Changes in Meteorological Elements and Its Impacts on Yunnan Plateau Lakes

Abstract

In this study, we analyzed the main climatic factors influencing nine plateau lakes in Yunnan Province over the past 40 years (from 1980 to 2019) based on trend and mutation analyses. The results revealed that the air temperatures of these plateau lakes have increased, with an annual average air temperature increase of 0.18 °C per decade, during this 40-year period. From 2000 to 2005, there was an abrupt change in the air temperature increase, the rate of which was 0.20 °C per decade, and there were reductions in long-wave radiation, relative humidity, wind speed, precipitation, and snowfall. In addition, the rising trend in the air temperature of these lakes located in northwest Yunnan (temperate climate, higher elevation) was found to be significantly higher than that of the lakes in central Yunnan (subtropical climate, lower elevation), whereas in contrast, the declining trends in long-wave radiation, relative humidity, wind speed, and precipitation were more pronounced in central than in northwestern Yunnan.

1. Introduction

Lakes are among the most significant geographical units affected by climate change, to which they are particularly sensitive, and are accordingly considered as sentinels of climatic and environmental change [1]. Climate change is among the most complex challenges currently facing humanity, and will become increasingly more so in the future. Since the early 1980s, much of the planet has witnessed the effects of rising temperatures due to significant global warming, resulting from anthropogenic-derived increases in the emission of carbon dioxide and other greenhouse gases [2]. Among aquatic ecosystems, lakes are perhaps the most sensitive to climate change. Climate factors such as air temperature, wind speed, relative humidity, precipitation, solar radiation, and long-wave radiation can directly or indirectly affect the physical, chemical, and biological processes of lakes, and even subtle changes in these meteorological factors can potentially lead to complex non-linear responses in lakes [3,4,5,6,7,8,9], including increases in air temperature and reductions in wind speed, which have been recorded in many lakes, and will contribute to increases in surface water temperature and algal biomass during the growth season, as well as a prolongation of lake stratification. The accumulation of organic matter in lake sediments and reductions in dissolved oxygen content in the lower water layers will lead to complex physical, chemical, and biological feedback processes, and when the cumulative effects of these processes reach or exceed a certain threshold, this may have catastrophic non-linear repercussions for lake ecosystems [10,11,12]. The potential occurrence of such adverse effects thus highlights the necessity of gaining a thorough understanding of the long-term change trends of climate factors, as basis for further examining the impact of climate change on lake ecosystems.

Globally, climate change is characterized by significant geographical heterogeneity [13,14,15,16,17]. For example, wind speeds have declined in the past 30 years globally. However in South America, Southeast Asia, Central Asia, Africa, and Australia, they have increased [18]. The warming of large inland waters is much greater in the northern hemisphere at middle and high latitudes than in the lower latitudes and southern hemisphere [19]. Previous studies have shown that climate change will have more severe impacts at higher elevations [20]. In this respect, although analysis based on global coarse-grain data can provide insights regarding global trends and geographical heterogeneity as a whole, it may fail to accurately predict changes at a more local scale. The average air temperature of the six lakes in Northern Poland increased from 0.02 °C to 0.025 °C per year, while the temperature of the lake water varies more, from 0.005 °C to 0.028 °C per year [21]. These values are considerably larger than estimates based on global data [8]. In addition, due to the extremely rich diversity of morphological characteristics, such as lake area and water depth, the responses of different lakes to climate change also differ to a considerable extent [22,23]. Consequently, a larger number of regional case studies are necessary to gain a more detailed understanding of climate change processes, and thereby provide a basis for clarifying the responses of lakes to such changes.

The Yunnan Plateau lakes form one of the five major lakes in China, with nine major lakes including Dianchi, Yangzonghai, Fuxian Lake, Xinyun Lake, Qilu Lake, Erhai, Lugu Lake, Chenghai, and Yilong Lake. These plateau lakes are semi-enclosed fault lakes, located in regions at high elevations that are exposed to high levels of radiation and are characterized by a shortage of natural water resources and long water exchange cycles. The climate and morphological characteristics of the plateau lakes in Yunnan Province are conducive to the growth of algae and the accumulation of organic materials, which may render these water bodies more sensitive to climate change. In addition, these lakes are notably diverse, particularly with respect to the influence of differences in elevation and climate type. For example, whereas the highest lake, Lake Lugu, is located at an elevation of 2690 m, the lowest, Lake Chenghai, lies at the elevation of 1501 m. With respect to climate, Lake Lugu lies within a temperate climatic zone, whereas lakes Dianchi, Fuxian, Yangzonghai, Yilong, Xingyun, and Qilu lie within a subtropical climate zone, and lakes Chenghai and Erhai Lake are located in two transitional zones. Given the diversity and typicality of the Yunnan Plateau lakes, we selected these lakes to undertake a case study, with the aim of gaining a further understanding of the change trends in key climate factors over the past 40 years in a plateau region characterized by a concentrated distribution of fault-depressed lakes. In addition, previous studies on climate change in Yunnan highland lakes have been limited to individual lakes and the trend analysis of a few meteorological factors, thus, the nine highland lakes in Yunnan lack an overall climate change trend analysis and variance analysis, and our study can fill this gap.

2. Methodology

2.1. Study Area

In this study, we focused on nine plateau lakes larger than 30 km² in Yunnan Province, namely, lakes Lugu, Chenghai, and Erhai in northwest Yunnan, and lakes Dianchi, Fuxian, Yangzonghai, Yilong, Xingyun, and Qilu in central Yunnan. The locations of these lakes in study area are shown in Figure 1, and data for associated morphological parameters are presented in

Table 1. Morphological parameters of the nine major lakes on the Yunnan Plateau.

Lake	Watershed Area/km2	Surface Area/km2	Water Surface Elevation/m	Average Depth/m	Maximum Depth/m	Water Change Cycle/Day
Chenghai	318	77.20	1501.00	24.98	35.87	5861
Dianchi	2920	297.90	1887.40	5.01	11.35	1540
Erhai	2565	249.00	1974.31	8.80	19.50	3500
Fuxian	675	211.00	1722.50	95.20	158.90	11,584
Lugu	248	48.50	2690.80	38.40	105.30	11,727
Yangzonghai	292	31.70	1770.00	21.50	28.59	4602
Yilong	360	38.00	1414.20	3.90	5.70	1295
Xingyun	378	34.70	1722.50	6.01	10.81	1642
Qilu	354	36.90	1797.65	4.20	6.84	316

. The total water area of these nine plateau lakes is approximately 1021 km², and the total drainage area is 8110 km². Among these lakes, Lake Lugu is a temperate climate lake, lakes Chenghai and Erhai are located in a transition area of temperate and subtropical zones, and lakes Dianchi, Fuxian, Yangzonghai, Qilu, Xingyun, and Yilong are subtropical climate lakes. Lake Dianchi is the largest lake in southwest China and the sixth largest freshwater lake in China, whereas Lake Fuxian is the second deepest lake in China and the lake containing the largest volume of water in Yunnan Province. Lake Erhai is the second largest plateau lake in Yunnan Province, and Lake Lugu is the third deepest lake in China. The areas of lakes Dianchi, Erhai, and Fuxian are each greater than 200 km², whereas those of the other six lakes are all less than 80 km².

2.2. Data Source

As a source of data for the analysis of climate change in the Yunnan Plateau lakes, we used an ERA5 re-analysis dataset, compiled by the European Centre for Medium-range Weather Forecasts (ECMWF), which was downloaded from the Copernicus Climate Change Service (C3S) Climatic Data Centre (https://cds.climate.copernicus.eu, accessed on 19 January 2023), and is the most qualitative re-analysis dataset worldwide. For analytical purposes, we downloaded and extracted data for the hourly near-surface air temperature, short-wave radiation, long-wave radiation, relative humidity, wind speed, rainfall, and snowfall of the ERA5 grid units nearest to each lake from January 1980 to December 2019. Due to the close location of the three lakes, Yilong Lake, Xingyun Lake and Qilu Lake, the data was extracted as the same grid, and Yilong Lake was used as a proxy for Qilu Lake and Xingyun Lake.

2.3. Innovation Trend Analysis (ITA)

In this study, innovative trend analysis (ITA) [24] was used to analyze the trends in climate factors over the past 40 years for the Yunnan Plateau region. The ITA method divides time series into two sub-series of equal length, arranges these series in ascending order and generates a corresponding scatter diagram. The first and second sections of the time series are placed on the X and Y axes, respectively. If the data points are distributed near the line of equality (1:1), the series show no distinct trend, if the points are plotted above the 1:1 line, the series are characterized by an upward trend, and if below the line, the trend is downward, as in Figure 2. The major advantage of the ITA method is that it provides a visual interpretation of time series trends that does not have restrictive assumptions and is not significantly influenced by a few missing data points in the time series.

The slope S of the linear trend plotted by ITA can be calculated as follows:

$$S=\frac{2(\overline{y_2}-\overline{y_1})}{n}$$

(1)

where $y_1$ and $y_2$ are the average of the two subseries and n is the total number of data points in the entire time series. The standard deviation of the slope (${\sigma }_s$) is given by the following equation:

$${\sigma }_s=\frac{2\sqrt{n}}{n\sqrt{n}}\sigma \sqrt{1-{\rho }_{y_1{y}_2}}$$

(2)

where $\sigma$ is the standard deviation of the entire time series, and ${\rho }_{y_1{y}_2}$ is the correlation between the subseries, calculated as follows:

$$\begin{array}{c}\hfill {\rho }_{y_1{y}_2}=\frac{\frac{n}{2}(\mathsf{\Sigma }{y}_1{y}_2-\mathsf{\Sigma }{y}_1\mathsf{\Sigma }{y}_2)}{\sqrt{\left[\frac{n}{2}\left(\mathsf{\Sigma }{y_1}^2\right)-{\left(\mathsf{\Sigma }{y}_1\right)}^2\right.\left[\frac{n}{2}\left(\mathsf{\Sigma }{y_2}^2\right)-{\left(\mathsf{\Sigma }{y}_2\right)}^2\right.}}\end{array}$$

(3)

The significance of the trend was also assessed using the ITA technique. For a given significance level, the critical slope ($S_{}$${}_{crit}$) is determined and if $\left|S\right|>\left|S_{crit}\right|$, then the trend is considered significant. For significance levels of 5% versus 1%, the critical slope is given by the following equation:

$$\begin{array}{c}\hfill S_{crit}=1.96\times {\sigma }_s\\ \hfill S_{crit}=2.58\times {\sigma }_s\end{array}$$

(4)

2.4. The Mann-Kendall Test

The Mann–Kendall test can be used to calculate trends and identify abrupt changes in temperature. Given a sequence x(t) of length n, and assuming that the time series x(t) is independently distributed and that it does not show the onset of a developing trend, the test calculates two standardized statistical sequences UF and UB and plots these with confidence lines. If the two curves of UB and UF curves intersect within the confidence zone, the time of intersection is the time of an abrupt change, where the $95\%$ confidence levels (1.96 and −1.96) are considered to be the boundaries of the confidence zone. To determine UF and UB, the normal distribution statistic $S_k$ should initially be computed first and then normalized to obtain $U{F}_k$. The equations used to calculate UF and UB are as follows:

$$S_k=\sum _{i=1}^k\sum _{j=1}^{i}logistic(x_i>x_j),(k=2,3,\dots ,n)$$

(5)

The logistic expression ($x_i$ > $x_j$) is 1 when $x_i$ > $x_j$ and otherwise 0. $S_k$ is then normalized using the follows equation:

$$U{F}_k=\frac{S_k-E\left(S_k\right)}{\sqrt{Var\left(S_k\right)}},(K=1,2,\dots ,n)$$

(6)

where: $U{F}_1$ = 0 and E($S_k$) and Var($S_k$) are the mean and variance of $S_k$, respectively, which can be calculated when $x_1$, $x_2$, …, $x_n$ are independent of each other and have the same continuous distribution, as determined using the following equation:

$$\begin{array}{c}\hfill \left\{\begin{array}{c}E(S_k=\frac{K(k+1)}{4})\\ Var\left(S_k\right)=\frac{K(k-1)(2k+5)}{72}\end{array}\right.\end{array}$$

(7)

x(t) is then inverted to obtain x’(t), and the above process is repeated to calculate UB. If the value for $U{F}_k$ is greater than 0, this indicates that the series has an upward trend, and if less than 0, the trend is downward. When the $U{F}_k$ or $U{B}_k$ curve exceeds the critical line, it indicates a significant upward or downward trend. If the $U{F}_k$ and $U{B}_k$ curves have a point of intersection and this is between the critical lines, the moment corresponding to the intersection point is the time at which the trend change commences.

2.5. Technical Route

Based on the above research background, research content and research method, this study is carried out according to the following technical route, as shown in Figure 3.

3. Results

3.1. Annual Analysis

Table 2. Results for innovation trend analysis of the annual mean values of climatic factors of lakes on the Yunnan Plateau.

Climate Factor	Lake	Slope s	Standard Deviation	Correlation	Slope Standard	Significant Level (p < 0.5)	Significant Level (p < 0.1)
	Chenghai	0.02182 **	0.370	0.934	0.00107	±0.00209	±0.00275
	Dianchi	0.01582 **	0.338	0.904	0.00117	±0.00230	±0.00302
	Erhai	0.02200 **	0.368	0.935	0.00105	±0.00206	±0.00271
Temperature	Fuxian	0.01560 **	0.339	0.920	0.00107	±0.00210	±0.00277
	Lugu	0.02596 **	0.394	0.925	0.00121	±0.00237	±0.00312
	Yangzonghai	0.01422 **	0.338	0.892	0.00124	±0.00243	±0.00320
	Yilong	0.01486 **	0.344	0.944	0.00091	±0.00179	±0.00236
	Chenghai	0.15551 **	5.125	0.966	0.01055	±0.02068	±0.02722
	Dianchi	−0.04296 **	4.218	0.977	0.00710	±0.01392	±0.01832
	Erhai	0.04423 **	4.619	0.970	0.00892	±0.01749	±0.02302
Short-wave	Fuxian	−0.03613 **	4.603	0.983	0.00666	±0.01306	±0.01719
	Lugu	0.09707 **	4.238	0.927	0.01281	±0.02510	±0.03305
	Yangzonghai	−0.01427 **	4.542	0.963	0.00973	±0.01907	±0.02510
	Yilong	−0.00252 **	4.915	0.966	0.01006	±0.01972	±0.02596
	Chenghai	−0.02745 **	2.766	0.970	0.00540	±0.01058	±0.01393
	Dianchi	−0.12040 **	2.786	0.944	0.00734	±0.01439	±0.01895
	Erhai	−0.06833 **	2.820	0.975	0.00502	±0.00984	±0.01295
Long-wave	Fuxian	−0.12115 **	2.527	0.931	0.00745	±0.01460	±0.01922
	Lugu	0.05118 **	2.804	0.953	0.00677	±0.01327	±0.01746
	Yangzonghai	−0.14777 **	2.484	0.902	0.00869	±0.01703	±0.02242
	Yilong	−0.15701 **	2.647	0.917	0.00852	±0.01670	±0.02198
	Chenghai	−0.08670 **	7.341	0.938	0.02052	±0.04021	±0.05293
	Dianchi	−0.13394 **	8.924	0.945	0.02337	±0.04581	±0.06030
	Erhai	−0.02906	6.840	0.741	0.03894	±0.07632	±0.10047
Snowfall	Fuxian	−0.13394 **	9.292	0.945	0.02433	±0.04769	±0.06278
	Lugu	−0.39074 **	14.145	0.918	0.04530	±0.08880	±0.11689
	Yangzonghai	0.08840 *	11.593	0.912	0.03839	±0.07523	±0.09903
	Yilong	0.03378	7.052	0.870	0.02845	±0.05576	±0.07340
	Chenghai	−0.00062 **	0.064	0.938	0.00018	±0.00035	±0.00046
	Dianchi	−0.00164 **	0.092	0.937	0.00026	±0.00051	±0.00067
	Erhai	−0.00040 *	0.078	0.949	0.00020	±0.00039	±0.00051
Wind speed	Fuxian	−0.00208 **	0.102	0.879	0.00040	±0.00078	±0.00102
	Lugu	−0.00017 *	0.049	0.978	0.00008	±0.00016	±0.00021
	Yangzonghai	−0.00173 **	0.113	0.944	0.00030	±0.00059	±0.00077
	Yilong	−0.00007	0.081	0.941	0.00022	±0.00043	±0.00057
	Chenghai	−0.00062 **	0.064	0.938	0.00018	±0.00035	±0.00046
	Dianchi	−0.00164 **	0.092	0.937	0.00026	±0.00051	±0.00067
	Erhai	−0.00040 *	0.078	0.949	0.00020	±0.00039	±0.00051
Relative humidity	Fuxian	−0.00208 **	0.102	0.879	0.00040	±0.00078	±0.00102
	Lugu	−0.00017 *	0.049	0.978	0.00008	±0.00016	±0.00021
	Yangzonghai	−0.00173 **	0.113	0.944	0.00030	±0.00059	±0.00077
	Yilong	−0.00007	0.081	0.941	0.00022	±0.00043	±0.00057
	Chenghai	−0.30418	109.764	0.963	0.23723	±0.46497	±0.61206
	Dianchi	−2.30199 **	140.530	0.977	0.24031	±0.47101	±0.62000
	Erhai	−1.26125 **	118.279	0.979	0.19376	±0.37976	±0.49989
Precipitation	Fuxian	−2.80528 **	143.718	0.963	0.30982	±0.60724	±0.79933
	Lugu	0.61480 **	101.235	0.969	0.19953	±0.39109	±0.51480
	Yangzonghai	−2.53510 **	155.331	0.976	0.27021	±0.52960	±0.69713
	Yilong	−2.59767 **	137.741	0.978	0.22977	±0.45035	±0.59280

* Significant at the 95% level, ** significant at the 99% level.

lists the statistical results of climate change for the Yunnan Plateau lakes from 1980 to 2019, based on the ITA method. Figure 4 shows the change trend of the annual average value of each of the assessed climate factors. The results indicate an upward trend in the air temperature for the plateau lakes, whereas the trends of wind speed and relative humidity are downward. With the exception of Lake Lugu, long-wave radiation and rainfall are also characterized by downward trends. In the case of short-wave radiation, a downward trend was detected for those lakes located in central Yunnan, whereas an upward trend was observed for the lakes in northwest Yunnan, and snowfall showed a mainly downward trend.

Figure 4 shows that all air temperature-related data points are located above the 1:1 line, and all lakes show a significant upward trend in air temperature from 1980 to 2019, among which, the rising trend of air temperature for the lakes in northwest Yunnan is higher than that of the central Yunnan lakes, with Lake Lugu in the northwest being characterized by the largest upward trend. The short-wave radiation for lakes in northwest Yunnan showed a similar upward trend, whereas in the case of the lakes in central Yunnan, the trend was downward. As in the case of air temperature, Lake Lugu in northwest Yunnan was characterized by the largest upward trend. With the exception of Lake Lugu, which showed an upward trend, all other lakes were characterized by a downward trend for long-wave radiation, although the descending trend for lakes in central Yunnan was found to be more pronounced than that of lakes in northwest Yunnan. Contrastingly, we were unable to detect any clear trend with respect to snowfall. All lakes were, however, characterized by a downward trend for wind speed, although again this trend was more pronounced for the lakes in central Yunnan than for those in the northwestern region. Both relative humidity and rainfall also showed downward trends, with the latter being characterized by a more pronounced trend in the central than in the northwestern Yunnan lakes.

3.2. Seasonal Analysis

Table 3. Results of innovation trend analysis of seasonal climate factors in the Yunnan Plateau lake.

Climate Factor	Lake	Spring	Summer	Autumn	Winter
	Chenghai Lake	0.02049 **	0.01493 **	0.02014	0.03005 **
	Dianchi Lake	0.01980 **	0.00974 **	0.01441 **	0.01849 **
	Erhai Lake	0.01981 **	0.01543 **	0.02211 **	0.03054 **
Temperature	Fuxian Lake	0.01922 **	0.00970 **	0.00033 **	0.00058 **
	Lugu Lake	0.02581 **	0.02127 **	0.02528 **	0.03182 **
	Yangzonghai Lake	0.01992 **	0.00827 **	0.01160 **	0.01568 **
	Yilong Lake	0.01584 **	0.00735 **	0.01381 **	0.02140 **
	Chenghai Lake	−0.03904 *	0.23385 **	0.08314 *	0.23946 **
	Dianchi Lake	−0.19659 **	−0.04175 *	−0.01318	0.07005 **
	Erhai Lake	−0.12159 **	0.08579 *	0.06798 *	0.13233 **
Short-wave	Fuxian Lake	−0.14171 **	−0.09139 **	0.00616 *	0.01657
	Lugu Lake	−0.01859	0.23997 **	0.12516 **	0.03119 *
	Yangzonghai Lake	−0.07571 *	−0.09379 **	−0.00705	0.13242 **
	Yilong Lake	0.05724	−0.11834 **	0.07549 *	−0.02627
	Chenghai Lake	0.05053 **	−0.04366 **	−0.05290 *	−0.02040 *
	Dianchi Lake	−0.00480	−0.10790 **	−0.21446 **	−0.14980 **
	Erhai Lake	0.03311 *	−0.08575 **	−0.14096 **	−0.07648 **
Long-wave	Fuxian Lake	−0.00996	−0.11464 **	0.00606 **	0.00841 **
	Lugu Lake	0.11121 **	0.05353 **	−0.00604	0.06507 **
	Yangzonghai Lake	−0.05279 **	−0.09544 **	−0.27318 **	−0.17481 **
	Yilong Lake	−0.09311 **	−0.18873 **	−0.25344 **	−0.10360 **
	Chenghai Lake	0.07843	0	−0.21231 **	−0.21291 *
	Dianchi Lake	−0.08717 *	0	−0.11769 **	−0.32043 **
	Erhai Lake	−0.03345	0	−0.01676 **	−0.03992 **
Snowfall	Fuxian Lake	−0.01380	0	−0.00255 **	0
	Lugu Lake	−0.06475	0	−0.86887 **	−0.75309 **
	Yangzonghai Lake	0.01290	0	−0.28136 **	0.64893 **
	Yilong Lake	0.06746 **	0	−0.10746 **	0.17069
	Chenghai Lake	−0.00273 **	−0.00111 **	0.00006	0.00051 *
	Dianchi Lake	−0.00497 **	−0.00312 **	0.00140 *	−0.00032
	Erhai Lake	−0.00352 **	−0.00128 **	0.00246 **	0.00055
Wind speed	Fuxian Lake	−0.00480 **	−0.00513 **	0.00018 **	0.00021
	Lugu Lake	−0.00117 **	−0.00103 **	−0.00017*	0.00161 **
	Yangzonghai Lake	−0.00454 **	−0.00575 **	0.00204 **	0.00057
	Yilong Lake	−0.00201 **	−0.00212 **	0.00354 **	−0.00002
	Chenghai Lake	−0.00090 **	−0.00106 **	−0.00100 **	−0.00120 **
	Dianchi Lake	−0.00119 **	−0.00102 **	−0.00138 **	−0.00157 **
	Erhai Lake	−0.00104 **	−0.00151 **	−0.00174 **	−0.00174 **
Relative humidity	Fuxian Lake	−0.00107 **	−0.00106 **	0.00002 **	0.00002 **
	Lugu Lake	−0.00067 **	−0.00069 **	−0.00062 **	−0.00051 *
	Yangzonghai Lake	−0.00101 **	−0.00080 **	−0.00121 **	−0.00126 **
	Yilong Lake	−0.00097 **	−0.00144 **	−0.00162 **	−0.00098 **
	Chenghai Lake	3.96296 **	−0.82457	−2.47601	−1.06519 **
	Dianchi Lake	1.75127	−4.83401 **	−5.67196 **	−0.51708 **
	Erhai Lake	4.57181 **	−2.43614	−5.70940 **	−2.12645 **
Precipitation	Fuxian Lake	0.76844 **	−3.72356	0.16865 **	0.24276
	Lugu Lake	4.14140 **	1.13157	−2.73790 **	−0.17256 **
	Yangzonghai Lake	0.40433	−2.08926	−7.39131 **	−1.13807 **
	Yilong Lake	1.73657	−5.55325 **	−7.35147 **	0.62853

* Significant at the 95% level, ** significant at the 99% level.

shows the results of ITA performed to detect seasonal trends in the climate. These indicate that air temperature has an upward trend, short-wave radiation a downward trend, and long-wave radiation shows a downward trend in the lakes of central Yunnan and an upward trend in the lakes in northwest Yunnan. No significant changes were detected for snowfall, whereas wind speed and relative humidity have downward trends. In spring, rainfall shows an upward trend, which is opposite to the trend detected for annual rainfall. In summer, air temperature shows an upward trend, albeit smaller than the trends detected for the other three seasons. The seasonal trend in short-wave radiation was found to be consistent with that of annual short-wave radiation, with the lakes in northwest and central Yunnan showing upward and downward trends, respectively. During autumn, air temperatures in this region are on the rise, and this is accompanied by a general increase in the amounts of short-wave radiation, whereas long-wave radiation and rainfall are generally on the decline, with the decline trend tending to be largest in this season. Snowfall and relative humidity are mainly declining in autumn, whereas there is an upward trend in wind speed, which is opposite to that seen in spring and summer, and also the overall annual trend. During winter, both air temperature and short-wave radiation are rising, and the rising trends for both tends to be highest in this season. Contrastingly, long-wave radiation, snowfall, relative humidity, and rainfall are mainly declining in this season. Collectively, these results for seasonal climate change trends among the Yunnan Plateau lakes indicate that these lakes are more sensitive to global warming during the colder seasons.

3.3. Analysis of Sudden Changes in Air Temperature

Sudden changes in climate are ubiquitous in the climatic system and occur when the climate changes from one stable state to another, which is a specific manifestation of the non-linearity of the climate system [25]. In our analysis of the trends in climatic factors over the 40 years from 1980 to 2019, we found that the air temperature was the only factor showing a significant upward trend, and we performed a Mann–Kendall analysis to determine the point at which the air temperature underwent an abrupt change among the Yunnan Plateau lakes.

The data presented in Figure 5 indicates that the abrupt change in the annual average air temperature of the Yunnan Plateau lakes occurred in the early 2000s, and the year in which this abrupt change occurred was earlier for lakes in northwest Yunnan than for the central lakes. Specifically, this change was initially detected in the northwest lakes in 2001, whereas for most lakes in central Yunnan, the change took place 2002, the exception being Lake Yilong, where air temperatures rose abruptly in 2004. This is later than the time point proposed by Cheng Qingping et al. [26], who reported that the annual average ground air temperature in Yunnan showed a significant abrupt warming commencing in 2000. We suspect that this disparity in findings could be attributable to the use of different time series data. In their comparison of the times of abrupt changes in air temperature in different regions of China, Ding Yihui et al. [27] identified clear geographical differences, with changes occurring earlier in higher latitude areas than in those at lower latitudes. This is consistent with our finding of the later abrupt change for Lake Yilong, which among the Yunnan Plateau lakes lies at the lowest latitude. In addition, we found that the air temperature of the Yunnan Plateau lakes showed an initial downward trend from the early 1980s until 1985, at which point the trend reversed. In 1997, we detected a sudden change from cold to warm in the lakes of central Yunnan, which is consistent with the findings of Liu Yu et al. [28], thereby indicating that the annual average air temperature in Yunnan has increased significantly in the past 46 years, and there was a sudden change from cold to warm in 1997.

Table 4. Maximum, minimum, and average air temperatures (°C) of the Yunnan Plateau lakes per decade.

Lake	1980–1989			1990–1999			2000–2009			2010–2019
	Max	Min	Mean	Max	Min	Mean	Max	Min	Mean	Max	Min	Mean
Chenghai	30.11	−15.99	11.51	29.17	−6.67	11.31	28.38	−7.09	11.26	30.75	−12.32	12.30
Dianchi	29.85	−14.15	14	31.85	−6.15	13.91	29.85	−10.15	13.66	31.85	−10.15	14.55
Erhai	33.34	−5.56	16.45	33.72	−2.87	16.23	33.31	−0.47	16.19	33.34	−2.16	17.19
Fuxian	32.20	−6.92	15.16	32.23	−3.47	15.07	29.97	−3.98	14.82	31.15	−6.61	15.72
Lugu	25.95	−27.49	7.14	25.08	−18.72	6.98	25.04	−20.29	7.02	26.98	−20.10	8.02
Yangzonghai	32.84	−6.91	14.98	34.52	−4.39	14.95	31.40	−4.98	14.6	32.99	−9.71	15.47
Yilong	34.23	−3.41	17.43	33.47	−1.34	17.22	32.65	−1.64	17	33.18	−8.35	18

lists the maximum, minimum, and average values of air temperature for each of the four decades during the period from 1980 to 2019, which reveal that for almost all lakes, air temperatures have been increasing continuously, and that the magnitude of change has been increasing since 1980, reaching the fastest growth rate of 0.2 ∘C per decade over the past 10 years. In addition, these data reveal considerable differences in change among the individual lakes, with the rates of annual average air temperature rise ranging from 0.08 ∘C per decade for Lake Yilong to 0.34 ∘C per decade for Lake Lugu, with the rates of air temperature rise per decade being greater among the lakes located in northwest Yunnan (0.2 to 0.23 ∘C) than those of lakes located in central Yunnan (0.14 to 0.17 ∘C). Overall, the rate of annual average air temperature rise has been 0.18 ∘C per decade.

The results of our analysis of the abrupt change points of seasonal air temperature in the Yunnan Plateau lakes reveal considerable differences in air temperature change between different regions and during different seasons (Figure 6). For spring, the abrupt change point occurred in 2005 in northwest Yunnan, and in 2008 in central Yunnan; the summer abrupt change point occurred after 2005 in the northwest lakes, and mainly occurred after 2010 in the central lakes; the autumn abrupt change points occurred from 1998 to 2000; and the winter abrupt change points occurred in 1997. Consistent with our previous analyses, these findings indicate that the Yunnan Plateau lakes tend to be more sensitive to global warming during the colder seasons. In addition, we found that there was an abrupt change from cold to warm in the central Yunnan lakes in 1997, which was consistent with the time of the winter abrupt change point. On the basis of these findings, it can thus be concluded that an abrupt change in the winter temperature is the main factor contributing to the abrupt change in average air temperature from cold to warm in 1997.

4. Discussion

4.1. Geographical Heterogeneity of Climate Factor Variation in Highland Lakes in Yunnan

According to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, the period from 1983 to 2012 may be the hottest 30 years in the Northern Hemisphere in the past 1400 years [29]. However, although global warming has become an accepted fact, the underlying changes in climate are characterized by geographical heterogeneity. The rate at which climate changes may differ with time in a single region, and may also vary in different regions within a single period of time [30], with climatic region and elevation being two important geographical factors influencing climate change. Studies have found that the most rapid global warming (>0.4 ∘C per decade) in recent decades has occurred in the mid-latitude air temperate regions of the Northern Hemisphere [31]. Moreover, it has been established that changes in climate have been more pronounced at high elevations, with the rate of increase in average annual temperature in the Himalayas, the region with the highest average elevation worldwide, being almost 3.5 times higher than that of the global average, 2.5 times that of China, and 1.5 times that of the Qinghai–Tibetan Plateau [25]. In southwest China, the Qinghai–Tibetan Plateau has been the first to respond to global climate change, followed by the Yunnan–Guizhou Plateau, with the Sichuan Basin and eastern Guizhou being identified as the slowest responder [32]. The plateau lakes in Yunnan have consistency with respect to genesis, but are characterized by varying degrees of diversity in terms of elevation, morphology, climate, and other factors. Whereas the average elevation of the lakes located in the northwest of Yunnan is 2055 m, that of lakes in central Yunnan is 1718 m. In terms of climate, Lake Lugu in northwest Yunnan is classified as a high-elevation temperate climate lake, whereas the lakes located in central Yunnan are low-elevation subtropical climate lakes, and lakes Chenghai and Erhai in northwest Yunnan are located in a transitional region between two climate zones, and their patterns of climate change show clear geographical heterogeneity. During the past 40 years, whereas the air temperature of all assessed lakes in the Yunnan Plateau has undergone a significant upward trend, the rate at which air temperature is rising in the lakes of the northwest of Yunnan has been significantly higher than that of the central Yunnan lakes. Moreover, the lakes in northwest Yunnan have a significantly earlier abrupt climate change than those in central Yunnan.

4.2. Possible Impacts of Climate Change on Highland Lakes in Yunnan

Climate is the main factor driving the physical, chemical, and biological processes of lakes, and in response to changes in climate, these processes undergo complex interactions [33]. With respect to the lakes of the Yunnan Plateau, the change trends are not simply confined to the common characteristics of rising air temperature and declining wind speed seen on a global scale, but also show distinct regional characteristics, with the most evident being the rising trend of air temperature and general decline in precipitation during winter, which are superimposed upon the morphological characteristics of the lakes on the Yunnan Plateau. It is predicted that ongoing climate change could have profound and complex impacts on Yunnan Plateau lakes in three main respects, namely, exacerbating water shortages, causing increases in algal biomass and growth periods, and promoting an increase in the thermal stability of deep-water lakes.

Given the fault subsidence characteristics of the plateau region, the runoff ratio of the Yunnan Plateau lakes is small, and the water resources of the lakes are limited [34]. Our findings in this study indicate that during the past 40 years, these lakes have shown a trend of increasing air temperature and solar radiation and reductions in relative humidity, which will lead to increases in the rate of surface water evaporation. These effect will be compounded by the general downward trend in precipitation in the Yunnan Plateau region, which will lead to reductions in both runoff into the lakes and direct precipitation onto the lake surfaces. Consequently, climate change may lead to an increase in evaporation and a reduced inflow of water into lakes on the Yunnan Plateau, which will further exacerbate the shortage of water resources in the lakes, potentially resulting in a reduction in water levels and a shrinkage in lake area. Indeed, analyses conducted by Li Haojie et al. [35] have indicated that water levels among the Yunnan Plateau lakes have been gradually lowering during the past 30 years. If such climate-induced changes continue in the future, this will put pressure on the water resources of the Yunnan plateau lakes, and may even promote marked ecosystem changes due to the continual reductions in water levels and area.

Further important direct effects of climate change on lake ecosystems are an increase in algal biomass and changes in algal community structure [36]. In this regard, Ho et al. [37] have established that climate warming is an important factor contributing to increases in the summer biomass of lakes on a global scale. Given differences in the adaptability of different algal populations to temperature, it is predicted that higher temperatures will be more conducive to the growth, and thus competitive ability, of cyanobacteria and other thermophilic algae, thereby leading to an increase in algal biomass and a shift in the composition of algal populations to those dominated by cyanobacteria [38]. Increases in temperature will also accelerate the rate of organic matter mineralization in lake sediments, particularly those in severely eutrophicated shallow lakes, which may promote the release of nitrogen and phosphorus from sediments into lake waters, thereby increasing nutrient contents in the water column and thus creating conditions that are more conducive to algal growth [39]. In addition, given that algae are an important source of organic matter accumulation in the sediments of eutrophic shallow lakes [40], the two aforementioned effects will contribute to generating a mutually reinforcing positive feedback, which will further accelerate the eutrophication of lakes. Climatically, the plateau lakes in Yunnan are exposed to high temperatures and strong radiation, with average summer and autumn air temperatures of 20.1 and 15.58 ∘C, respectively, and corresponding values of 17.17 and 12.39 ∘C for the lakes in northwest Yunnan. The results obtained in this study indicate that the Yunnan Plateau lakes are undergoing a long-term increase in rising air temperatures, particularly in the spring and winter, whereas wind speeds are declining, thereby providing conditions that are conducive to an increase in algal biomass during the growth season, along with an extension of the growth period of thermophilic algae such as cyanobacteria.

The thermodynamic stratification and mixing of deep-water lakes are the most important limnological characteristics influencing the vertical heterogeneity of water quality and aquatic biological components of deep-water lakes [41]. Increases in lake heat input caused by climate warming will directly modify the thermodynamic structure of deep-sea lakes [42,43], thereby resulting in a prolonged period of stratification, lower thermocline depth, and increased thermal stability [44,45,46,47], which could even result in the conversion of some seasonally stratified lakes to perennially stratified lakes. Dissolved oxygen is among the key factors necessary for the maintenance of lake ecosystem health, and the rate of vertical diffusion of substances such as oxygen in the thermocline will be significantly reduced under conditions of long-term stable stratification [48,49]. Given that dissolved oxygen does not readily diffuse to the bottom of lakes and that oxygen is continually being consumed by organic matter in lake sediments, an increase in the thermal stability of deep-water lakes will strengthen the stratification of dissolved oxygen, and thereby lead to hypoxia in the lower layers of lakes [50]. Among the Yunnan Plateau lakes, Lake Erhai is exceptional, in that it is completely thermally mixed, owing to specific geothermal conduction [51. Of the other lakes, the thermal stratification and circulation in lakes Fuxian, Lugu, Chenghai, and Yangzonghai are characterized by mixing in winter and stratification in spring, summer, and autumn, which is typical of warm single-mixed lakes [52,53,54]. Combined with reductions in wind speed, the clear increases in the autumn and winter air temperatures of the Yunnan Plateau lakes will inevitably lead to increases in thermal stability and extension of the stratification period in lakes Fuxian, Lugu, Chenghai, and Yangzonghai, whereas the semi-enclosed characteristics of the Yunnan Plateau lakes will be more conducive to the accumulation of organic matter in lake sediments. Climate change is thus predicted to have adverse effects on dissolved oxygen contents in the lower layers of deep-water lakes on the Yunnan Plateau, thereby contributing to a decline in the ecosystem health of these lakes.

5. Conclusions

In this study, ERA5 data were used to analyze variations in trends of the main meteorological factors relevant to the physical, chemical, and biological processes of nine plateau lakes in Yunnan Province during the past 40 years (1980 to 2019), including near-surface air temperature, short- and long-wave radiation, relative humidity, wind speed, and precipitation. The results revealed that the near-surface air temperature of the Yunnan Plateau lakes has shown a significant upward tendency during the past 40 years, with the largest increases occurring in winter air temperature, whereas in contrast, long-wave radiation, relative humidity, wind speed, precipitation, and other meteorological factors were all characterized by downward trends. The observed trends in climate change among the Yunnan Plateau lakes show evidence of the influence of geographical heterogeneity, with the rises in air temperature of a temperate lake (Lugu) and temperate–subtropical transitional lakes (Chenghai and Erhai) located in the northwest of Yunnan Province being significantly higher than those of subtropical lakes (Dianchi, Fuxian, Yangzonghai, Qilu, Xingyun, and Yilong) located in the central region of Yunnan Province. Conversely, the trends of decline in long-wave radiation, wind speed, and precipitation in the lakes of central Yunnan were found to be more pronounced than those recorded for the lakes of northwestern Yunnan. In addition, whereas the short-wave radiation of lakes in northwest Yunnan showed an upward trend, that of lakes in central Yunnan showed a downward trend. The abrupt change of air temperature in the highland lakes in Yunnan occurred between 2000 and 2005, among which the earliest year of abrupt air temperature change in the lakes in northwest Yunnan (Lugu, Chenghai and Erhai) was 2001, the year of abrupt air temperature change in the lakes in central Yunnan (Dianchi, Fuxian, Yangzonghai, Qilu and Xingyun) was 2002, and the year of abrupt air temperature change in Yilong Lake was 2004. We also found that these abrupt changes in air temperatures occurred earliest in the winter, followed by autumn, spring, and summer. Given that these lakes are semi-enclosed lakes of a plateau fault subsidence origin, they tend to be exposed to high air temperatures and high levels of radiation, and are characterized long water exchange cycles. In the future, climate change may further exacerbate the shortage of water resources in these plateau lakes, accompanied by increases in the biomass and growth period of algae, as well as increases in the thermal stability of deep-water lakes, which will contribute to enhancing hypoxic conditions in the lower layers of these lakes. Finally, our experiment is still flawed, as we only performed an analysis of meteorological factors, without further in-depth analysis in conjunction with the response of the lake ecosystem. In the follow-up work, we will further collect the lake water temperature and other monitoring data to explore the relationship between climate change and the response of the lake ecosystem, as well as the future impact of climate change on the lake ecosystem.