Assimilation of snow cover is an important method to improve the accuracy of snow simulation. However, the effects of snow assimilation are poor because satellite observed snow cover data contain erroneous information...Assimilation of snow cover is an important method to improve the accuracy of snow simulation. However, the effects of snow assimilation are poor because satellite observed snow cover data contain erroneous information, such as cloud contamination. In this paper, an improved approach is proposed to reduce the effects of observational errors during assimilation of snow cover fraction acquired by the Fengyun-3(FY-3) satellite in northeastern China. A snow depth constraint was imposed on quality control of a snow depth product from a microwave radiation imager. The assimilation experiments were carried out before and after quality control(denoted as SCFDA and SCFDA_WSD, respectively). The snow cover fraction results were evaluated against the Moderate Resolution Imaging Spectroradiometer(MODIS) snow cover products. When assimilating the snow cover fraction with the snow depth constraint(i.e., SCFDA_WSD), substantially larger improvement was obtained than that without such a constraint/quality control(SCFDA), and the deviation and root mean square error of the snow cover fraction were significantly reduced.The assimilation performance was also evaluated against in-situ snow depth observations. The SCFDA_WSD also showed greater improvements during the snow accumulation and snowmelt periods than the SCFDA. The SCFDA_WSD improvements in woodland and shrubland were the most obvious. At different altitudes, the effects of the SCFDA_WSD were basically equivalent, and the deeper the snow depth was, the better the effect. In addition, the SCFDA_WSD method was found in close agreement with the observations during a sudden snowfall event.展开更多
Snow and cloud discrimination is a main factor contributing to errors in satellite-based snow cover.To address the error,satellite-based snow cover performs snow reclassification tests on the cloud pixels of the cloud...Snow and cloud discrimination is a main factor contributing to errors in satellite-based snow cover.To address the error,satellite-based snow cover performs snow reclassification tests on the cloud pixels of the cloud mask,but the error still remains.Machine Learning(ML)has recently been applied to remote sensing to calculate satellite-based meteorological data,and its utility has been demonstrated.In this study,snow and cloud discrimination errors were analyzed for GK-2A/AMI snow cover,and ML models(Random Forest and Deep Neural Network)were applied to accurately distinguish snow and clouds.The ML-based snow reclassified was integrated with the GK-2A/AMI snow cover through post-processing.We used the S-NPP/VIIRS snow cover and ASOS in situ snow observation data,which are satellite-based snow cover and ground truth data,as validation data to evaluate whether the snow/cloud discrimination is improved.The ML-based integrated snow cover detected 33–53%more snow compared to the GK-2A/AMI snow cover.In terms of performance,the F1-score and overall accuracy of the GK-2A/AMI snow cover was 73.06%and 89.99%,respectively,and those of the integrated snow cover were 76.78–78.28%and 90.93–91.26%,respectively.展开更多
基金Supported by the National Natural Science Foundation of China(91437220)National Key Research and Development Program of China(2018YFC1506601)China Meteorological Administration Special Public Welfare Research Fund(GYHY201506002)
文摘Assimilation of snow cover is an important method to improve the accuracy of snow simulation. However, the effects of snow assimilation are poor because satellite observed snow cover data contain erroneous information, such as cloud contamination. In this paper, an improved approach is proposed to reduce the effects of observational errors during assimilation of snow cover fraction acquired by the Fengyun-3(FY-3) satellite in northeastern China. A snow depth constraint was imposed on quality control of a snow depth product from a microwave radiation imager. The assimilation experiments were carried out before and after quality control(denoted as SCFDA and SCFDA_WSD, respectively). The snow cover fraction results were evaluated against the Moderate Resolution Imaging Spectroradiometer(MODIS) snow cover products. When assimilating the snow cover fraction with the snow depth constraint(i.e., SCFDA_WSD), substantially larger improvement was obtained than that without such a constraint/quality control(SCFDA), and the deviation and root mean square error of the snow cover fraction were significantly reduced.The assimilation performance was also evaluated against in-situ snow depth observations. The SCFDA_WSD also showed greater improvements during the snow accumulation and snowmelt periods than the SCFDA. The SCFDA_WSD improvements in woodland and shrubland were the most obvious. At different altitudes, the effects of the SCFDA_WSD were basically equivalent, and the deeper the snow depth was, the better the effect. In addition, the SCFDA_WSD method was found in close agreement with the observations during a sudden snowfall event.
基金supported by the National Research Foundation of Korea(NRF)grant funded by the Korean government(MSIT)[grant number 2021R1A2C2010976].
文摘Snow and cloud discrimination is a main factor contributing to errors in satellite-based snow cover.To address the error,satellite-based snow cover performs snow reclassification tests on the cloud pixels of the cloud mask,but the error still remains.Machine Learning(ML)has recently been applied to remote sensing to calculate satellite-based meteorological data,and its utility has been demonstrated.In this study,snow and cloud discrimination errors were analyzed for GK-2A/AMI snow cover,and ML models(Random Forest and Deep Neural Network)were applied to accurately distinguish snow and clouds.The ML-based snow reclassified was integrated with the GK-2A/AMI snow cover through post-processing.We used the S-NPP/VIIRS snow cover and ASOS in situ snow observation data,which are satellite-based snow cover and ground truth data,as validation data to evaluate whether the snow/cloud discrimination is improved.The ML-based integrated snow cover detected 33–53%more snow compared to the GK-2A/AMI snow cover.In terms of performance,the F1-score and overall accuracy of the GK-2A/AMI snow cover was 73.06%and 89.99%,respectively,and those of the integrated snow cover were 76.78–78.28%and 90.93–91.26%,respectively.
