| 摘要: |
| 为了更好的揭示长江源区积雪变化的空间异质性和气候驱动机制，将长江源区划分为沱沱河、当曲、楚玛尔河及其他4个子区域，探究各个子区域积雪指标的变化特征及与气候因子的关系。基于1980-2020年积雪及同期气温、降水数据，应用Mann-Kendall检验和ESMD方法分析子区域积雪时空变化特征，结合偏相关法分析积雪与气候因子的关系，并采用XGBoost-SHAP模型解析了积雪变化的物理机制。结果表明积雪的变化具有显著的空间异质性：雪深和雪水当量在沱沱河、当曲边缘高海拔区及楚玛尔河西北部、其他区域东南少部分区域有上升趋势，其余区域普遍下降，积雪覆盖率整体多为下降趋势；偏相关分析显示，四个子区域积雪的变化与降水和气温的变化关系并不一致：控制气温的影响，积雪与降水整体偏相关系数多为正值。控制降水，雪深和雪水当量与气温在沱沱河、当曲均为负相关，楚玛尔河西北及其他区域东南少部分区域为正相关，其余区域为负相关，积雪覆盖率与气温整体多为负相关，说明气候暖湿化对寒区积雪的影响具有一定的空间差异性；XGBoost-SHAP模型表明：历史降水状况比瞬时降水更能影响当前雪深。低温区间降水对积雪有正向贡献，温度升高，降水的正向贡献减弱直至消失，0℃左右时，积雪变化达到临界，之后温度升高，对积雪的减少作用不再加剧。 |
| 关键词: 长江源区 积雪变化 气候因子 空间异质性 XGBoost-SHAP |
| DOI： |
| 分类号:P426.63 |
| 基金项目: |
|
| Trend and causes of snow cover in the headwaters of the Yangtze River from 1980 to 2020 |
|
LI Bing-xin1, CHEN Lian-gang1, WU Nan1, TIAN Zi-he1, MIAO Qing1, MA Yiming2, LIU Yanli3
|
|
1.Nanjing Hydraulic Research Institute,Hydrology and Water Resources Department;2.China Yangtze Power Co.,Ltd.;3.Nanjing Hydraulic Research Institute
|
| Abstract: |
| Abstract: In order to better reveal the spatial heterogeneity and climate-driven mechanism of snow cover change in the source area of the Yangtze River, the source area of the Yangtze River was divided into four sub-regions: Tuotuo River, Dangqu, Chumar River and other four sub-regions, and the variation characteristics of snow cover indexes in each sub-region and their relationship with climatic factors were explored. Based on the snow cover and temperature and precipitation data from 1980 to 2020, the spatiotemporal variation characteristics of snow cover in sub-regions were analyzed by using the Mann-Kendall test and ESMD method, the relationship between snow cover and climate factors was analyzed by partial correlation method, and the physical mechanism of snow cover change was analyzed by XGBoost-SHAP model. The results showed that the changes of snow cover had significant spatial heterogeneity: snow depth and snow water equivalent increased in the high altitude area of Tuotuo River, Dangqu edge, northwest of Chumar River, and a small part of the southeast of other regions, while the rest of the area generally decreased, and the snow cover rate was mostly downward. Correlation analysis showed that the changes of snow cover in the four sub-regions were not consistent with the changes of precipitation and temperature: snow cover and precipitation as a whole were positively correlated, snow depth and snow water equivalent were negatively correlated with air temperature in Tuotuo River and Dangqu, a few areas in the northwest of the Chumar River and other regions in the southeast, and the rest of the regions were negatively correlated, and snow coverage was mostly negatively correlated with the overall temperature, indicating that the impact of climate warming and humidification on snow cover in cold regions had certain spatial differences. The XGBoost-SHAP model shows that historical precipitation conditions can affect the current snow depth more than instantaneous precipitation. Precipitation in the low temperature range has a positive contribution to snow cover, and when the temperature rises, the positive contribution of precipitation weakens until it disappears, and when it is about 0°C, the snow cover change reaches a critical point, and then the temperature rises, and the reduction effect on snow cover is no longer intensified. |
| Key words: source area of the Yangtze River watershed division changes in snow cover climatic factors XGBoost-SHAP |
