Research And Application Of Collaborating Acquisition Of Evapotranspiration And Surface Soil Moisture Over Irrigated Area

Evapotranspiration and soil moisture are two essentials in the water cycle over agricultural area.The continuously spatial and temporal evapotranspiration and soil moisture,combined with the information of meteorology and crop,can help obtaining crop water shortage and crop water requirement real-timely.And then when and how much to irrigate are determined.Therefore,evapotranspiration and soil moisture provide the basis for the irrigation water planning,implementation and dynamic regulation,irrigation water monitoring and irrigation effect assessment.There are three ways to obtain evapotranspiration and soil moisture,including ground-based observation,remote sensing and modeling.However,neither one could satisfy the demand on the time and space scale in agricultural water management.Assimilating remote sensing or ground measurements into model can achieve complementary advantages of remote sensing,ground-based observation and modeling.Nowadays,there are many remote-sensing inversion methods for evapotranspiration and soil moisture.And abundant remote-sensing products for soil moisture are free to use.However,when combined with agricultural water management business applications,these methods lack a csystematic,in-depth analysis and comprehensive integration.Problems to be solved are as follows:(1)Evapotranspiration at the field sacle are lacking in agricultural water management;(2)There are a variety of data resources for water cycle elements,but the spatial and tempral scale of these data resources are different,so it is necessary to solve the problem of spatio-temporal fusion of multi-source data;(3)The existing methods for retrieving evapotranspiration and soil moisture tend to focus on a single element to seek model improvement,ignoring the relationship between the two water cycle elements.Aiming at solving the problems mentioned above,some research were conducted,including:(1)Considering the influence of water stress on evapotranspiration,the soil moisture indicator was coupled with the SEBS to construct the SM-SEBS,which is suitable for evapotranspiration irn arid and semi-arid areas;(3)Considering the complementary advantages in temporal and spatial resolution of multi-source remote sensing images,the space-time adaptive fusion algorithm(STARFM)is used to spatially downscale evapotranspiration to obtain field evapotranspiration with high temporal resolution and high spatial resolution;(4)Based on the Noah-MP and EnKF,the simultaneous assimilation model of evapotranspiration and soil moisture was constructed.And the data assimilation simulation effect under the three schemes of assimilation including assimilating evapotranspiration,assimilating soil moisture and simultaneously assimilating evapotranspiration and soil moisture were deeply discussed,respectively.The research was conducted in the Ganzhou Irrigation District of Heihe River Basin,Northwest China.And some conclusions were drawn,including:(1)The modified SEBS(SM-SEBS)model improved the underestimation of sensible heat and overestimation of evapotranspiration in the original SEBS model under soil water stress conditions;(2)Based on the improved STARFM algorithm,the field evapotranspiration with high temporal and high spatial resolution was obtained.The results show that the improved STARFM model has higher integration precision and a 0.5 mm and 4.3%improvement in RMSE(Root Mean Square Error)and MAP(Mean Absolute Percentage Error),respectively;(3)Based on the Noah-MP and EnKF,a combined assimilation scheme of evapotranspiration and soil water content was constructed,and the assimilation scheme was run on the field scale to obtain evapotranspiration in a continuous field-time scale.In the sensitivity analysis of model parameters,the sensitivity of the parameters was examined from the two aspects of evapotranspiration and soil water content.The parameter sets which were sensitive to both state variables were selected.And the sensitivity of EnKF was also analyzed.The best solution for data assimilation was established from the aspects of atmospheric disturbance setting,background error,observation error and aggregate data;(4)A comparative analysis of Noah-MP simulation,assimilation of evapotranspiration,assimilation of soil moisture,and simultaneous assimilation evapotranspiration and soil moisture was conducted.The results showed that the evapotranspiration and soil moisture obtained by the combined assimilation were better than the single factor assimilation and for a certain period of time,and the combined assimilation had better stability.The combined assimilation of evapotranspiration and soil moisture can obtain accurate,time-space continuous evapotranspiration and soil moisture,thus providing data for agricultural water management.