| 摘要: |
| 全球变暖以及人类活动导致世界各地出现了更为频繁与强烈的热浪事件，给湖泊生态系统稳定和健康带来了潜在的挑战。全球范围内具有满足热浪事件影响研究的长序列逐日连续水温数据记录的湖泊不多。鉴于此，本文利用MODIS11A1逐日地表温度产品结合BP神经网络获取太湖2000年至2022年的逐日昼、夜表层水温，并分析太湖昼、夜间热浪特征与趋势。结果显示：基于此得到的逐日水温数据与实测值高度吻合（昼: r=0.99，RMSE=1.97 ℃；夜: r=0.98，RMSE=1.99 ℃）。在季节上，春季和夏季昼、夜间热浪较为频繁，同时持续时间较长、热浪重现天数较短；相比于昼间热浪，夏季夜间热浪表现出更为频繁、持续时间更长、强度较弱的特性。年尺度上，太湖昼、夜间热浪的频次（昼: Theil-Sen: 2 events decade-1，p<0.01；夜: Theil-Sen: 3 events decade-1 ，p<0.01）和年持续时间显著增加（昼: Theil-Sen: 17.5 days decade-1，p<0.01；夜: Theil-Sen: 22.0 days decade-1，p<0.001），热浪的重现天数缩短（昼: Theil-Sen: -29.2 days decade-1，p<0.001；夜: Theil-Sen: -28.2 days decade-1，p<0.001）。通过相关性与贡献度分析发现风速对太湖水体热浪次数、持续时间和上升与下降速率具有较高且显著的贡献（昼: 8.9%~60.5%（p<0.05）；夜: 16.4%~53.4%（p<0.05））。对于昼、夜同时发生的复合热浪事件，结果显示多发生于春季和夏季，并表现出更长的持续时间与更高的强度。本文通过太湖的研究，为其它缺少长序列实测水温数据的湖泊开展热浪规律及其生态效应的研究提供了思路和参考。 |
| 关键词: MODIS11A1 BP神经网络 逐日表层水温 热浪特征与趋势 昼夜复合热浪 |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金项目（42371016，42220104010）、江苏省科技计划项目（BK20220041） |
|
| Research on the Patterns of Heatwaves in Lake Taihu Based on BP Neural Network and Remote Sensing Inversion |
|
Yang Yifan,Yue Lintan,Deng Jianming,Lu Yingcheng,Zhu Guangwei,Qin Boqiang
|
|
1.School of Geography and Ocean Science,Nanjing University;2.School of Atmospheric Sciences, Nanjing University、Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences;3.International Institute for Earth System Science,Nanjing University
|
| Abstract: |
| Global warming and human activities have led to increasingly frequent and intense heatwaves worldwide, posing potential challenges to the stability and health of lake ecosystems. Few lakes globally have continuous, long-term daily water temperature records suitable for studying the impacts of heatwaves. In this context, this study utilized the MODIS11A1 daily surface temperature product combined with a BP neural network to derive daily diurnal and nocturnal surface water temperatures of Lake Taihu from 2000 to 2022, and analyzed the characteristics and trends of diurnal and nocturnal heatwaves in Lake Taihu. The results showed high concordance between the derived daily water temperature data and measured values (diurnal: r=0.99, RMSE=1.97°C; nocturnal: r=0.98, RMSE=1.99°C). Seasonally, heatwaves were more frequent and lasted longer in spring and summer, with shorter recurrence intervals. Compared to diurnal heatwaves, nocturnal heatwaves in summer were more frequent, lasted longer, but were less intense. On an annual scale, the frequency of diurnal and nocturnal heatwaves in Lake Taihu significantly increased (diurnal: Theil-Sen: 2 events decade-1, p<0.01; nocturnal: Theil-Sen: 3 events decade-1 , p<0.01) as did their annual duration (diurnal: Theil-Sen: 17.5 days decade-1, p<0.01; nocturnal: Theil-Sen: 22.0 days decade-1, p<0.001), while the recurrence intervals decreased (diurnal: Theil-Sen: -29.2 days decade-1, p<0.001; nocturnal: Theil-Sen: -28.2 days decade-1, p<0.001). Correlation and contribution analyses revealed that wind speed significantly influenced the number of heatwave occurrences, duration, and rate of increase and decrease in water temperature (diurnal: 8.9%-60.5% (p<0.05); nocturnal: 16.4%-53.4% (p<0.05)). Compound heatwaves occurring simultaneously during the day and night were mostly observed in spring and summer, displaying longer durations and greater intensity. This study on Lake Taihu provides insights and references for investigating heatwave patterns and their ecological effects in other lakes lacking long-term measured water temperature data. |
| Key words: MODIS11A1 BP Neural Network Daily Surface Water Temperature Heatwave Characteristics and Trends Diurnal and Nocturnal Compound Heatwaves |
