Water-saving Effect Of Wheat Fallow Based On Multi-Temporal Remote Sensing Data

Cultivated land fallow is an effective method to reduce groundwater consumption and conserve groundwater sources.In recent years,the national and local governments have issued relevant winter wheat fallow policies,and Hebei Province has carried out cropland fallow for many years.In this context,quantitative research on the water-saving effects of winter wheat fallow can provide data references and theoretical basis for adjusting agricultural structure,and help to adjust crop layout rationally,maintain land quality,have practical significance in water resources rationally used,and improveing agricultural water resources Rate and so on.In this paper,Hebei Province is the study area,Sentinel 2 data was using to construct a time series vegetation index,image segmentation method was applying to segment image objects,then match the time-series vegetation index of the image object with the reference vegetation index curve to extract the 2019 winter wheat planting area and winter wheat fallow area,at the same time,for the characteristics of phenological migration caused by large study area,the time-weighted dynamic time warping algorithm(TWDTW)was introduced as a matching algorithm,and the accuracy was verified with high-resolution images.The winter wheat area published by Hebei Province is used as the total verification.In view of the current situation that it is difficult to reconcile evapotranspiration in both time and space scales,this paper uses Sentinel 2 data as the main data to retrieve evapotranspiration and Landsat 8 and MODIS temperature products are fused to generate surface temperature as a supplement to Sentinel 2's lack of surface temperature.Comparing the inversion results of Sentinel 2 and Landsat 8 in the same region on the same day,it is found that the inversion results of Sentinel 2 and Landsat 8 are highly consistent in spatial distribution,numerical distribution,detail texture;The root-mean-square error of the inversion values of the two is 0.1193 which is very small,and the linear regression coefficient is 0.9575,and the detail texture of Sentinel 2 is better than the Landsat8 inversion result.Qualitative test was carried out using site evaporating dish measurements and quantitative test was carried out using site meteorological data calculations,effectively proving the feasibility of using Sentinel 2 to retrieve evapotranspiration.Based on the SEBS model,all low cloud sentinel 2 data from January to July 2019 in Hebei Province were used to calculate the surface albedo,vegetation index,leaf area index,vegetation coverage,and surface emissivity,then combined with the fusion-land-surface-temperature and meteorological data to calculate evapotranspiration on 115 dates in Hebei Province,and synthesize monthly evapotranspiration every month.The results showed that the evapotranspiration of winter wheat planting areas and fallow areas in Hebei Province increased gradually from January to June,and slightly decreased in July.Winter wheat evapotranspiration is higher than fallow land in February-June,the same in January,and lower than fallow land in July.The difference value between the evapotranspiration of winter wheat and fallow land in most cities is April>May>March>June>February>January>July.Overall,the difference in evapotranspiration between the winter wheat planting area and the fallow land during the entire growing season is 66.87 mm,which translates into an average water saving of 44 m3 per unit area.Beside evapotranspiration,irrigation of winter wheat is 180m3 per unit area,so the winter wheat has a water saving of 224m3 relative to the acres of fallow land throughout the growing season.