Variations of Extreme Temperature Event Indices in Six Temperature Zones in China from 1961 to 2020

Abstract

In this study, eight extreme temperature event indices were calculated based on daily maximum, minimum, and mean temperature data recorded at 699 National Reference Stations in China during 1961–2020. The yearly change of mean temperature and the magnitude, frequency, and duration of extreme temperature events in six temperature zones were evaluated. All temperature zones had a trend of an increase in mean temperature (rate: 2.1–3.3 °C per 10 years), and the warming was more significant in the warm temperate zone and the Qinghai–Tibet Plateau zone (QPZ). For extreme temperature events, the extreme maximum and minimum temperatures in most temperature zones showed significant trends of increase, and the rates of increase were greater in the northern zones and QPZ. The rate of increase in extreme minimum temperature was substantially (up to three times) higher than the rate of increase in extreme maximum temperature in the same temperature zone; however, the finding was the opposite for the cold temperate zone (CTZ), which is the northernmost region of China. The rate of increase in extreme maximum temperatures was the greatest (0.35 °C per 10 years), whereas the rate of increase in extreme minimum temperatures was the smallest (0.17 °C per 10 years). The number of warm days/nights and the warm spell duration index also showed significant trends of increase that were most obvious in the southern zones and QPZ. In the tropical zone (TZ), which is the southernmost part of mainland China, the number of warm nights was only 15.3 days in 1961–1970, whereas it increased to 61.9 days in 2011–2020 (an increase of 303.9%). The rate of increase in warm nights in TZ (8.8 days per 10 years) was four times that in CTZ (2.2 days per 10 years). The number of cold days/nights and the cold spell duration index showed significant trends of decrease, with the greatest rates of reduction in QPZ and TZ. In evaluating the frequency of extreme temperature events, the amplitude of warming of the night index was found to be greater than that of the day index. In evaluating the duration of extreme temperature events, the variation of the cold index was found to be greater than that of the warm index. The notable asymmetries found in the variations of the minimum/maximum temperatures, day/night indices, and cold/warm spell durations in China are direct manifestations of global warming.

1. Introduction

Global warming has attracted widespread attention [1,2,3]. According to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [4], the 50 years since 1970 have been the warmest of the previous 2000 years. The global mean surface temperature was 1.1 °C higher in 2011–2020 than in 1850–1900, and, between 2003 and 2012, the temperature increased by 0.2 °C. Global warming and associated climatic changes will very likely intensify over the coming decades. Based on simulation results of global climate models, as the global average temperature increases, extreme temperature events in most regions are projected to increase and become enhanced on both daily and seasonal scales [5,6,7]. Extreme temperature events, which are caused by natural factors, urbanization, and deforestation [8,9], are also likely to become more frequent and, consequently, pose severe threats to human lives and property [10].

To formulate unified definitions and standards for extreme temperature event indicators and thresholds, the World Meteorological Organization and the World Climate Research Programme jointly established the Expert Team on Climate Change Detection and Indices (ETCCDI) in the early 21st century. They defined 27 representative climate indices and promoted research on extreme climate change. Many in-depth studies on extreme temperature events have been based on ETCCDI data. Based on observational data recorded at meteorological stations in 50 major urban areas of the United States during 1961–2010, Habeeb et al. [11] found that the frequency (+0.6 times per 10 years), intensity (+0.1 °C), and duration (+6 d) of extreme temperature events in the United States are all increasing. The Hadley Centre Global Historical Climatology Network Daily dataset also shows a significant increase in extreme temperature events in most regions of the world [12]. According to Zhai and Pan [13], the number of hot/cold days in China had a slight negative/positive trend during 1951–1999. However, the frequency of warm nights has increased significantly over most parts of the country, and the frequency of cool nights has decreased over almost all of China. Zhou and Ren [14] analyzed the spatial and temporal changes of extreme temperature events in mainland China between 1961 and 2008. They reported a considerable reduction in the values of extreme cold indices after the mid-1980s and an increase in the values of extreme warm indices after the mid-1990s. Kong [15] analyzed extreme climatic events in China between 1961 and 2018, and found that the temperature thresholds of cold and warm days and nights were higher in southern parts of China than in northern areas; the characteristics of the indices also varied depending on location. Under Global Warming Scenarios of temperature increase of 1.5 and 2.0 °C, extreme maximum temperature events in China would continue to increase to varying degrees [16]. Changes in extreme temperature are an important aspect of climate change because their impact on natural ecosystems and human society can be more profound than that of the change in mean temperature [17,18]. To elaborate on the recent temperature changes in China, we studied the change in mean temperature and the changes in eight extreme temperature indices.

Many studies have been conducted on extreme temperature indices in relation to China; some were conducted at the national level [19,20,21,22,23], whereas others were conducted at the regional level [24,25]. Generally, the regions considered were either administrative regions (e.g., North China, Northeast China, and East China), or certain key cities or climate-critical regions (e.g., Beijing, Shanghai, and the Tibet Plateau). As with the entire country, large administrative regions cover large areas with marked internal climatic differences that make it difficult for an overall temperature trend to reflect detailed information. Conversely, while the study of a small region might reflect the temperature variation characteristics of that region, the small geographic scope might promote a lack of representativeness. In summary, the characteristics of climate change in China could be better reflected by adopting the more reasonable and scientific approach of climate zoning, in which areas with broadly similar climatic features are assigned to the same zone.

According to the characteristics of temperature distribution, China can be divided broadly into six temperature zones [26,27,28] that could facilitate research on extreme temperature events. Variations in extreme temperature indicators differ among the different temperature zones in China. A comparative analysis of meteorological and climatic variables in different climate zones can highlight current trends in climate change. Previous studies on extreme temperature events generally focused on a single region or on year-to-year changes, and no studies have yet focused on the change of the climatic state in terms of extreme temperature events and across different temperature zones. To improve our understanding and provide a reference for further research on climate change and extreme climate events, we conducted temperature regionalization for China and examined the characteristics of extreme temperature events in six derived temperature zones using observational data from 1961 to 2020. The findings of this study could contribute toward improving national disaster prevention and mitigation capabilities, government decision-making processes, and socioeconomic development.

2. Data and Methods

2.1. Data

This study used the daily minimum temperature (TN), daily maximum temperature (TX), and daily mean temperature (TM) data for 1961–2020 derived from observations recorded at 699 National Reference Stations in China. Stations with more than 0.3% of air temperature data missing from their records and stations located in Hong Kong, Macao, Taiwan, and on other islands were excluded from our analysis. The distribution of the stations is shown in Figure 1c.

2.2. Regionalization Method

To understand the characteristics of climate change in different regions, we divided China into several temperature zones from the north to the south. Temperature regionalization reflects climate regionalization and it is generally based on specific climate states. According to Bian et al. [29], substantial differences exist between the division of temperature zones in China based on data from 1981–2010 in comparison with that derived based on data from 1951–1980. To ensure use of the latest temperature zones, we used data from 1991–2020 to conduct regionalization.

In China, most crops can only grow actively when the daily average temperature is ≥10 °C. A continuous daily average temperature of ≥10 °C (hereafter referred to as the accumulated temperature), which can reflect the thermal condition of a region, has long been an indicator used for climate zoning and agricultural climate resource evaluation in China. In 1982, Chen [30] identified that the accumulated temperature has certain limitations regarding temperature regionalization for China, which has a wide range of terrain and vast territory, and proposed using the number of days with continuous daily average temperature of ≥10 °C (hereafter referred to as accumulated temperature days) as an indicator, because it can more accurately delineate regional differences in the temperature conditions of China. Use of the indicator of accumulated temperature days has been adopted by the Central Meteorological Administration of China and other departments since the 1980s.

The regionalization method used in this study refers to the national standard for defining temperature zones in China from 1991 to 2020 [27,28]. First, on the basis of the topographic differences between the Qinghai–Tibet Plateau and other regions, the Qinghai–Tibet Plateau was classified as an independent temperature zone. Then, the remaining regions were simply divided into five temperature zones using accumulated temperature days as the main index, and accumulated temperature as the auxiliary reference index.

Using the above procedure in this study, mainland China was divided into six zones: the cold temperate zone (CTZ), middle temperate zone (MTZ), warm temperate zone (WTZ), subtropical zone (SZ), tropical zone (TZ), and the Qinghai–Tibet Plateau zone (QPZ). Details of the specific classification criteria are listed in

Table 1. Criteria for defining temperature zones in China [27,28].

Temperature Zone Name	Main Index: Number of Days When the Daily Average Temperature Is Stable ≥10 °C	Reference Index: Accumulated Temperature When the Daily Average Temperature Is Stable ≥10 °C
cold temperate zone (CTZ)	<100	<1600
middle temperate zone (MTZ)	100~170	1600~3200/3400
warm temperate zone (WTZ)	170~220	3200/3400~4500/4800
Qinghai–Tibet Plateau zone (QPZ)	\	\
subtropical zone (SZ)	220~360	4500/4800~8000
tropical zone (TZ)	>360	>8000

.

2.3. Evaluation Methods

2.3.1. Extreme Temperature Event Indices

We computed eight indices for 1961–2020 following the extreme temperature event indices of the ETCCDI (http://etccdi.pacificclimate.org, accessed on 1 January 2022) [17,31,32], as recommended by the World Meteorological Organization (

Table 2. Definition of extreme temperature indices [17,31,32].

Index	Name	Definition	Units
TXx	Extreme maximum temperature	Annual 95th percentile of maximum temperature (TX)	°C
TNn	Extreme maximum temperature	Annual 5th percentile of minimum temperature (TN)	°C
TX10p	Cold days	Days when TX < 10th percentile	Day
TN10p	Cold nights	Days when TN < 10th percentile	Day
TX90p	Warm days	Days when TX > 90th percentile	Day
TN90p	Warm nights	Days when TN > 90th percentile	Day
CSDI	Cold spell duration index	Annual count of days with at least 6 consecutivedays when TN < 10th percentile	Day
WSDI	Warm spell duration index	Annual count of days with at least 6 consecutivedays when TX > 90th percentile	Day

).

To limit the influence of single occurrences of extreme temperatures on the mean results, modifications were made to some indicators. In this study, we used the annual 95th percentile of TX as the extreme maximum temperature index (TXx) and the annual 5th percentile of TN as the extreme minimum temperature index (TNn).

2.3.2. Change Rate and Percentage

In this study, we chose the Sen’s slope estimation method to calculate the rate of change of each index with time in the different temperature zones and to analyze the data [33,34]. It is a nonparametric linear regression method that chooses the median slope among all lines through pairs of sample points. It is insensitive to measurement errors and outliers, and it has high computational efficiency.

And we also calculated the change percentage to further understand its change characteristics. In order to filter out the impact of annual fluctuations, the calculation is based on the percentage change of the average value of the index from 2011 to 2020 relative to the change from 1961 to 1970.

3. Results

3.1. Temperature Regionalization

Excluding the Qinghai–Tibet Plateau, the number of days and the accumulated temperature when the daily average temperature was stable and ≥10 °C (hereafter referred to as days and accumulated temperature, respectively) both show a decreasing trend from the south to the north (Figure 1a,b). We divided mainland China into broad temperature zones according to the regionalization method introduced in 2.2. For example, we classified the area with days of >360 days and accumulated temperature of >8000 °C as TZ. The final temperature regionalization result is shown in Figure 1c. The number of National Reference Stations varies between the zones, ranging from 6 in CTZ to 309 in SZ (

Table 3. Number of National Reference Stations in each of the six temperature zones.

Temperature Zone Name	Area Code	Number of Stations
cold temperate zone (CTZ)	1	6
middle temperate zone (MTZ)	2	135
warm temperate zone (WTZ)	3	175
Qinghai–Tibet Plateau zone (QPZ)	4	62
subtropical zone (SZ)	5	309
tropical zone (TZ)	6	12

). The focus of this study was on understanding the characteristics of climate change in different regions of China, and, despite the relatively coarse division of the six temperature zones adopted, no obvious differences were found in comparison with the temperature regionalization result of Zheng et al. [27,28], which was based on a more refined temperature regionalization scheme.

3.2. Evaluation of Mean Temperature (TM)

Before evaluating the extreme temperature events, we first needed to understand the differences in TM between the different temperature zones. The changes in TM in the different temperature zones from 1961 to 2020 are shown in Figure 2. The TM in each temperature zone has an upward trend, but the range of variation and the rate of increase in TM vary between the zones. Generally, the range of TM shows an upward trend from the north to the south, and the range of TM in CTZ, MTZ, WTZ, QPZ, SZ, and TZ is −6.7 to 1.2, −4.6 to 12.7, −6.3 to 18.0, −7.2 to 14.4, 2.3 to 25.8, and 17.1 to 26.8 °C, respectively. Owing to its high elevation, the TM of QPZ is relatively low in comparison with that of other areas at similar latitudes. During 1961–2020, the increase in mean TM was larger in the north and in QPZ. To filter out annual fluctuations, we calculated the 10-year mean TM to determine the percentage growth from 2011 to 2020 for a comparison with that from 1961 to 1970 (hereafter referred to as the percentage growth). During 1961–2020, MTZ and QPZ had the largest rates of increase of 0.34 and 0.30 °C per 10 years, respectively, and their average value in 2010–2020 represents an increase of 36.0% and 41.0%, respectively, in comparison with that in 1961–1970. The rates of increase in WTZ and CTZ were 0.29 and 0.26 °C per 10 years and the percentage growth was 12.1% and 36.0%, respectively. The lowest rates of increase were in TZ (0.26 °C per 10 years) and SZ (0.21 °C per 10 years) in southern China with a percentage growth of 5.2% and 4.9%, respectively. These results indicate that the effects of global warming are more pronounced in high-latitude and high-elevation regions.

3.3. Evaluation of Extreme Temperature Events

3.3.1. Magnitude of Extreme Temperature Events

Extreme Maximum Temperature (TXx)

One widely used indictor of extreme temperature events is TXx, which represents the extreme maximum value of temperature in a year. During 1961–2020, the range of TXx in CTZ, MTZ, WTZ, QPZ, SZ, and TZ was 24.3 to 31.3, 20.8 to 40.0, 13.2 to 44.4, 11.4 to 34.5, 15.7 to 41.3, and 28.9 to 37.3 °C, respectively (Figure 3). For the northernmost and southernmost regions in China, the range of TXx was relatively small; i.e., it did not exceed 10 °C in CTZ and TZ, but large ranges are evident in the other regions, especially in WTZ where the largest range of TXx (44.4 °C) occurred. All mean TXx values increased significantly (α = 0.05) in all zones during 1961–2020. The rates of increase in CTZ and QPZ were largest at approximately 0.35 and 0.25 °C per 10 years, together with the biggest percentage growth of 4.1% and 5.4%, respectively. Similar rates of increase of 0.18, 0.17, and 0.18 °C per 10 years were found in MTZ, SZ, and TZ with a percentage growth of 2.7%, 1.8%, and 2.5%, respectively. The rate of increase and percentage growth were smallest in WTZ (0.14 °C per 10 years and 1.7%, respectively), i.e., less than half the values in CTZ. The increase in extreme maximum temperature was greater in the northernmost and plateau regions of China.

Extreme Minimum Temperature (TNn)

The annual characteristics of extreme minimum temperature (TNn) in the six zones are shown in Figure 4. The range of TNn in CTZ, MTZ, WTZ, QPZ, SZ, and TZ during 1961–2020 was −45.7 to −26.5, −41.9 to −9.6, −42.3 to −1.0, −40.9 to −3.0, −17.6 to 14.0, and 2.8 to 17.2 °C, respectively. Except in QPZ, the mean TNn was negatively correlated with latitude. The mean TNn increased significantly (α = 0.05) during 1961–2020 in all zones except CTZ. The rate of increase was highest in MTZ (0.58 °C per 10 years) with a percentage growth of 10.8%, second highest in QPZ (0.55 °C per 10 years) with a percentage growth of 15.2%, third highest in WTZ (0.53 °C per 10 years) with a percentage growth of 18.0%, fourth highest in TZ (0.44 °C per 10 years) with a percentage growth of 19.9%, fifth highest in SZ (0.38 °C per 10 years) whose mean TNn in 1961–1970 was only 0.27 with no meaning in terms of calculating the percentage growth, and lowest in CTZ (0.17 °C per 10 years) with a percentage growth of 2.4%. Overall, the rate of growth in temperature zones with smaller TNn values was greater than that in temperature zones with larger TNn values. In other words, warming was more pronounced in northern China. In comparison with the rates of increase in TXx, the rates of increase in TNn were considerably higher, with a difference of two to three times in most zones. The changes between TXx and TNn exhibit notable asymmetry; i.e., the increase in amplitude of TNn was greater than that of TXx in most regions. However, this phenomenon was reversed in CTZ, where the rate of increase in extreme maximum temperature was highest and the rate of increase in extreme minimum temperature was lowest, reflecting an obvious imbalance in the increase of extreme temperature.

3.3.2. Frequency of Extreme Temperature Events

Warm Days (TX90p)

The warm days index (TX90p) represents the number of days on which the daily maximum temperature exceeded the 90th percentile of the maximum temperature during 1961–2020; a high TX90p value is associated with a large number of warm days. The range of TX90p during 1961–2020 in CTZ, MTZ, WTZ, QPZ, SZ, and TZ was 12–65, 5–72, 0–125, 0–115, 0–102, and 5–101 d, respectively (Figure 5). For each temperature zone, the mean TX90p increased significantly (α = 0.05) during 1961–2020, and the increase was most obvious after the 1990s, which is consistent with the finding that the frequency of high-temperature heat wave events in China showed a clear increase after the 1990s [25]. The rate of increase in TX90p was highest in TZ (4.6 days per 10 years) with a percentage growth of 78.8%, second highest in QPZ (3.0 days per 10 years) with a percentage growth of 43.0%, third highest in SZ (2.8 days per 10 years) with a percentage growth of 28.8%, fourth highest in CTZ (2.6 days per 10 years) with a percentage growth of 29.9%, fifth highest in MTZ (2.2 days per 10 years) with a percentage growth of 28.9%, and lowest in WTZ (1.9 days per 10 years) with a percentage growth of 20.7%. In China, during 1961–2020, the rate of increase in warm days was greater in the south than in the north, especially in TZ.

Warm Nights (TN90p)

The range of TN90p during 1961–2020 in CTZ, MTZ, WTZ, QPZ, SZ, and TZ was 14–59, 1–78, 1–92, 0–107, 0–121, and 5–117 d, respectively (Figure 6). Consistent with the finding regarding the warm days index (TX90p), the mean TN90p increased significantly (α = 0.05) during 1961–2020, and the increase was more obvious after the 1990s. The rate of increase was highest in TZ (8.8 days per 10 years), increasing from 15.3 days in 1961–1970 to 61.9 days in 2011–2020 with a percentage growth of 303.9%, second highest in QPZ (4.9 days per 10 years) with a percentage growth of 85.0%, joint third highest in SZ (4.0 days per 10 years) with a percentage growth of 61.9% and MTZ (4.0 days per 10 years) with a percentage growth of 66.3%, fifth highest in WTZ (3.4 days per 10 years) with a percentage growth of 46.4%, and lowest in CTZ (2.2 days per 10 years) with a percentage growth of 28.0%. Overall, the rate of increase in warm nights in southern China was greater than that in northern China, especially in TZ. Except for CTZ, the rate of increase in TN90p was higher than the rate of increase in TX90p. Our results indicate that an imbalance exists not only between extreme maximum and minimum temperatures, but also in terms of the frequency of extreme warm events both during the day and night.

Cold Days (TX10p)

The cold days index (TX10p) represents the frequency of daytime extreme cold events, as shown in Figure 7. The range of TX10p during 1961–2020 in CTZ, MTZ, WTZ, QPZ, SZ, and TZ was 6–78, 0–83, 1–86, 0–106, 1–126, and 6–85 d, respectively. The mean TX10p decreased significantly (α = 0.05) during 1961–2020 in all temperature zones, and the decrease was more obvious after the 1990s. The rate of decrease in TX90p was highest in QPZ (−3.4 days per 10 years) with a percentage reduction of 32.6% from 2011 to 2020 in comparison with 1961–1970 (hereafter referred to as percentage reduction), second highest in CTZ (−3.0 days per 10 years) with a percentage reduction of 38.0%, third highest in TZ (−2.3 days per 10 years) with a percentage reduction of 15.1%, fourth highest in WTZ (−2.2 days per 10 years) with a percentage reduction of 24.4%, fifth highest in SZ (−2.1 days per 10 years) with a percentage reduction of 15.7%, and lowest in MTZ (−1.9 days per 10 years) with a percentage reduction of 20.1%. Overall, the rate of decrease in cold days was more obvious in the northernmost and plateau regions of China. In comparison with the rates of increase in warm days, the rates of decrease in cold days in MTZ, SZ, and TZ during the study period were smaller, especially in TZ where the difference was up to two times; in CTZ, WTZ, and QPZ, the opposite relation was true.

Cold Nights (TN10p)

The cold nights index (TN10p) represents the frequency of extreme cold events at night, as shown in Figure 8. The range of TN10p in CTZ, MTZ, WTZ, QPZ, SZ, and TZ during 1961–2020 was 4–82, 0–96, 0–84, 0–155, 0–113, and 3–74 d, respectively. Except for CTZ, the mean TN10p decreased significantly (α = 0.05) during 1961–2020, and the rate of reduction was more obvious after the 1990s. The rate of reduction in TX90p was joint highest in QPZ (−5.1 days per 10 years) with a percentage reduction of 52.0% and TZ (−5.1 days per 10 years) with a percentage reduction of 40.4%, third highest in WTZ (−4.4 days per 10 years) with a percentage reduction of 40.0%, fourth highest in SZ (−3.9 days per 10 years) with a percentage reduction of 40.4%, fifth highest in MTZ (−3.8 days per 10 years) with a percentage reduction of 37.0%, and lowest in CTZ (−1.3 days per 10 years) with a percentage reduction of 17.5%. Overall, the rate of reduction in cold nights was more obvious in the southernmost and plateau regions of China. During the study period, in comparison with cold days and cold nights, the rate of reduction in cold nights in all temperature zones (except CTZ) was larger than the corresponding rate of reduction in cold days, especially in TZ where the difference was up to 2 times. Thus, asymmetry also clearly exists in the frequency of extreme cold events both during the day and night.

3.3.3. Duration of Extreme Temperature Events

Warm Spell Duration Index (WSDI)

The general trend of the warm spell duration index (WSDI) during 1961–2020 was similar to that of TX90p (see Figure A1). The range of the WSDI during 1961–2020 in CTZ, MTZ, WTZ, QPZ, SZ, and TZ was 0–49, 0–58, 0–104, 0–104, 0–69, and 0–70 d, respectively. In all temperature zones, the mean WSDI increased significantly (α = 0.05) during 1961–2020, and the rate of increase was more obvious after the 1990s. The rate of increase was highest in TZ (1.9 days per 10 years) where the mean WSDI of 9.7 days in 1961–1970 increased to 23.8 days in 2011–2020 with a percentage growth of 146.6%, second highest in QPZ (1.6 days per 10 years) with a percentage growth of 87.6%, third highest in SZ (1.4 days per 10 years) with a percentage growth of 26.4%, joint fourth highest in MTZ (1.4 days per 10 years) with a percentage growth of 70.8% and CTZ (1.4 days per 10 years) with a percentage growth of 51.5%, and lowest in WTZ (1.0 days per 10 years) with a percentage growth of 46.6%. The overall rate of growth in southern China was slightly higher than that in the north, especially in TZ.

Cold Spell Duration Index (CSDI)

The variations in the cold spell duration index (CSDI) were consistent with those of TN10p (see Figure A2). The range of the CSDI in CTZ, MTZ, WTZ, QPZ, SZ, and TZ was 0–54, 0–78, 0–72, 0–132, 0–90, and 0–72 d, respectively. In all temperature zones (except CTZ), the mean WSDI decreased significantly (α = 0.05) during 1961–2020, and the rate of reduction was more obvious after the 1990s. The rate of change in the CSDI in TZ, QPZ, WTZ, MTZ, SZ, and CTZ was −2.9, −2.8, −2.6, −2.2, −2.1, and −1.1 days per 10 years, respectively, with percentage reductions of 59.4%, 78.1%, 59.8%, 54.0%, 54.6%, and 41.8%, respectively. The rate of reduction in the CSDI was higher than the rate of increase in the WSDI in all zones except CTZ. Thus, asymmetry also clearly exists in the duration of extreme cold and warm events.

4. Discussion and Conclusions

In this study, we divided mainland China into six temperature zones, and used data recorded at National Reference Stations during 1961–2020 to evaluate the annual characteristics of the mean temperature and extreme temperature event indicators, based on the extreme climate indices of the ETCCDI.

The overall mean temperature showed a decreasing trend with increasing latitude, but the rate of increase in mean temperature was more pronounced in the northern region and QPZ, consistent with the conclusions of other studies that found the effects of global warming to be more significant in high-latitude and high-elevation regions [4].

To analyze the annual variation characteristics of extreme temperature events, we evaluated the magnitude, frequency, and duration of extreme temperature events in the six temperature zones during 1961–2020. In all temperature zones (except CTZ), extreme maximum/minimum temperature had a significant trend of increase (α = 0.05) with time, and the rates of increase were more obvious in the northern and QPZ regions, consistent with previous research [13,20]. We also found that the rate of increase in extreme maximum temperature was lower than the rate of increase in extreme minimum temperature. Unlike other studies, for extreme maximum temperatures, the rate of increase in the northernmost region of China was the most significant, but, for extreme minimum temperatures, the rate of increase was the smallest, which is contrary to the conclusions reached in relation to the other temperature zones. In each temperature zone, the difference between the rates of increase in extreme maximum temperature and extreme minimum temperature was up to an astonishing two to three times, which might represent evidence of global warming.

For most temperature zones, the frequency and duration of extreme warm events (including warm days/nights and the WSDI) increased significantly (α = 0.05) with time, and the frequency and duration of extreme cold events (including cold days/nights and the CSDI) decreased significantly (α = 0.05) with time; the increase/decrease was more obvious after the 1990s. Similar to our conclusions, previous work also found that the increase or decrease in the national average of warm days/nights and cold days/nights across all of China was more significant after the mid–late 1980s [13,14]. We also found that, for the indicators of warm days, warm nights, and the WSDI, the rate of increase in TZ (the southernmost part of mainland China) was significantly higher than that in other regions, especially for warm nights; i.e., the rate of increase in TZ (8.8 days per 10 years) was four times that of CTZ (2.2 days per 10 years). The number of warm nights in TZ was only 15.3 days in 1961–1970, but it increased to 61.9 days in 2011–2020, which is an increase of 303.9%.

Interestingly, Zhai and Pan [13] found that the trend of increase (decrease) in the national average number of warm (cold) days/nights was more pronounced in the northern and southwestern regions, contrary to our conclusion. In terms of warm days/nights and the WSDI, the place with the highest rate of growth was TZ, i.e., the southernmost point of mainland China, followed by QPZ, and then SZ in the south. Irrespective of either warm nights or warm days, the rate of growth in the south was greater than that in the north. This contradiction might be related to change in the selected interval; i.e., some studies found that the WSDI in China gradually decreased from the 1960s to the 1980s and gradually developed toward the south [15]. For cold days/nights and the CSDI, QPZ had a relatively large rate of growth, and the rate of increase in TZ and WTZ was also larger than that in either MTZ or SZ. Surprisingly, the rate of reduction in cold days in CTZ was higher than that in the southern region.

Overall, the rate of growth in warm nights was significantly higher than the rate of growth in warm days, the rate of reduction in cold nights was significantly higher than the rate of reduction in cold days, and the rate of decrease in the CDSI was significantly higher than the rate of increase in the WSDI. The findings of other recent studies support our results [35]. In recent years, extreme weather events have shown increasing trends in frequency, strength, and duration. For example, summer 2022 was the hottest driest summer in China on record; on 11 June 2023, the Ayding Lake Meteorological Station experienced a maximum temperature of 48.5 °C, breaking the national record for the highest temperature recorded in June.

Notably, QPZ ranked in the top two zones in terms of the rate of change in all extreme temperature event indicators. In QPZ, the climate is strongly influenced by the unique topography, and the rate of reduction in cold days/night was especially large, which might be associated with an amplification of the effects of climate change on the Qinghai–Tibet Plateau [25,35]. The most notable finding is that the asymmetry observed in CTZ is the opposite to that found in the other temperature zones. Few studies have examined this phenomenon, and further research on extreme temperature events in CTZ should be pursued in the future.