A Multiscale Deep Learning Model for Soil Moisture Integrating Satellite and In Situ Data
A Multiscale Deep Learning Model for Soil Moisture Integrating Satellite and In Situ Data
复制标题
DOI：
10.1029/2021gl096847
复制
发表时间：
2022-03
影响因子：
5.2
通讯作者：
Jiangtao Liu;F. Rahmani;K. Lawson;Chaopeng Shen
中科院分区：
文献类型：
--
作者：
Jiangtao Liu;F. Rahmani;K. Lawson;Chaopeng Shen
在水文观测上训练的深度学习（DL）模型可以表现得非常好，但它们可以继承训练数据的缺陷，例如原位数据的覆盖范围有限或卫星数据的低分辨率/精度。在这里，我们提出了一种新的多尺度深度学习方案，同时从卫星和原位数据中学习，以预测每天9公里（5厘米深）的土壤湿度。基于对美国邻近地区的空间交叉验证，多尺度方案的中位数相关系数为0.901，均方根误差为0.034 m3/m3。它的表现优于土壤湿度主动式被动卫星任务的9公里产品、仅在原位数据上训练的深度学习模型和陆地表面模型。我们的9公里产品显示出比以前的1公里卫星降尺度产品更好的精度，突出了提高分辨率的有限影响。我们的产品不仅对防洪、干旱和虫害的规划有用，而且我们的方案一般适用于具有多尺度数据的地球科学领域，打破了单个数据集的限制。
Deep learning (DL) models trained on hydrologic observations can perform extraordinarily well, but they can inherit deficiencies of the training data, such as limited coverage of in situ data or low resolution/accuracy of satellite data. Here we propose a novel multiscale DL scheme learning simultaneously from satellite and in situ data to predict 9 km daily soil moisture (5 cm depth). Based on spatial cross‐validation over sites in the conterminous United States, the multiscale scheme obtained a median correlation of 0.901 and root‐mean‐square error of 0.034 m3/m3. It outperformed the Soil Moisture Active Passive satellite mission's 9 km product, DL models trained on in situ data alone, and land surface models. Our 9 km product showed better accuracy than previous 1 km satellite downscaling products, highlighting limited impacts of improving resolution. Not only is our product useful for planning against floods, droughts, and pests, our scheme is generically applicable to geoscientific domains with data on multiple scales, breaking the confines of individual data sets.