Simulation And Verification Of Crop Growth Models With Different Drip Irrigation Models Of Xinjiang Spring Wheat

Wheat is one of the three major crops in China,and its output has a critical impact on China's food security.In recent years,with the continuous in-depth research and application of information technology in agriculture,the development of traditional agriculture has broken a new ground.The DSSAT(Decision Support System for Agrotechnology Transfer)model is a mechanism model based on simulating crop growth.It can comprehensively simulate crop growth and development process with many factors considered.It is mainly used for agricultural experiment analysis and yield Forecasting,production risk assessment,and climate impact assessment on agriculture.it can also be used for crop breeding,production potential simulation,and optimization of crop cultivation programs.In this thesis,DSSAT model v4.6 was used for spring wheat in drip irrigation in northern Xinjiang.Field experiments of different pipe-row ratio configuration modes were carried out in Shihezi Experimental Farm from 2018 to 2019.Model parameter adjustment,model localization,and wheat growth simulation verification studies were carried out.The spring wheat varieties to be tested were two varieties screened by the research group in the previous research: Xinchun 22 and Xinchun 44(the former is more sensitive to water stress and the latter is less sensitive).Set three different drip irrigation belt configuration modes,namely one pipe with four rows(TR4),one pipe with six rows(TR6),and one pipe with eight rows(TR8).Field management measures are consistent.The simulation first selected the test data of spring wheat's key growth period,grain yield,dry matter accumulation and distribution in 2018,and used the "GLUE" tool in the DSSAT model combined with the trial and error method to determine the genetic parameters of the crop varieties in the model,then it completed the parameter adjustment and Localize the model,using field test data of 2019 for verification.The conclusion is as follows:(1)The simulated spring wheat growth period of the model is basically consistent with the measured value: the simulation error from 2018 to the seedling stage is 2 days,and the time in2019 is completely consistent;the simulated value from the measured value in 2018 and 2019 to the three-leaf one-heart period is different from the measured value.The deviation is about3 days and 2 days respectively;the simulated and measured values of the jointing period and the flowering period in 2018 are completely consistent,and the simulation errors of the two periods in 2019 are 2 and 3 days;the simulation error of the full growth period 2018 Xinchun22 is 2 days,and the error of the Xinchun 44 is 1 day.The error in 2019 is slightly larger,which is 3 days and 4 days respectively.(2)The simulation effect of grain yield has a better effect: the simulation accuracy of Xinchun 22 in 2018 is TR4> TR6> TR8,and the simulation accuracy of Xinchun 44 is TR4>TR8> TR6.The simulation accuracy of Xinchun 22 in 2019 is the highest It is TR8,and the highest simulation accuracy of Xinchun 44 is TR6.The simulation accuracy of Xinchun 22 is TR8> TR4> TR6 and the Xinchun 44 is TR6> TR8> TR4.(3)The accuracy of wheat dry matter accumulation and distribution in the flowering stage and the mature stage is not much different in two years: the leaf simulation shows that the nRMSE is 1.29% in the flowering stage in 2019,and the nRMSE is about 3.5% in otherperiods;stem simulation results shows that the nRMSE in 2018 is relatively close,at 1.30%and 1.57% respectively,the nRMSE flowering period in 2019 is 0.51%,and the maturity period is 4.24%;the results of the simulating shell show that the simulation effect of the flowering period in 2018 is the worst,nRMSE It was 6.29%,and the mature period nRMSE was 1.64%,while the flowering stage and mature stage in 2019 were 0.88% and 2.75%,respectively;grain simulation showed that the nRMSE was 3.50% in 2018 mature stage and3.83% in 2019.The overall simulation effect is better.Based on the above research results,the DSSAT model after adjusting the variety parameters has good applicability to different spring irrigation models of Xinchun 22 and Xinchun 44 in the Shihezi area in different years,and can be used to predict local conventional and large pipe row specific drip irrigation.Spring wheat growth and grain yield under the model.