Passive Microwave Remote Sensing Of Lake Ice Freeze-Thaw Monitoring Over High Asia

Lakes spread all over the High Asia,which is a sensitive zone for researches in global changes.The parameter of lake ice freeze-thaw is one of the key sensitive factors of global changes,and it was always expected to record and reveal the climate change furthermore and response to climate change.It is found that there is a vacancy in the study of monitoring lake ice freeze-thaw in High Asia,due to the limitations of existing remote sensing techniques and the lack of corresponding monitoring methods.Considering the characteristics of spaceborne microwave radiometer sensor transit scan feature and the traditional method inadequacies of monitoring lake ice freeze-thaw,the Nearest Neighbor method was put forward which used passive microwave data as the main data source and combined with the Landsat data and some other auxiliary data.The Nearest Neighbor method which is a method to gain the brightness temperature value of the pixel that it nearest neighbor with the lake center point when the satellite is crossed,and then constructs the lake's brightness temperature value for time series to extract the freeze-thaw information of the lake ice.It is found that the Nearest Neighbor method haves a good accuracy to detect the lake ice freeze-thaw conditions of the super large lakes such as Qinghai Lake and Siling Co.Then,based on the Nearest Neighbor method,a Pixel Unmixing method was developed to detect the lake freezing ice freeze-thaw information of the medium-large lakes.The Pixel Unmixing method was based on the proportion of the lake area and the land area in one Passive microwave pixel which as the weight coefficient to restore the lake area brightness temperature information(sub-pixel level)with linear decomposition,and this method has succeeded to extending the scope of lake freeze-thaw monitoring over the high Asian region with passive microwave remote sensing.This paper employs the MODIS optical data products to conduct a comparison validation the lake ice freeze-thaw detection of Hoh Xil Lake,Dagze Lake and Saiku Lake,and the results show that the Pearson Correlation Coefficients of judging the parameters of lake ice freeze and thaw time through two methods are 0.968 and 0.987,and it shows that the Pixel Unmixing method is proved to be reliable.At last,expanding the application of the nearest neighbor method and pixel decomposition method,the lake ice freeze-thaw change information of 51 medium-large lakes in High Asia region was obtained,and the Mann-Kendall Technique was used to quantify the change trend of lake ice freeze-thaw in 51 lakes.We analyze the change of the spatial trend from the lake ice freeze-up start,freeze-up end,break-up start and break-up end.The result shows that,from 2002 to 2016,under the influence of global climate change,the lake ice freeze-thaw change of 51 lakes in the High Asia region were spatial agglomeration in strip and less affected by latitude distribution.