Abstract:Abstract. The dynamic characteristics of seasonal snow cover are critical for hydrology management, climate system, and ecosystem function. Although optical satellite remote sensing has proved to be an effective tool for monitoring global and regional variations of snow cover, it is still problematic to accurately capture the snow dynamics characteristics at a finer spatiotemporal resolution, because the observations from optical satellite sensors are seriously affected by clouds and solar illumination. Besides, traditional methods of mapping snow cover from passive microwave data only provide binary information with a 25-km spatial resolution. In this study, we first present an approach to predict fractional snow cover over North America under all-weather conditions, derived from the enhanced resolution passive microwave brightness temperature data (6.25 km). This estimation algorithm used Moderate Resolution Imaging Spectroradiometer (MODIS) snow cover products between 2010 and 2017 to create the reference fractional snow cover data as the "true" observations. Further, the influence of many factors, including land cover, topography, and location, were incorporated into the retrieval models. The results show that the proposed retrieval models based on random forest regression technique perform much better using independent test data for all land cover classes, with higher accuracy and no out-of-range estimated values, when compared to the other three approaches (linear regression, artificial neural networks (ANN), and multivariate adaptive regression splines (MARS)). The results of the output evaluated by using independent data indicate that the root-mean-square error (RMSE) of the estimated fractional snow cover ranges from 16.7 % to 19.8 %. In addition, the estimated fractional snow cover is verified in the snow mapping aspect by using snow cover observation data from meteorological stations (more than 0.31 million records). The result shows that the binary snow cover obtained by the proposed retrieval algorithm is in a good agreement with the ground measurements (kappa: 0.67). The accuracy of our algorithm estimation in the snow cover identification shows significant improvement when benchmarked against the Grody’s snow cover mapping algorithm: overall accuracy is increased by 18 % (from 0.71 to 0.84), and omission error is reduced by 71 % (from 0.48 to 0.14). Daily time-series and full space-covered sub-pixel snow cover area data are urgently needed for climate and reanalysis studies. According to our experiment results, we can conclude that it is feasible for estimating fractional snow cover from passive microwave brightness temperature data, and this strategy also has a great advantage in detecting snow cover area.

An Effective Fractional Snow Cover Estimation Method Using Deep Feature Snow Index

Comparison of Snow Indexes in Estimating Snow Cover Fraction in a Mountainous Area in Northwestern China.

Monitoring Snow Cover in Typical Forested Areas Using a Multi-Spectral Feature Fusion Approach

Fractional Snow Cover Mapping with High Spatiotemporal Resolution based on Landsat, Sentinel-2 And Modis Observation

A Spatially‐Distributed Machine Learning Approach for Fractional Snow Covered Area Estimation

IMPROVED ON SNOW COVER EXTRACTION IN MOUNTAINOUS AREAS BASED ON MULTI-FACTOR NDSI DYNAMIC THRESHOLD

Downscaling MODIS NDSI to Sentinel-2 fractional snow cover by random forest regression

A novel fine-resolution snow depth retrieval model to reveal detailed spatiotemporal patterns of snow cover in Northeast China

Snow Cover Monitoring With Chinese Gaofen-4 Pms Imagery And The Restored Snow Index (Rsi) Method: Case Studies

Estimating fractional snow cover from passive microwave brightness temperature data using MODIS snow cover product over North America

Binary and Fractional MODIS Snow Cover Mapping Boosted by Machine Learning and Big Landsat Data

Estimating AVHRR snow cover fraction by coupling physical constraints into a deep learning framework

Estimation of High-Resolution Fractional Tree Cover Using Landsat Time-Series Observations

Gap-Filling of MODIS Fractional Snow Cover Products via Non-Local Spatio-Temporal Filtering Based on Machine Learning Techniques

An Approach to Improve the Spatial Resolution and Accuracy of AMSR2 Passive Microwave Snow Depth Product Using Machine Learning in Northeast China

PCSSR‐DNNWA: A Physical Constraints Based Surface Snowfall Rate Retrieval Algorithm Using Deep Neural Networks With Attention Module

Snow depth mapping in agricultural areas of Northeast China based on deep learning and multi-temporal Sentinel-1 data

Reconstruction of Snow Cover in Kaidu River Basin Via Snow Grain Size Gap-Filling Based on Machine Learning

Support Vector Machine-Based Decision Tree for Snow Cover Extraction in Mountain Areas Using High Spatial Resolution Remote Sensing Image

DSRSS-Net: Improved-Resolution Snow Cover Mapping from FY-4A Satellite Images Using the Dual-Branch Super-Resolution Semantic Segmentation Network

A New Operational Northern Hemisphere Snow Water Equivalent Retrieval Algorithm for FY-3F/MWRI-II Based on Pixel-Based Regression Coefficients