The research of agricultural non-point sources apportionment based on hydrology model and remote sensing technology

Non-points sources pollution is caused by rainfall or snowmelt moving over and through the ground. As the runoff moves, it picks up and carries away natural and human-made pollutants, finally depositing them into lakes, rivers, wetlands, coastal waters and ground waters. The agricultural non-point sources apportionment is the premise of non-point sources pollution prevention. Crop planting is an important part of agricultural non-point source. However, traditional non-point sources apportionment methods can’t quantify nutrient load originating from different kind of crops. The apportionment accuracy of traditional methods can’t meet the demand of more precise environment management. This study selected Shahe River Basin as the study area and used remote sensing and hydrology model to apportion the total phosphorus (TP) load aiming at establish a high-precision agricultural non-point sources apportionment method. The results indicate that classification accuracy of the Google Earth Engine (GEE) is above the 88%, the kappa coefficients are above 0.81, and classification results of different crops are credible. The main crops of Shahe River Basin include wheat-maize, chestnut, fruit, and other crops. The planting area of wheat-maize is the biggest and accounts for the 44%−67% cropland area. The planting area of fruit is the secondary and accounts for the 11%−29% cropland area. The area of wheat-maize shows a decreasing trend and the area of fruit shows an increasing trend during 2006—2012. The increased planting area of chestnut and fruit are manly from wheat-maize cropland. The farmlands of wheat-maize are mainly distributed in the south of the Shahe River Basin and the farmland of the chestnut are mainly distributed in the north of the Shahe River Basin. The Generalized Watershed Loading Function (GWLF) can simulate the streamflow and total phosphorus load of Shahe River Basin and the Nash-Sutcliffe efficiency and Correlation coefficient of simulated VS observed streamflow in calibration period is above 0.59 and 0.79 and the Nash-Sutcliffe efficiency and Correlation coefficient of simulated VS observed streamflow in validation period is above 0.92 and 0.96. The Nash-Sutcliffe efficiency and Correlation coefficient of simulated VS observed total phosphorus in calibration period is above 0.61 and 0.89 and the Nash-Sutcliffe efficiency and Correlation coefficient of simulated VS observed total phosphorus in validation period is above 0.81 and 0.94. Agricultural planting is the largest non-point source total phosphorus load in Shahe River Basin, accounting for 61% of the total phosphorus load. The resident is the secondary non-point source total phosphorus load in Shahe River Basin, accounting for 29% of the total phosphorus load. Among agricultural planting sources total phosphorus load, wheat-maize accounted for the largest proportion (52%) in total phosphorus load coming from agricultural planting activities and chestnut accounted for the second proportion (20%) in total phosphorus load coming from agricultural planting activities. Although the planting of chestnut is not the biggest the non-point source total phosphorus load, considering the increasing trend of chestnut area in recent years, there is still a risk in agricultural non-point sources pollution in Shahe River Basin. As a consequence, we also should pay enough attention for the planting of chestnut and other cash crops, which may pose a pollution risk into Yuqiao Reservoir.