基于特征选择和机器学习的阳澄湖叶绿素a遥感反演研究

利用2016-2020年Sentinel-2多光谱遥感影像和同步实测叶绿素a浓度数据,提出了一种基于特征选择和机器学习的叶绿素a遥感反演方法,并应用于阳澄湖.结果表明,特征选择方法在反演模型的自变量选取上具有较好的应用效果,基于此建立的随机森林模型在阳澄湖叶绿素a反演上具有较优的验证精度;2016-2020年阳澄湖叶绿...

Study on Remote Sensing Inversion of Chlorophyll A in Yangcheng Lake Based on Feature Selection and Machine Learning

By using Sentinel 2 multispectral remote sensing images and synchronously measured chlorophyll a concentration data from 2016 to 2020, a chlorophyll a remote sensing inversion method based on feature selection and machine learning was proposed and applied to Yangcheng Lake. The results indicated that the feature selection method had a good application effect on the selection of independent variables of the inversion model. The random forest model based on this had a better verification accuracy in the inversion of chlorophyll a in Yangcheng Lake. The concentration of chlorophyll a in Yangcheng Lake generally showed an upward trend from 2016 to 2020, and decreased in 2018. The concentration of chlorophyll a was relatively low in Yangcheng East Lake, and the high concentration areas concentrated in the middle and north of Yangcheng West Lake. The nutrients from inflow rivers might be the main reason for the increase of chlorophyll a concentration in this region.