This paper aims to efficiently invert the water turbidity of Xiaoxingkai Lake.The study is focused on the efforts to introduce dummy variables and spatial autoregression theory to optimize the re-mote sensing inversion model of water turbidity of Xiaoxingkai Lake for its higher accuracy on the basis of the ordinary regression model.The results show that the spatial autoregressive model with dummy varia-bles can effectively improve the turbidity inversion accuracy.Compared with the ordinary regression mod-el,the dummy variable model and spatial autoregressive model,the R2 increases by 12.39%、6.55%and 1.97%,respectively,and the eRMSE is reduced by 39.59%、30.21%and 1.45%,respectively.The water turbidity distribution of Xiaoxingkai Lake presents a trend of high in the north and low in the south,which is consistent with the distribution law of water turbidity,that is higher near bank and lower far from bank.In May,the water turbidity distribution is more balanced,with the high turbidity value accumula-ted in Chengzi River estuary,Xingkai Lake farm and other areas,and in August,the water turbidity is higher than that in May.
关键词
水质反演/空间自回归/小兴凯湖/哑变量
Key words
remote sensing retrieval of water quality/spatial autoregressive model/Xiaoxingkai Lake/dummy argument
维普
万方数据