The original data mainly comes from ERA5 product data, and ERA5 products are processed. First, read the ERA5 data set (shallow soil temperature data, snow depth data, and shallow soil moisture data) in 1-hour steps, and obtain the daily average of 10km in spring in typical agricultural areas in Northeast China × Shallow soil temperature data, snow depth data and shallow soil moisture data with 10km resolution;

Test the outliers of the measured soil moisture data, shallow soil temperature data, snow depth data and shallow soil moisture data at meteorological stations in 2017 – 2018, and eliminate the outliers. Then, according to the multiple linear regression analysis in mathematical statistics, determine the relationship between the measured soil moisture data at meteorological stations in 2017 – 2018 and the shallow soil temperature data, snow depth data, shallow soil moisture data, longitude information, and latitude information, so as to obtain the daily 0 – 10cm soil moisture fitting model of typical agricultural areas in the northeast of China in spring. The model is as follows: Y=– 266.2921+0.2732 * (ERA5_Temp – 273.15)+9.2996 * ERA5_ SD + 0.0540 * ERA5_ Mois * 100 + 2.0190 * Lon+0.5245 * Lat;

Where ERA5_ Temp represents the 0 – 7cm soil temperature value of the pixel ERA5 product, ERA5_ SD represents the snow depth value of ERA5 product at the pixel point, ERA5_ Mois represents the 0 – 7cm soil humidity value of the ERA5 product of the pixel point, Lon represents the longitude of the pixel point, Lat represents the latitude of the pixel point, and Y represents the fitted 0 – 10cm soil humidity;

Finally, 10km from 2017 to 2020 is obtained according to the constructed soil moisture fitting model × The thematic map data of soil moisture in spring (from February 1 to March 20) under the influence of snow cover in the typical farmland area in northeast China with a resolution of 10km.

The data quality control stage is divided into two steps: first, the data provider checked the data accuracy of the validation data set, removed the outliers of the measured soil temperature with the residual error greater than 95%, and verified the soil moisture fitting model with the measured soil moisture value, indicating that the results have high simulation accuracy (RMSE=0.04 m3/m3);

On the other hand, unify and standardize the data storage mode and data format. In the second stage, experts are invited to review the data to verify the integrity and accuracy of the data.