青藏工程走廊沿线多年冻土区融化季高分辨率地表土壤水分反演
Thawing Period High Resolution Surface Soil Moisture Retrieval over the Permafrost Region along Qinghai-Tibet Engineering Corridor
袁启连 1赵林 2幸赞品 3邹德富 3赵建婷 1刘世博 3樊磊4
作者信息
- 1. 南京信息工程大学 地理科学学院,江苏 南京 210044
- 2. 南京信息工程大学 地理科学学院,江苏 南京 210044;中国科学院西北生态环境资源研究院,甘肃 兰州 730000
- 3. 中国科学院西北生态环境资源研究院,甘肃 兰州 730000
- 4. 西南大学 地理科学学院,重庆 400715
- 折叠
摘要
青藏高原多年冻土区地表土壤水分是该区地表各圈层之间物质和能量迁移转化的重要媒介,准确的地表土壤水分时空信息有助于分析该区土壤冻融循环过程、多年冻土退化特征以及在探讨地表土壤水分对气候变化的响应等方面具有重要科学意义.现阶段,青藏高原多年冻土发生着不同程度的退化,对区域土壤水分产生了显著影响.受限于该区高寒、缺氧以及恶劣天气的影响,青藏高原多年冻土区大范围的高分辨率土壤水分信息相对缺乏.为此,以青藏工程走廊沿线区域为研究区,以土壤水分的相关地表变量(地表温度、归一化植被指数、增强植被指数、坡度、坡向、高程、土壤质地、经纬度)为输入变量,以实测地表土壤水分为目标变量,基于随机森林回归方法分别建立了白天和夜间的土壤水分反演模型,得到了研究区2015～2018年5～9月地表土壤水分含量数据.结果表明:两个土壤水分反演模型均具有较好的反演精度,模型训练结果和验证结果与实测土壤水分含量的相关性均大于0.97,误差较小(RMSE=0.03 m3/m3),偏差接近于0 m3/m3,对各站点模拟结果较优,反演得到的地表土壤水分空间分布特征与植被分布特征一致,自东南向西北呈递减趋势.
Abstract
Surface soil moisture in the permafrost area of the Qinghai-Tibet Plateau(QTP)is an important bridge for the transfer of material and energy between the surface layers in the area.Accurate spatiotemporal in-formation of surface soil moisture is of great scientific significance for analyzing the soil freeze-thaw cycle pro-cess,permafrost degradation characteristics,and exploring the response of surface soil moisture to climate change.Under global warming,permafrost degradations have significantly influenced soil moisture over the per-mafrost regions of the QTP.The high-resolution soil moisture products over the QTP permafrost regions are limited because of the harsh natural environment.This study takes the permafrost regions along the Qinghai-Ti-bet engineering corridor as the study area,takes the relevant surface variables of soil moisture(i.e.,land surface temperature,normalized difference vegetation index,enhanced vegetation index,slope,aspect,elevation,soil texture,longitude and latitude)as the input variables,and takes the in-situ soil moisture as the target variable,to establish two kinds of soil moisture retrieval models at the daytime and nighttime based on the random forest regression method.The soil moisture products in the study area from May to September 2015 to 2018 were ob-tained.The results showed that the two soil moisture retrieval models have good retrieval accuracy.The correla-tion coefficients R between the estimated and in-situ soil moisture in the training and validation period were higher than 0.97,root mean square error both are 0.03 m3/m3,the bias both are close to 0 m3/m3.In addition,the spatial distribution characteristics of the predicted surface soil moisture were consistent with that of the vege-tation types,showing a decreasing trend from southeast to northwest.
关键词
土壤水分/青藏高原/多年冻土/高分辨率/随机森林回归Key words
Soil moisture/Qinghai-Tibet Plateau/Permafrost/High resolution/Random Forest regression引用本文复制引用
基金项目
国家自然科学基金重点项目(41931180)
出版年
2023
NETL
NSTL
维普
万方数据