摘要
多种土壤水分产品的综合评估有助于了解产品的特性与差异,对产品的算法改进及合理应用有重要意义。从空间分布,站点评估,土地覆盖类型及干湿分类等多方面对2010—2011年中国北方典型区域遥感土壤水分产品(SMOS_L3、AMSR-E_LPRM、ESACCI v04.5)和模型土壤水分产品(ECMWF_ERA5、GLDAS_Noah v2.1、GLDAS_CLSM v2.2)进行差异性及适用性分析,并从多角度讨论了影响土壤水分产品准确性的可能原因。结果表明:(1)在年尺度上,各产品均能有效表征西部干旱区土壤水分分布情况。在季节尺度上,ESACCI和3种模型产品夏秋季土壤水分较高且空间分布相似。(2)在站点评估方面,ERA5产品整体性能最优,平均相关系数R值最高为0.582,无偏均方根误差ubRMSE最低为0.045 m^(3)/m^(3)。模型产品在ubRMSE和R方面均优于遥感产品,能有效刻画站点观测的动态特征,但容易出现干湿偏差。ESACCI产品在遥感产品中准确性最高。AMSR-E与观测值之间的偏差最小(-0.015 m^(3)/m^(3)),但受天气影响其与观测值的相关性较低。SMOS产品受无线频射干扰影响,整体表现一般。(3)SMOS产品和AMSR-E产品分别对农田和林地最为敏感,其余产品在不同土地类型下土壤水分分布与实际情况基本一致且能较好地反映干湿分布情况。
The comprehensive assessment of multiple Soil Moisture(SM)products is helpful to understand the characteristics and differences of products,and is of great significance to the algorithm improvement and rational application of products.The differences and applicability of three remote sensing SM products(SMOS_L3,AMSR-E_LPRM and ESACCI v04.5)and three model-based SM products(ECMWF_ERA5,GLDAS_Noah v2.1 and GLDAS_CLSM v2.2)in typical regions of North China from 2010 to 2011 were analyzed from the aspects of spatial distribution,in-situ evaluation,land cover type and dry and wet classification.The possible reasons affecting the accuracy of soil moisture products were discussed from multi-angle.Results show that:(1)On the annual scale,all products can effectively characterize the distribution of soil moisture in the arid region of the West.On the seasonal scale,ESACCI product and three model-based SM products had high soil moisture and similar spatial distribution in summer and autumn;(2)In terms of in-situ evaluation,ERA5 product outperformed other products with the highest average Pearson correlation coefficient(0.582)and the lowest unbiased root mean square error(0.045 m^(3)/m^(3)).The model-based SM products were superior to remote sensing SM products in terms of ubRMSE and R and can effectively represent the dynamic characteristics of in-situ observations.However,the time variations range of model-based SM products was low,which may lead to dry or wet bias.ESACCI product had the highest accuracy among remote sensing SM products.AMSR-E product performed well in Bias(-0.015 m^(3)/m^(3)),but the correlation with in-situ observations was low due to the influence of weather.SMOS product was affected by Radio-frequency Interference,and its overall performance was average;(3)SMOS and AMSR-E products were sensitive to farmland and forest respectively.The soil moisture distribution of other products under different land types was consistent with the actual situation,and can show dry and wet distribution.
作者
黄钰玲
刘凯
王树东
王大成
苑峰
王保林
景文
王伟
Huang Yuling;Liu Kai;Wang Shudong;Wang Dacheng;Yuan Feng;Wang Baolin;Jing Wen;Wang wei(Aerospace Information Research Institute,Beijing 100094,China;University of Chinese Academy of Sciences,College of Resources and Environment,Beijing 100049,China;Institute of Geographic Sciences and Natural Resources Research,Beijing 100101,China;Inner Mongolia Xiaocao Digital Ecological Industry Limited Company,Hohhot 010000,China;Inner Mongolia Fengmao Technology Limited Company,Hohhot 010000,China;Hebei Finance University,Baoding 071000,China)
出处
《遥感技术与应用》
CSCD
北大核心
2022年第6期1414-1426,共13页
Remote Sensing Technology and Application
基金
国家自然科学基金项目(42141007)
内蒙古自治区科技成果转化专项资金课题(2021CG0045)。
作者简介
黄钰玲(1998-),女,福建三明人,硕士研究生,主要从事土壤水分产品评估和干旱指数研究,E-mail:huangyuling20@mails.ucas.ac.cn;通讯作者:王伟(1973-),男,河北河间人,高级工程师,主要从事遥感在环境工程中的应用研究,E-mail:228398973@qq.com。
