Assimilation And Irrigation Simulation Of Crop Models And Remote Sensing Data Based On LAI Of Maizi In Jilin

The issue of food security is related with the economic development of our country,which is also the core of agricultural production.Timely and accurately yield forecast information and monitoring of crop prospects is the key of regional and national food security,cash crop trade and management decision-making guarantee.Compared with the existing problems of economic and time cost in traditional information of agricultural survey,satellite remote sensing has the advantages of objective data,timely detection and large coverage area,but it is poorly described in the long-term growth and development of crops.Therefore,in order to improve the accuracy of simulation yield,the use of assimilation method to coupling remote sensing data with crop model is the research hotspot and key point of remote sensing and information agriculture.In this paper,the aim of this study were using of data assimilation to combine remote sensing data with crop model to improve the yield accuracy and find irrigation schemes of Jilin spring maize,specifically,the research objectives were spring maize from 2002 to 2011 in Yushu and Baicheng stations in Jilin Province.Leaf area index(LAI)was selected as the combination point of the two methods.Based on this assimilation method,the spring corn yield in the study area was simulated and analyzed.According to the simulated yield of spring maize,the use of crop water requirement and effective rainfall calculation of irrigation index to determine the key irrigation demand growth period further set irrigation schemes,on the basis of spring maize yield simulation were carried out to determine the optimal irrigation schemes.The specific research contents were as follows:(1)WOFOST model localization.It needs preliminary preparation because of some crop model inputs data cannot be measured directly.Then,the parameter sensitivity of WOFOST model was determined by the initial variable perturbation method.Based on the results of sensitivity analysis and relevant references,the varieties and soil of WOFOST model were determined parameters;the results of the model were validated from the data of crop development period,leaf area index and yield from 2002 to 2011 in Yushu and Baicheng stations,and the WOFOST model for Jilin Province stations was established.(2)Remote sensing data processing.In this paper,the remote sensing data were MODIS15 leaf area index product.The remote sensing data were easily anomalous due to unknown factors such as weather,aerosol and clouds,using the Savitzky-Golay filtering method to process the time series curve of MODIS remote sensing products for the data smoothing and removal of abnormal values.At the same time,combining with the study area leaf area measured values and Logistic equation to modify data after smoothing,get with the actual birth during the period of growth was more close to the continuous time series curve of spring maize leaf area index for the data assimilation.(3)The establishment of WOFOST-ENKF nested model.A set of Kalman filter algorithm was utilized to assimilate the MOD 15 leaf area index of 8-day time series as external assimilation data.A single point study was conducted on the Yushu.The results of the assimilation simulation yields and the yields simulation of the WOFOST model were compared.The model simulation results based on the assimilation method were closer to the measured values than the conventional model simulation results.(4)The setting of irrigation schemes and crop model simulation of main spring corn planting areas in Jilin Province.The planting in Jilin Province of spring maize is mainly restricted by the local water resources.Therefore,the study on water requirement and irrigation of maize had great significance to the steady and high yields of local grain.Through calculating the water requirement and effective precipitation during each growth period of spring corn,the index of irrigation demand during the whole growth period was obtained to determine the key irrigation periods.The irrigation amount during the growth period of different precipitation years was calculated to set up the irrigation simulation plans.WOFOST-EnKF nesting model was used to simulate spring maize yields.The results showed that spring maize in central and western Jilin Province had the highest irrigation efficiency at maturity-maturity stage,followed by seeding-emergence stage,which could increase the maturity-maturity irrigation and key irrigation conditions The number of irrigation during the growth period.