High-resolution Reconstruction Of The Maximum Snow Water Equivalent Based On Remote Sensing Data And Degree-day Snowmelt Model In Mountainous Watershed

Accurate estimation of the maximum SWE in mountainous areas has always been an important concern in the fields of remote sensing and hydrology,but real-time measurements of scientific research are difficult to obtain.The reconstruction method can obtain the seasonal maximum SWE and its spatial distribution characteristics with high accuracy and resolution without any ground measured data,to provide a more reliable data source for hydrological research in mountainous and cold regions.Previous studies have used the MODIS data as the snow cover data source,and its spatial resolution of 500 meters cannot reflect the fine SWE distribution characteristics in alpine areas.In order to further improve the retrieval precision and the spatial resolution of the snow in mountains,this study reconstruct the maximum SWE accumulation of Xinjiang,the Caiertes river basin in the snowmelt season of 2017-2018,using high resolution Landsat8 and Sentinel-2to replace the conventional MODIS image to get the space and time continuous high-resolution snow cover,in combination with the reconstruction model and the restricted degree-day Snowmelt model,and evaluate the effect of different sources of parameters on the SWE valuations.Finally,the results of reconstruction in this study are compared with other SWE products.This study is the first to obtain the spatial and temporal distribution of the accumulative snow water equivalent in the snowmelt period and the maximum SWE accumulation in the early snowmelt period with the resolution of tens of meters using reconstruction technology and the method of combining remote sensing data from multiple satellites.Conclusions are as follows:(1)In the early stage of snowmelt in the study area in 2017-2018,the average maximum SWE accumulation was 377.83 mm.The snowmelt period is from early march to the end of July,and the snow water equivalent gradually decreases to 0 from low altitude to high altitude and from low latitude to high latitude.(2)In the early stage of snowmelt,the maximum cumulative snow water equivalent of the whole basin gradually increased with the increase of elevation,but there was an inflection point at the elevation of 2800-2900 m.And the distribution of snow cover in local areas is related to the redistribution of snow cover caused by gravity,such as wind-blown snow and slope.(3)The accuracy of snow cover,air temperature and radiation parameters all affect the distribution of the maximum SWE accumulation in magnitude,elevation and slope.(4)Even forcing parameter accuracy remains to be improved,compared with other snow water equivalent products,the joint usage of high-resolution Sentinel-2 and Landsat8 data can greatly improve the spatial and temporal resolution of mountainous snow and the estimated accuracy of local maximum SWE.This method provides a more accurate hydrologic model input parameters,which is of great significance to hydrology and water resources research in mountainous areas.