全球36 km格网土壤水分逐日估算

发布时间：2018-05-20 10:00

本文选题：SMAP + 土壤水分　；参考：《地球信息科学学报》2017年06期

【摘要】：土壤水分是陆面生态系统和能量循环的核心变量之一,利用微波遥感技术获得的土壤水分产品的时间分辨率一般是2-3 d,因此精确地获得具有较高时间分辨率的土壤水分成了人们关注的焦点。本文尝试将SMAP(the Soil Moisture Passive and Active)土壤水分和MODIS光学数据相结合,利用广义回归神经网络进行全球36 km土壤水分的估算,提升SMAP土壤水分的时间分辨率。结果显示,广义回归神经网络估算土壤水分与SMAP保持了高相关性(r=0.7528),但其却保留了较高的误差(rmse=0.0914 m3/m3)。尽管如此,估算的土壤水分能够很好地保持SMAP土壤水分的整体空间变化,并且提升了土壤水分的时间分辨率(1 d)。此处,本文研究了SMAP土壤水分与MODIS光学数据之间的关系,这对今后利用机器学习进行SMAP土壤水分降尺度研究提供了重要的参考价值。
[Abstract]:Soil moisture is one of the core variables of land surface ecosystem and energy cycle. The time resolution of soil moisture products obtained by microwave remote sensing technology is usually 2-3 days, so it has become the focus of attention to accurately obtain soil moisture with high temporal resolution. This paper attempts to combine SMAP(the Soil Moisture Passive and Active) soil moisture with MODIS optical data and use generalized regression neural network to estimate the global soil water content of 36km in order to improve the temporal resolution of SMAP soil moisture. The results showed that the estimation of soil moisture by generalized regression neural network maintained a high correlation with SMAP 0.7528, but retained a high error of 0.0914 m3 / m3. Nevertheless, the estimated soil moisture can maintain the overall spatial variation of SMAP soil moisture and improve the temporal resolution of soil moisture by 1 dU. In this paper, the relationship between SMAP soil moisture and MODIS optical data is studied, which provides an important reference value for the future study of SMAP soil moisture downscaling using machine learning.
【作者单位】：北京科技大学土木与资源工程学院;武汉大学资源与环境科学学院;
【基金】：国家自然科学基金项目(41372370、41572274)
【分类号】：S152.7

【相似文献】