Development And Optimization Of The Retrieval Models Of Soil Water Supplement In The Whole Growth Period Of Winter Wheat

There is always the exchange of matter and energy throughout the whole growth period for crops with the environment, mainly through photosynthesis, respiration and transpiration processes. Soil water is not only an important factor of soil-plant-atmosphere continuum, but also the basis to achieve series of physiological processes for crops. The variation of soil water affects transpiration, photosynthesis and biomass accumulation of crops. Excluding the influences of irrigation and precipitation, the variation of water of the soil profile (usually refers to the soil water that can be absorbed by plant roots, directly or indirectly utilized) during a certain period which can be defined as the soil water supplement. There are close links between the soil water supplement with crop biomass and leaf area index.The quantitative remote sensing has advantages on the retrieval of regional vegetation index, biomass and leaf area index, and the CERES-Wheat model can accurately simulate biomass, LAI and evapotranspiration in sample points during the whole growth period of winter wheat. Therefore, making full use of the data from remote sensing and CERES-Wheat model has important significance for improving the precision of retrieval soil water supplement.The characteristic of the soil water supplement model is the couple of the transpiration coefficient and the ratio of leaf transpiration and soil evaporation by using the remotely sensed and dynamically simulated biomass and LAI, while the coefficient and the ratio are often assigned to constant values during crop growth and development period in past studies. The model is used to quantitatively retrieve the soil water supplement in single day and the amount of results during the whole growth period of winter wheat in 2013-2014. The results in single days show that the soil water supplement retrieved on March 24 is higher than on May 11, which is consistent with the regulation of variation of the crop evapotranspiration. Also according to the spatial distribution of the soil water supplement, the numerical range of the soil water supplement is 100-200mm, and gradually decrease from the east to the west in Guanzhong Plain, which is roughly consistent with the variation trend of the water stress level in study area.During the whole growth period of crops, there is a very long stage that the vegetation coverage is low. During this stage, in order to accurately estimate the amount of crop evapotranspiration, considering the roles of soil evaporation and crop transpiration and the relationship between actual and reference evapotranspiration are both necessary. The soil water supplement model has been able to deal with the relationship between soil evaporation and crop transpiration, but the results are smaller which resulted by the low growth rate of biomass. The ratio of simulated crop coefficient and regional crop coefficient can be used as an optimization factor for the retrieval of evapotranspiration. The result of the optimized model in 2013-2014 shows that the precision has been significantly improved, especially under low vegetation coverage conditions.Using Four-dimensional Variational and Ensemble Kalman Filter assimilation algorithm to optimize the regional biomass and LAI which are inputs of the soil water supplement model. The results show that the EnKF assimilation algorithm can effectively combine the advantages of observed and simulated values to optimize the inputs of soil water supplement model. According to the point-based verification, the average relative error is from 18.4% to 14.1%, and RMSE from 23 mm to 19 mm.Further applying the soil water supplement model on 2000-2001 and 2007-2008 growth period of winter wheat, the performance confirms that model has reliable stability in different conditions, and the accuracy of the model result is closely linked with the number of remote sensing data and the information of precipitation, irrigation, surface runoff and groundwater resource.In summary, the results of the soil water supplement model in three growth periods of winter wheat have proved that the model consisting of the transpiration coefficient, the ratio of leaf transpiration and soil evaporation, the crop coefficient factor and the assimilated inputs has reliable stability and practicality.