with significance testing conducted at three levels: non-significant (P>0.05), significant (P<0.05), and highly significant (P<0.01). The Mann-Kendall mutation test [32] and sliding T-test were used to identify abrupt characteristics of each indicator. Dominance analysis [33] was applied to determine the primary geographical factors influencing ≥10 ℃ integrated temperature indicators and to calculate the contribution rates (C) of FD and ED changes to DD changes, following the computational principles described in the literature [33]. Origin 2021 and ArcGIS 10.8 were used to produce temporal variation charts and spatial distribution maps for all integrated temperature indicators.

2. Results and Analysis

2.1 Temporal Variation Characteristics of ≥10 ℃ Integrated Temperature Indicators

2.1.1 Interannual Variation

Overall, from 1960 to 2019, the average temperature on the Loess Plateau increased at a rate of 0.031 ℃·a⁻¹ (P<0.01), exceeding the national warming rate of 0.022 ℃·a⁻¹. Against this warming background, the FD for both ≥10 ℃ and ≥0 ℃ thresholds showed highly significant advancing trends at rates of 0.186 d·a⁻¹ and 0.158 d·a⁻¹, respectively (P<0.01), fluctuating between early April and late April. The ED exhibited delaying trends at rates of 0.130 d·a⁻¹ and 0.128 d·a⁻¹ (P<0.01), varying between late October and mid-November. Consequently, the DD for both thresholds showed significant prolongation at rates of 0.288 d·a⁻¹ and 0.314 d·a⁻¹ (P<0.01). The AIT also showed highly significant increasing trends of 7.50 ℃·d·a⁻¹ and 7.55 ℃·d·a⁻¹ (P<0.01). Compared with the ≥0 ℃ threshold, the ≥10 ℃ threshold showed more pronounced changes in all indicators except FD.

2.1.2 Interdecadal Variation

Interdecadal analysis reveals that the FD for ≥10 ℃ was delayed during the 1960s–1980s but shifted to earlier dates after the 1990s, with the most significant advancement in the 2010s (4.39 days earlier). The ED was delayed during the 1960s–1970s, advanced in the 1980s, and then delayed again after the 1990s, with the most pronounced delay in the 2010s (5.89 days). The DD and AIT showed patterns similar to the ED, with the highest values in the 2010s. For the ≥0 ℃ threshold, the interdecadal patterns were similar but phase transitions occurred one decade later than for ≥10 ℃. The mutation times for FD (1990s) were earlier than for ED (2000s), while DD and AIT showed the earliest mutation times, indicating a lag effect of temperature on phenology [35].

2.1.3 Abrupt Characteristics

All integrated temperature indicators on the Loess Plateau exhibited significant abrupt changes concentrated between the late 1990s and early 2000s (Table 2 [TABLE:2]), slightly later than the average temperature mutation. Compared with pre-mutation periods, post-mutation periods showed FD advancing by 4.63–5.06 days, ED delaying by 6.75–6.91 days, DD prolonging by 11.12–12.96 days, and AIT increasing by 296.39–340.23 ℃·d.

2.2 Spatial Variation Characteristics of ≥10 ℃ Integrated Temperature Indicators

2.2.1 Spatial Patterns of FD, ED, and DD

The multiyear average FD and ED for ≥10 ℃ integrated temperature showed consistent spatial patterns, gradually advancing and delaying from northwest to southeast (Fig. 3 [FIGURE:3]). The earliest FD occurred at Wugong Station in the Fenwei Plain (March 27), while the latest occurred at Huajialing Station in the central Gansu cold mountainous area (April 25). The earliest and latest ED were at Huajialing Station (September 25) and Sanmenxia Station (November 19), respectively. The average DD showed a south-long, north-short pattern, ranging from 133.1 days at Huajialing Station to 216 days at Sanmenxia Station.

Spatially, 98.28% of stations showed advancing trends for ≥10 ℃ FD, with significant (P<0.05) and highly significant (P<0.01) trends at 39 and 51 stations, respectively. The advancement magnitude showed a "higher-lower-higher" pattern along the southeast-northwest direction, with the greatest advancement at Linfen Station (−0.338 d·a⁻¹). For ED, 94.5% of stations showed delaying trends, with 58 stations reaching highly significant levels. The delay magnitude followed an east-west "higher-lower-higher" pattern, with the most pronounced delays at Tongren Station (0.262 d·a⁻¹) and Yuanping Station (0.320 d·a⁻¹). The DD prolongation trends were highly significant (P<0.01) at 58 stations, showing a west-high, east-low pattern, with the greatest extension at Tongren Station (0.601 d·a⁻¹).

2.2.2 Spatial Patterns of Active Integrated Temperature

The multiyear average AIT for ≥10 ℃ and ≥0 ℃ thresholds ranged from 2,160.2–5,226.2 ℃·d and 2,160.2–5,226.2 ℃·d, respectively, showing a west-high, east-low pattern. High-value centers formed in the southeastern Fenwei Plain, while low-value areas occurred in the southwestern Qinghai-Tibet Plateau eco-region. Most stations (94.5%) showed highly significant increasing trends (P<0.01), with the greatest increases at Yushe Station (12.92 ℃·d·a⁻¹) and Yuanping Station (13.41 ℃·d·a⁻¹). Compared with 1960–1989, stations during 1990–2019 showed AIT increases of 56.98–392.50 ℃·d for ≥10 ℃ and 44.42–421.60 ℃·d for ≥0 ℃, with the most pronounced increases at Yuanping Station. The spatial distribution of AIT increase was similar to its trend pattern, showing an east-high, west-low pattern, with the 250–300 ℃·d range being most extensive.

3. Discussion

3.1 Impacts of Integrated Temperature Indicator Changes on the Loess Plateau

Under global warming, integrated temperature indicators across China [11-13,15] and various regions [9-10,14-16] consistently show advancing FD, delayed ED, prolonged DD, and increased AIT, leading to significant phenological changes. Our findings align with these trends, though the Loess Plateau shows distinct characteristics. The FD advancement rate for ≥10 ℃ (−0.186 d·a⁻¹) was notably higher than the national average (−0.078 d·a⁻¹) [11], while the ED delay rate (0.130 d·a⁻¹) was comparable to national levels. The DD prolongation rates (0.288–0.314 d·a⁻¹) exceeded the national average but were lower than those on the Tibetan Plateau [14]. The AIT increase rates (7.50–7.55 ℃·d·a⁻¹) approached the national average, indicating a significant response to global warming.

These changes have important agricultural implications. The advancing FD and delayed ED extend the growing season for thermophilic crops and delay winter planting for cool-season crops [22], potentially increasing multiple cropping indices [37]. However, they may also expand the occurrence of crop pests and diseases [39], increasing agricultural costs. During the global warming hiatus (1998–2013), the Loess Plateau experienced a slight temperature decrease (−0.049 ℃·a⁻¹, P>0.05), causing integrated temperature indicators to show delayed FD, shortened DD, and reduced AIT, demonstrating the region's sensitivity to global climate fluctuations.

3.2 Relationship Between Integrated Temperature Indicators and Geographical Factors

Meteorological elements exhibit significant regional variations related to latitude, longitude, and altitude [28]. Correlation analysis (Table 5 [TABLE:5]) shows that ≥10 ℃ FD is significantly negatively correlated with latitude and significantly positively correlated with longitude (α=0.001), while ED and DD are significantly negatively correlated with latitude (α=0.001). This indicates that lower latitudes lead to earlier FD, later ED, and longer DD. Except for FD, all other indicators are positively correlated with longitude, reaching highly significant levels (α=0.01), suggesting that eastern locations experience later ED and longer DD. All indicators are highly significantly correlated with altitude (α=0.001), consistent with findings from Tibet [14].

Sub-regional analysis reveals distinct differences. The East Asian Monsoon sub-region shows the earliest FD, latest ED, longest DD, and highest AIT for ≥10 ℃, while the Northwest Arid and Qinghai-Tibet Plateau sub-regions show progressively later FD, earlier ED, shorter DD, and lower AIT. Dominance analysis indicates that altitude is the primary factor influencing integrated temperature changes, with contribution rates of 65.59%–72.17%, higher than the 55.60%–58.90% reported for the Tibetan Plateau [14]. Evidence of elevation-dependent warming in Northwest China [41], particularly pronounced at 1000–2000 m on the Loess Plateau, explains altitude's dominant role. Human activities, including greenhouse gas emissions and urbanization [16,43], also contribute significantly and warrant further quantification.

3.3 Influence of First Date and Ending Date Changes on Duration

Dominance analysis reveals that changes in FD, ED, and DD for ≥10 ℃ integrated temperature are jointly influenced by latitude and altitude, while changes in other indicators are primarily altitude-driven. The contribution rate of FD change to DD change was 65.1% for ≥10 ℃ and 68.4% for ≥0 ℃ during 1960–2019. This contribution decreased slightly to 57.4% for ≥10 ℃ but increased substantially to 67.3% for ≥0 ℃ during 1990–2019, indicating that DD prolongation is mainly driven by significant FD advancement.

Spatially, the contribution rate of FD change to DD change for ≥10 ℃ shows a decreasing trend from east to west, exceeding 65.0% in the northeastern region. The pattern is "higher-lower-higher" along the south-north axis, with the highest values in the eastern plateau and Fen River valley, reaching 81.6% at Yuncheng Station. Some stations (Yushe, Guyuan, Xiji) showed negative contributions (−31.9% to −24.3%), indicating that DD changes at these locations were primarily influenced by ED changes. Compared with 1960–1989, the contribution rate during 1990–2019 showed a "higher-lower-higher" pattern from south to north and a southeast-high, northwest-low pattern, highlighting the complex regional and temporal dynamics of these relationships.

4. Conclusion

Based on daily average temperature data from 55 meteorological stations during 1960–2019, this study employed linear fitting, mutation testing, dominance analysis, and spatial visualization to analyze the spatio-temporal variation characteristics of ≥10 ℃ and ≥0 ℃ integrated temperature indicators on the Loess Plateau. The main conclusions are:

1. Temporal trends: From 1960–2019, both thresholds showed synchronous changes with FD advancing at 0.186 d·a⁻¹ and 0.158 d·a⁻¹, ED delaying at 0.130 d·a⁻¹ and 0.128 d·a⁻¹, DD prolonging at 0.288 d·a⁻¹ and 0.314 d·a⁻¹, and AIT increasing at 7.50 ℃·d·a⁻¹ and 7.55 ℃·d·a⁻¹ (all P<0.01). Interdecadal shifts mainly occurred in the 1990s, with abrupt changes concentrated in the late 1990s to early 2000s.

2. Spatial patterns: Mean values for both thresholds showed consistent distributions, with FD advancing, ED delaying, DD prolonging, and AIT increasing from northwest to southeast. Trend magnitudes varied spatially: FD advancement showed a central-high, peripheral-low pattern; ED delay followed an east-west "higher-lower-higher" pattern; DD prolongation was west-high, east-low; and AIT increase was east-high, west-low.

3. Geographical influences: Changes in FD, ED, and DD for ≥10 ℃ were jointly controlled by latitude and altitude, while other indicators were primarily altitude-driven (contribution rates 65.59%–72.17%). The contribution of FD change to DD change was 65.1% for ≥10 ℃ and 68.4% for ≥0 ℃, showing a decreasing trend from east to west and a "higher-lower-higher" pattern from south to north.

4. Period comparisons: Compared with 1960–1989, most indicators during 1990–2019 showed significant shifts, with more pronounced changes at the ≥10 ℃ threshold. The contribution of FD change to DD change decreased by 2.3% for ≥10 ℃ but increased by 15.2% for ≥0 ℃, with opposite spatial distribution patterns between the two periods.

Overall, integrated temperature indicators on the Loess Plateau respond significantly to climate warming, exhibiting distinct regional and temporal characteristics that have important implications for agricultural planning and climate risk management.

References

[1] CMA Climate Change Centre. Blue Book on Climate Change in China (2023) [M]. Beijing: Science Press, 2023.

[2] Wang Yan, Wang Hao, Cui Peng, et al. Disaster effects of climate change and the associated scientific challenges[J]. Chinese Science Bulletin, 2024, 69(2): 286-300.

[3] IPCC. Climate Change 2021: The Physical Science Basis[C]//Lee J Y, Marotzke J, Bala G, et al. Future global climate: Scenario-based projections and near term information. Cambridge: Cambridge University Press, 2021.

[4] Du Jun, Huang Zhicheng, Tsewangthondup, et al. Impact of climate warming on the threshold temperature in Xizang under global climate change[J]. Chinese Journal of Agrometeorology, 2024, 45(2): 111-123.

[5] Bao Guang, Qu Yong, Liu Na, et al. Accumulated temperature variation characteristics of the boundary temperature ≥10 ℃ in Yan'an[J]. Journal of Earth Environment, 2019, 10(2): 149-155.

[6] Bai Lei, Zhang Fan, Shang Ming, et al. Evolution of the multiple accumulated temperature across China's mainland in 1961-2018 with the gridded meteorological dataset[J]. Journal of Geo-information Science, 2021, 23(8): 1446-1460.

[7] Kong Feng, Wang Yifei. Temporal and spatial evolution and temporal difference of positive accumulated temperature in China[J]. Journal of Arid Land Resources and Environment, 2021, 35(11): 103-110.

[8] Wang Zhiying, Zang Shuying, Zhou Daowei, et al. Temporal and spatial variation of agricultural threshold temperature from 1957 to 2016[J]. Scientia Geographica Sinica, 2020, 40(1): 137-148.

[9] Zhang Shanhong, Bai Yi, Song Lianhuan, et al. Assessing the boundary between subtropical and warm temperate zones and the spatio-temporal evolution of the active accumulated temperature of ≥10 ℃ in the Qinling Mountains from 1960 to 2019[J]. Journal of Ecology and Rural Environment, 2022, 38(6): 755-764.

[10] He Huayun, Hu Qi, Tang Shuyue, et al. Analysis of spatio-temporal evolution of the boundary temperature of thermophilic crops in China's Mainland based on site data[J]. Chinese Journal of Agrometeorology, 2023, 44(2): 85-95.

[11] Tao F, Zhang S, Zhang Z. Spatiotemporal changes of wheat phenology in China under the effects of temperature, day length and cultivar thermal characteristics[J]. European Journal of Agronomy, 2012, 43: 201-212.

[12] Liu Y J, Zhou W M, Ge Q S. Spatiotemporal changes of rice phenology in China under climate change from 1981 to 2010[J]. Climatic Change, 2019, 157(2): 261-277.

[13] Lü Tong, Guo Qian, Ding Yongxia, et al. Suitable planting areas of wheat and maize under climate changes over China[J]. Journal of Lanzhou University (Natural Sciences), 2022, 58(5): 588-599.

[14] Shi Guangxun, Ding Mingjun. Spatial and temporal variation of ≥10 ℃ accumulated temperature in the Yangtze River Basin for the past 40 years[J]. Tropical Geography, 2016, 36(4): 682-691.

[15] Wang Guoqiang, Zhang Bo, Zhang Yaozong, et al. Variation tendency of accumulated temperature in Gansu Province in recent 55 years[J]. Research of Soil and Water Conservation, 2016, 23(5): 193-198.

[16] Sun Yijie, Liu Xianfeng, Ren Zhiyuan, et al. Spatiotemporal variations of multi-scale drought and its influencing factors across the Loess Plateau from 1960 to 2016[J]. Geographical Research, 2019, 38(7): 1820-1832.

[17] An Bin, Xiao Weiwei, Zhu Ni, et al. Temporal and spatial variations of precipitation concentration degree and precipitation concentration period on the Loess Plateau from 1960 to 2019[J]. Arid Zone Research, 2022, 39(5): 1333-1344.

[18] Zhang Ruixiang, Lyu Yihe, Fu Bojie, et al. The effects of vegetation coverage changes on ecosystem service and their threshold in the Loess Plateau[J]. Acta Geographica Sinica, 2020, 75(5): 949-960.

[19] IPCC. Climate Change 2022: Impacts, Adaptation, and Vulnerability[R]. Cambridge, UK and New York, USA: Cambridge University Press, 2022.

[20] Jiang Tong, Zhai Jianqing, Luo Yong, et al. Understandings of assessment reports on climate change impacts, adaptation and vulnerability: Progress from IPCC AR5 to AR6[J]. Transactions of Atmospheric Sciences, 2022, 45(4): 502-511.

[21] IPCC AR5. Climate Change 2013: The Physical Science Basis[R]. Cambridge: Cambridge University Press, 2013.

[22] He G, Wang Z H, Shen J B, et al. Transformation of agriculture on the Loess Plateau of China towards green development[J]. Frontiers of Agricultural Science & Engineering, 2021, 8(4): 491-500.

[23] Xu Gang, Ren Jizhou, Feng Qisheng, et al. Origin of agriculture in Loess Plateau[J]. Pratacultural Science, 2015, 32(10): 1695-1701.

[24] Wang Xiwen, Wang Yirong. Sensitivity of total temperature change in China Loess Plateau[J]. Arid Land Geography, 2006, 29(6): 817-822.

[25] Deng Chenhui, Bai Hongying, Gao Shan, et al. Comprehensive effect of climatic factors on plant phenology in Qinling Mountains region during 1964-2015[J]. Acta Geographica Sinica, 2018, 73(5): 917-931.

[26] Yao Yubi, Wang Runyuan, Yang Jinghu, et al. Impacts of climatic change on spring wheat growth in an arid region of the Loess Plateau: A case study in Dingxi, Gansu Province[J]. Acta Ecologica Sinica, 2011, 31(15): 4225-4234.

[27] Zhang Weimin, Wang Jinghong. Variation characteristics of the accumulated temperature of Jujube planting zones in the hilly region of the Loess Plateau over the past 40 years[J]. Research of Soil and Water Conservation, 2016, 23(6): 232-237.

[28] Fu Y S H, Zhao H F, Piao S L, et al. Declining global warming effects on the phenology of spring leaf unfolding[J]. Nature, 2015, 526: 104-107.

[29] Zhang Zhigao, Zhang Xiuli, Jia Mengwei, et al. Spatio-temporal evolution of agricultural thermal resources in the Yellow River Basin during 1960-2020[J]. Journal of Shanxi Agricultural University (Natural Science Edition), 2024, 44(3): 119-130.

[30] Liu Pan, Zhao Xining, Gao Xiaodong, et al. Characteristics of extreme temperature variation in the Loess Plateau and its correlation with average temperature[J]. Chinese Journal of Applied Ecology, 2022, 33(7): 1975-1982.

[31] Liu Chuang, Shi Kun. GIS dataset of boundaries among four geo-eco regions of China[J]. Journal of Global Change Data & Discovery, 2018, 2(1): 42-50.

[32] Budescu D V. Dominance analysis: A new approach to the problem of relative importance of predictors in multiple regression[J]. Psychological Bulletin, 1993, 114: 542-551.

[33] Tang Qiyi. DPS Data Processing System: Experimental Design, Statistical Analysis and Data Mining[M]. 2nd ed. Beijing: Science Press, 2010.

[34] Chen Tingting, Yu Wenjun, Li Yanzhong, et al. The spatio-temporal changes and risk analysis of apparent temperature in China from 1960 to 2019[J]. Climate Change Research, 2024, 20(3): 265-277.

[35] Zhang Yaozong, Zhang Bo, Liu Yanyan, et al. Response of variability of temperature in the Loess Plateau to the Hiatus in the process of global warming in the period from 1960 and 2013[J]. Research of Soil and Water Conservation, 2020, 27(4): 213-219.

[36] Yan Libin. Characteristics of temperature and precipitation on the Loess Plateau from 1961 to 2014[J]. Journal of Earth Environment, 2015, 6(5): 276-282.

[37] Li Yu, Qiu Bingwen, He Yuhua, et al. Cropping intensity based on MODIS data in China during 2010-2018[J]. Progress in Geography, 2020, 39(11): 1874-1883.

[38] Hao Ruichao, Wang Qibao, Wang Chen, et al. Alterations in drying and wetting spells and their determinants in the Loess Plateau under the influence of climate change[J]. Journal of Irrigation and Drainage, 2024, 43(12): 74-83.

[39] Wang C Z, Wang X H, Jin Z L, et al. Large increase in crop pests and diseases over China since 1970[J]. Nature Food, 2022, 3: 57-65.

[40] Ding Yihui, Liu Yanju, Xu Ying, et al. Regional responses to global climate change: Progress and prospects for trend, causes, and projection of climatic warm-wetting in Northwest China[J]. Advances in Earth Science, 2023, 38(6): 551-562.

[41] Li Wen, Hou Litao, Zhang Shiqiang. Study on the relationship between elevation dependent warming and snowfall/rainfall ratio in Northwest China from 1980 to 2019[J]. Geographical Research, 2023, 42(7): 1957-1970.

[42] Liu Kai, Nie Gege, Zhang Sen. Study on the spatio-temporal evolution of temperature and precipitation in China from 1951 to 2018[J]. Advances in Earth Science, 2020, 35(11): 1113-1126.

[43] Li Yu, Zhou Decheng, Yan Zhangmei, et al. Contribution of urbanization to local warming in major cities of China[J]. Environmental Science, 2022, 43(5): 2822-2830.