受全球气候变暖的影响，青藏高原湖表温度、湖泊热浪的总天数和平均强度呈现显著增加，使得热力分层期间的湖表温度更易被加热，导致夏季湖表温度升高更快，湖表可能出现缺氧。现有研究在分析湖泊热浪的变化特征时是将较大区域内的多个湖泊的热浪特征进行平均，而青海湖的热浪特征尚不清楚。因此，本研究使用青海湖水温和湖表温度的原位观测数据、刚察气象站观测数据、MODIS地表温度观测数据、第三极地区长时间序列高分辨率地面气象要素驱动数据集（A high-resolution near-surface meteorological forcing dataset for the Third Pole region，TPMFD）和一维湖泊模式（Freshwater Lake Model，FLake）研究了青海湖 1980-2022年湖表温度的变化和热浪特征，利用相关性分析和去趋势分析法揭示了湖表温度和湖泊热浪变化的原因。研究表明：（1）TPMFD再分析数据的气温、比湿和风速与刚察气象站观测的气温、比湿和风速相关性较好且偏差和均方根误差较小，两者的相关系数分别为 0. 96、0. 84、0. 74，偏差分别为 0. 55 ℃、0. 00068 g·g-1、-0. 31 m·s-1，均方根误差分别为 0. 59 ℃、0. 00069 g·g-1、0. 38 m·s-1，TPMFD 气温的变化速率［0. 48 ℃·（10a）-1］与观测气温的变化速率［0. 44 ℃·（10a）-1］接近，TPMFD比湿的变率［0. 0001 g·g-1·（10a）-1］与观测值一致，TPMFD风速的变率［-0. 1 m·s-1·（10a）-1］较观测［-0. 25 m·s-1·（10a）-1］略小，并且TPMFD和刚察气象站的气温、比湿和风速的变化速率均通过了95%的显著性检验。模拟的青海湖湖水、湖表温度与青海湖原位观测的湖水和湖表温度相关性很好且偏差及均方根误差较小，长时间序列的模拟湖表温度与MODIS地表温度的相关性也较好且偏差和均方根误差均在合理范围，模拟结果与三种观测的相关系数分别为 0. 99、0. 96、0. 98，偏差分别为 0. 25 ℃、-0. 1 ℃、0. 87 ℃，均方根误差分别为0. 58 ℃、2. 65 ℃、2. 20 ℃。（2）1980-2022年的青海湖湖表温度和湖泊热浪特征均呈现出显著的升高趋势（p<0. 05），湖泊热浪的频次在0~6次之间波动，每年发生湖泊热浪的总天数明显增多，2022 年的总天数达到 150 天，多数年份的平均持续时间都超过了 10 d·time-1，2022年的热浪最长持续时间甚至达到76天，平均强度也显著增强，其中2016年和2022年的青海湖热浪强度等级已处于比多年平均强度等级（“中等”等级）强两个量级的“严重”等级状态。（3）气温、比湿、向下长波辐射、向下短波辐射及气压与模拟湖表温度的升高、湖泊热浪总天数的增加和平均强度的增强呈现正相关关系，而风速则与模拟湖表温度的升高呈负相关，与湖泊热浪总天数的增加和平均强度的增强呈正相关。对湖表温度的升高呈正贡献的气象要素从大到小依次为气温（23. 83%）、比湿（20. 52%）、风速（16. 05%）、向下长波辐射（14. 79%）和向下短波辐射（10. 68%）；对湖泊热浪总天数的增加呈正贡献的气象要素分别为气温（37. 54%）、风速（35. 86%）、比湿（30. 03%）、向下长波辐射（28. 27%）、向下短波辐射（27. 72%）；对湖泊热浪强度的增强呈正贡献的气象要素分别为气温（13. 25%）、风速（13. 07%）、比湿（12. 35%）、向下长波辐射（11. 05%）、向下短波辐射（10. 98%），气压则对湖表温度、湖泊热浪总天数和平均强度的升高呈现抑制作用。

Under the influence of global warming，the surface temperature of lakes on the Qinghai-Xizang Plateau，along with the total duration and mean intensity of lake heatwaves，has exhibited significant increases. These trends amplify the susceptibility of lake surface temperatures to heating during thermal stratification periods，accelerating summer warming rates and potentially inducing surface hypoxia. Previous studies analyzing lake heatwave characteristics have predominantly focused on spatially averaged metrics across broad regions， leaving the specific heatwave dynamics of Qinghai Lake poorly characterized. To address this knowledge gap， this study integrates in-situ observations of Qinghai Lake's water temperature and surface temperature，meteorological data from Gangcha Station，MODIS land surface temperature products，the Third Pole high-resolution near-surface meteorological forcing dataset（TPMFD），and simulations from the one-dimensional Freshwater Lake Model（FLake）to investigate long-term changes in surface temperature and heatwave characteristics of Qinghai Lake from 1980 to 2022. Through correlation analysis and detrended decomposition methods，the driving mechanisms underlying these changes were systematically elucidated. The research shows that：（1）The air temperature，specific humidity and wind speed of TPMFD reanalysis data are highly correlated with those observed by Gangcha meteorological station，and the biases（BIAS）and root mean square errors（RMSE）are small. The correlation coefficients of the two data are 0. 96，0. 84 and 0. 74，respectively，and the BIAS is 0. 55 ℃，0. 00068 g·g-1 and -0. 31 m·s-1，respectively. The RMSE is 0. 59 ℃，0. 00069 g·g-1 and 0. 38 m·s-1，respectively. The change rate of the air temperature in TPMFD［0. 48 °C·（10a）-1］is close to that of the observed air temperature［0. 44 °C·（10a）-1］. The variation rate of the specific humidity in TPMFD［0. 0001 g·g-1·（10a） -1］is consistent with the observed variation rate. The variation rate of the wind speed in TPMFD［-0. 1 m·s-1· （10a）-1］is slightly smaller than that of the observation［-0. 25 m·s-1·（10a）-1］. Moreover，the change rates of the air temperature，specific humidity，and wind speed in both TPMFD and the Gangcha Meteorological Station have passed the significance test at the 95% confidence level. The simulated water temperature and lake surface temperature of Qinghai Lake have a very good correlation with the in-situ observed water temperature and lake surface temperature of Qinghai Lake，and the biases and root mean square errors are relatively small. The longterm sequential simulated lake surface temperature also has a good correlation with the MODIS surface temperature，and both the BIAS and RMSE are within a reasonable range. The correlation coefficients between the simulation results and the three kinds of observations are 0. 99，0. 96，and 0. 98 respectively，the BIAS are 0. 25 °C， -0. 1 °C，and 0. 87 °C respectively，and the RMSE are 0. 58 °C，2. 65 °C，and 2. 20 °C respectively.（2）From 1980 to 2022，both the characteristics of the lake surface temperature and lake heatwaves in Qinghai Lake showed a significant increasing trend（p<0. 05）. The frequency of lake heatwaves fluctuated between 0 and 6 times. The total number of days with lake heatwaves each year increased significantly. The total number of days in 2022 reached 150 days. The mean lake heatwave duration in most years exceeded 10 days per occurrence. The maximum lake heatwave duration in 2022 even reached 76 days，and the mean lake heatwave intensity increased significantly. The intensity levels of heatwaves in Qinghai Lake in 2016 and 2022 were in a "severe" state，which is two orders of magnitude stronger than the multi-year average intensity level（"moderate" level）.（3）Air temperature，specific humidity，downward long-wave radiation，downward short-wave radiation，and air pressure are positively correlated with the increase in the simulated lake surface temperature. Moreover，they are positively correlated with the increase in the total days of lake heatwaves. Furthermore，they are positively correlated with the enhancement of the mean intensity. However wind speed is negatively correlated with the increase in the simulated lake surface temperature and positively correlated with the increase in the total days of lake heatwaves and the enhancement of the mean intensity. The meteorological factors that positively contribute to the increase in lake surface temperature，from largest to smallest，are air temperature（23. 83%），specific humidity（20. 52%）， wind speed （16. 05%），downward long-wave radiation （14. 79%），and downward short-wave radiation （10. 68%）. The meteorological factors that positively contribute to the increase in the total days of lake heat‐ waves are air temperature（37. 54%），wind speed（35. 86%），specific humidity（30. 03%），downward longwave radiation（28. 27%），and downward short-wave radiation（27. 72%）. The meteorological factors that positively contribute to the enhancement of lake heatwave intensity are air temperature（13. 25%），wind speed （13. 07%），specific humidity（12. 35%），downward long-wave radiation（11. 05%），and downward shortwave radiation（10. 98%）. Air pressure has an inhibitory effect on the increase in lake surface temperature，the total days of lake heatwaves，and the mean intensity.