| 中国1980—2019年1 m土层贮水量的时空变化特征分析 |
| |
| 引用本文： | 王宗侠,刘苏峡,邱建秀,莫兴国.中国1980—2019年1 m土层贮水量的时空变化特征分析[J].地理研究,2022,41(11):2979-2999. |
| |
| 作者姓名： | 王宗侠 刘苏峡 邱建秀 莫兴国 |
| |
| 作者单位： | 1.中国科学院地理科学与资源研究所 陆地水循环及地表过程重点实验室，北京1001012.中国科学院大学资源与环境学院，北京 1000493.中国科学院大学中丹学院，北京 1000494.中山大学地理科学与规划学院，广州 510275 |
| |
| 基金项目： | 国家重点研发计划项目(2018YFE0106500);第二次青藏高原综合科学考察研究课题(2019QZKK0903);第三次新疆综合科学考察项目课题(2021xjkk0803) |
| |
| 摘 要： | 土壤剖面水分信息比表层土壤水分信息难以获取，但对全面认识整个土层的水分含量至关重要。融合多源数据是估算区域土壤剖面水分的有效途径。本文采用随机森林回归算法，利用中国实测土壤水分数据建立了不同季节的表层-深层土壤水分关系模型。据此采用ESA CCI SM遥感表层土壤水分产品估算获得了中国1980—2019年0~10、0~20、0~30、0~40、0~50、0~60、0~70、0~80、0~90和0~100 cm 共10个深度层次土壤水分的时空变化特征。ESA CCI SM产品与实测数据整体上匹配较好但普遍高估，本文提出采用饱和含水量和凋萎系数信息对其进行值域控制，有效降低了该产品的高估误差。随机森林回归模型的精度在秋季最高，夏季和春季次之，冬季最低。模型对干带土壤水分的估算最准确，暖温带和冷温带次之，青藏带准确性最低。计算了中国10个深度层次的土壤贮水量，其多年平均值和标准差分别为1.64±0.11、3.50±0.21、5.29±0.30、7.13±0.38、10.04±0.46、12.25±0.54、14.47±0.62、16.75±0.69、19.05±0.76和21.36±0.83 cm。各深度的土壤水分呈明显的分层，即波动层（0~40 cm）、跃变层（40~60 cm）和稳定层（60~100 cm）。中国1m土层贮水量呈自西北向东北和东南方向递增的分布格局，寒旱区该值较低且空间变异明显，暖湿区该值较高且空间分布更均一。热带、干带和青藏带的1 m土层贮水量在夏季最高，暖温带和冷温带该值在夏季最低。近40年来中国1m土层贮水量在空间上“湿区愈湿，干区愈干”，在时间上“湿季愈湿，干季愈干”。热带土壤在2004—2009年显著变湿，干带土壤显著变湿和变干的转折年份分别为1985—1986年和2013—2014年。中国1m土层贮水量序列最常见的周期是5年和11年。
|
| 关 键 词： | 表层土壤水分 土壤剖面水分 随机森林回归 Köppen-Geiger气候带 饱和含水量 凋萎系数 |
| 收稿时间： | 2021-12-09 |
Spatiotemporal variation of water storage in 1-m soil layer in China from 1980 to 2019 |
| |
| WANG Zongxia,LIU Suxia,QIU Jianxiu,MO Xingguo.Spatiotemporal variation of water storage in 1-m soil layer in China from 1980 to 2019[J].Geographical Research,2022,41(11):2979-2999. |
| |
| Authors: | WANG Zongxia LIU Suxia QIU Jianxiu MO Xingguo |
| |
| Institution: | 1. Key Laboratory of Water Cycle and Related Land Surface Processes, Institute of Geographic Sciences and Natural Resources Research, CAS, Beijing 100101, China2. College of Resources and Environment, University of Chinese Academy of Sciences, Beijing 100049, China3. Sino-Danish College, University of Chinese Academy of Sciences, Beijing 100049, China4. School of Geography and Planning, Sun Yat-sen University, Guangzhou 510275, China |
| |
| Abstract: | Profile soil moisture (PSM), which is essential for a comprehensive understanding of the entire soil layer water content, is more difficult to retrieve than surface soil moisture (SSM). It is an effective approach to estimate PSM at regional scale by combining multi-source data. Based on random forest regression (RFR), this study established SSM-deep layer soil moisture (DLSM) relationship models for different seasons with in-situ observations over China. European Space Agency Climate Change Initiative Soil Moisture (ESA CCI SM) product was used to estimate DLSM. Spatiotemporal variation of SM in 10 soil layers, i.e., 0-10, 0-20, 0-30, 0-40, 0-50, 0-60, 0-70, 0-80, 0-90 and 0-100 cm, in China from 1980 to 2019 was analyzed in detail. ESA CCI SM product matched well with in-situ observations, but the former was generally higher than the latter. A method using saturated soil water content and wilting point for range constraining was proposed, which effectively reduced the overestimation error of ESA CCI SM product. As a whole, accuracy of RFR models was the highest in autumn, followed by summer and spring, and the lowest in winter. The models performed best in arid zone (ARZ), followed by temperate zone (TEZ) and cold zone (COZ), and worst in Qinghai-Tibet zone (QTZ). The multi-year mean and standard deviation of soil water storage of 10 soil layers were 1.64±0.11, 3.50±0.21, 5.29±0.30, 7.13±0.38, 10.04±0.46, 12.25±0.54, 14.47±0.62, 16.75±0.69, 19.05±0.76, and 21.36±0.83 cm, respectively. Soil profile was divided into fluctuating layer (0-40 cm), leap layer (40-60 cm) and stable layer (60-100 cm). Water storage of 1-m soil layer (WS-1m) over China increased from northwest to northeast and southeast, with lower PSM and greater heterogeneity in cold and arid regions and higher PSM and lower heterogeneity in warm and humid regions. WS-1m in tropical zone (TRZ), ARZ and QTZ peaked in summer, while that in TEZ and COZ was the lowest in summer. Soil profile became wetter in wet zone and wet season and drier in dry zone and dry season over the last 40 years. WS-1m in TRZ significantly increased after 2004-2009, and that in ARZ increased and decreased during 1985-1986 and 2013-2014, respectively. The main cycles of WS-1m over China were 5 a and 11 a. |
| |
| Keywords: | surface soil moisture profile soil moisture random forest regression Köppen-Geiger climatic zones saturated soil water content wilting point |
|
| 点击此处可从《地理研究》浏览原始摘要信息 |
| 点击此处可从《地理研究》下载免费的PDF全文 |
|
