Soybean yield estimation using HJ-1 CCD data in Northeast China

In this study, soybean biomass was estimated using HJ-1 CCD data, combined with a radiation use efficiency model. HJ-1 CCD data provide time-serial parameters, with which we can get the information that describe the growth process of soybean, so inputted these information into a biomass model with the real time meteorological data, and then we could get the biomass. And then a fix value of harvest index (HI) was used to calculate the soybean yield. The results were validated with ground measure data. The coefficients of determination R² is 0.425, and the average fractional differences between estimated and observed yield of soybean is 29.73%. We concluded that estimation with remote-sensing data, such as the HJ CCD data that have a high spatial resolution and a shorter revisit cycle, can show more detail in its spatial patterns and improve the application of remote sensing on a local scale. The potential also exists for applying the approach to many other studies, including agricultural production estimation, crop growth monitoring, and agricultural ecosystem carbon cycle studies.