The Yield Estimation Research Of Maize Under Drought Based On Remote Sensing And AquaCrop Model

Maize which is one of the major food crops in China is an important raw material in food and medicines. As water is an important factor in the growth of maize, much or little moisture both can affect the normal growth of maize. With the optimization of the field management method and fertigation strategy, crop yield has been increasing in China. But drought disasters occur frequently of arid and semi-arid area of northwest in recent years, which have caused great damage to crop production. Crop growth model can monitor the condition of crop growth by dynamically simulating in different weather conditions and different management process. It can provide decision-making for agricultural producers and can be used to estimate production of maize under drought stress.This study combined with AquaCrop model and HJ-1 remote sensing images to estimate the output of summer maize: 1) AquaCrop model validation study was based on maize measured canopy cover, biomass, and yield of Beijing Xiaotangshan laboratory in 2013; 2) The ground ASD hyperspectral data was used for estimating the maize biomass; 3) Maize yield was estimated based on the assimilation of HJ-1 remote sensing and AquaCrop model under drought stress.1) The R2 and RMSE between the simulated and the measured summer maize canopy cover, biomass and yield were 0.936 and 4.0%; 0.972 and 0.954 t·ha-1; 0.940 and 0.356 t·ha-1, respectively. The AquaCrop model was used for simulating the summer maize canopy cover, biomass and yield with the measured values and the AquaCrop model had a good applicability in the Beijing area.2) Combining with previous researchs, ground ASD hyperspectral data was used for estimation of maize biomass. The R2 and RMSE between the simulated summer maize biomass and the estimated biomass based on ground ASD hyperspectral data were 0.806 and 2.838 t·ha-1, respectively. TINDVI which was Integral values of NDVI had a good relationship with the Biomass. This research adopted the Tim's biomass estimation model which had certain rationality and estimated values of biomass were relatively close to the measured values.3) The R2 and RMSE between the assimilation and measured summer maize yield were 0.4816 and 1.224 t·ha-1, respectively. The assimilation values of the maize yield had good correlation with the actually measured values, and the assimilation method based on particle swarm optimization algorithm and AquaCrop model to calculate maize yield of Yang-ling area was feasible.