Simulation Of Spatial Heterogeneity Of Winter Wheat Yields Based On Gridded Parameteriaztion And Simulation System Of Crop Model

This research aims to establish a crop model simulation system based on grid data.and then evaluate and apply it.Firstly,analyzed the sensitivity of the model parameters to the model output of the CERES(Crop Environment Resource Synthesis)-Wheat model under different water deficit conditions;on this basis,established the DSSAT-PEST software package and evaluated the performance of the software package.Established a crop model simulation system based on grid data,and used the DSSAT-PEST software package to calibrate and verify the parameters of winter wheat in the Loess Plateau;then used the calibrated system to study the multi-year yield difference of winter wheat in the Loess Plateau.The main conclusions of the research were as follows:(1)Sensitivity evaluation of CERES-Wheat model parametersThe CERES-Wheat model was used to study the parameter sensitivity of different drought treatment experiments in the winter wheat growing season.Variety parameters and soil parameters were analyzed.A total of 24 parameters were used to evaluate the Sensitivity of outputs of phenological period(anthesis period and maturity period),maximum leaf area index,and biological parameters.By evaluating the results of each treatment,it was found that the sensitivity parameter of the CERES-Wheat model to the anthesis period of winter wheat was P1D,and the parameters sensitive to the maturity period were P1D and P5,and there was no difference between the treatments.For the maximum leaf area index,the parameter sensitivity differed between treatments.As the drought period passed,the sensitivity of SLLL1 gradually decreased,and the importance of G2 and PHINT gradually increased.For yield,the more sensitive parameter was P1D,and for biomass,the parameters P1D and PHINT were more sensitive.Generally speaking,the parameters that control the phenological period were more sensitive,and the soil parameters that control moisture were more sensitive.This research not only to analyzed the sensitivity of the output value at the end of the year,but also analyzed the leaf area index and biomass of the output value of the time series by used CERES-Wheat model.As a result,the sensitivity of the 24 parameters to output changed over time during the total growth period.For the leaf area index,the sensitive parameter in the early growth period was SLLL1,the sensitive parameter in the middle period was PHINT,and the sensitive parameter in the later period was P1D.For biomass,the more sensitive parameter in the early stage was SLLL1,while the parameter sensitivity of PHINT gradually increases in the later stage.At about the jointing stage,the sensitivity of PHINT rised to the maximum,and then begins to decreased.At the same time,it was also at the jointing stage.Sensitivity of P1D was gradually rises.(2)Design and evaluation of model parameters automatic optimization software(DSSAT-PEST)Quickly determining accurate crop genetic parameters for crop model applications can be difficult.In this study,we coupled the independent automatic parameter optimization tool PEST(Parameter ESTimation)with the crop growth model of DSSAT(Decision Support System for Agrotechnology Transfer)using the R programming language.A new DSSAT-PEST package was developed to perform automatic optimization of the crop genetic parameters.In addition,the PEST tool was modified to reduce problems associated with local optima and model runtime.The DSSAT-PEST package was used to estimate the genetic coefficients for five crops(i.e.,maize(Zea mays L.),soybean(Glycine max L.Merrill),wheat(Triticum aestivum L.),rice(Oryza sativa L.),and cotton(Gossypium hirsutum L.))based on existing experiments in the DSSAT database.Three parameter optimization methods were compared based on their efficiency and accuracy for estimating crop genetic parameters:1)the traditional trial-and-error method(default crop genetic parameters in the DSSAT database);2)DSSAT-GLUE(general likelihood uncertainty estimation,an existing parameter estimation package in DSSAT),and 3)DSSATPEST.The DSSAT-PEST optimization method produced reasonably accurate optimization results and improved optimization efficiency compared with the other two methods.For example,the average absolute relative error(AREs)between relevant field observations and model simulations obtained with DSSAT-PEST were 12%,7%,18%,4%,and 19%for the five crops,respectively,which were similar to or better than the results with DSSATGLUE and the default method.Additionally,average runtime for DSSAT-PEST was about 65%of the runtime for DSSAT-GLUE.In general,the DSSAT-PEST package performed similarly to or better than the traditional trial-and-error method and DSSAT-GLUE in terms of both optimization efficiency and accuracy,which should promote wider application of the DSSAT model in agricultural and environmental research.(3)Gridding of CERES-Wheat model at regional scaleIn the regional application of the model,calibration and verification are essential and important links.In the grid crop model,the calibration and verification data also needed to use the grid regional measured data.Using the measured phenological data of stations for many years,the regression model was established,and the regression model was calibrated and verified.The calibrated model predicted the temporal and spatial distribution of regional phenology,and expanded the station phenology data into spatial grid phenology data.Because there were few factors affecting phenological period,the prediction accuracy of phenological period was high,and the fitting degree of flowering period and mature period was good.The R²of the fitting curve of measured value and simulated value could reach more than 0.75,among which the simulation accuracy of mature period was higher.In terms of spatial distribution,elevation was a sensitive parameter to the value of phenological period,which was positively correlated with the length of phenological period.In terms of time series,the phenology obtained by simulation fluctuates was less in time series,and the standard deviation was less than 1 day.This provided data support for the calibration and verification of spatial model in the next step.In the regional application of the model,calibration and verification were indispensable and important part.In large-scale large areas,the grid points increased exponentially,the trial-and-error method was inefficient,and it was difficult to optimize the grid parameters at the regional scale.Although the automatic calibration method did not require huge manpower input,when the automatic calibration method was applied to a large area and a large scale,it required more calculations and iterations than the site scale.At the regional scale,a large number of parameter optimizations were required to increase the optimization time and increase the difficulty of parameter optimization.This was a problem that must be solved in the application of the regional scale model.The DSSAT-PEST software package was improved,and the Alibaba Cloud platform was used for calculation,which reduced the calculation time to about 1.8%of the original time,which greatly improved the efficiency and provides technical support for regional calibration verification.Using the optimized DSSAT-PEST software package,the parameters of winter wheat crops at 1,718 grid points in the Loess Plateau region of the study area were optimized,and the flowering period and phenological period were calibrated and verified on a grid scale,and the whole area was calibrated during the flowering period.The average error of verification were 5.2 days and6.0 days,and the average error of full-region calibration verification during the mature period were 5 days and 6 days.The average value of the simulated and measured output values were 3636 kg ha^-1and 3437 kg ha^-1,and good parameter optimization results had been obtained.(4)Simulation and analysis of regional gridding yield gapThe grid crop model simulation system was used for application analysis to calculate the yield under the stress and potential yield of winter wheat of many years in the Loess Plateau,and calculate the yield gap between the two yield.The yield gap in most regions of the whole region was between 4000-7500 kg ha^-1.The potential yield of most regions was an upward trend for many years,and a small part was a downward trend.Under the stress yield of many years,the yield of most regions was an upward trend.The yield gap between the two showed an upward trend in about half of the regions and a downward trend in half.Generally speaking,there was still about 52%for improvement in winter wheat yield.This article mainly used multi-source data to establish a grid crop model simulation system,and conducted parameter calibration verification on it,and then applied it.The established grid crop model simulation system could have good simulation calibration verification results for regional phenology and yield,and could obtain regional large-scale,long-term grid simulation results.The application had important and important theoretical and practical significance,and it had a positive effect on the application of crop models at the regional scale.It is improve of regional scale simulation accuracy,and the further construction of a large-scale crop planting management decision-making system.