| Maize is the first crop in China,and its planting footprint is all over the world.Its planting area reaches 4.126×10~7km~2,and the annual output is 2.6067×10~8t.Maize is an important food crop and a key economic crop providing industrial raw materials.The utility of maize is not only based on the safety of rations,but also based on the consideration of feed processing and resource reserve.China is still faced with the contradiction between the rigid growth of maize demand and the hard constraints of production factors.At the same time,the outbreak of new coronavirus pneumonia has impacted the global trade system,and the interruption of the international food supply chain has intensified.Therefore,accurate,timely and large-scale acquisition of maize yield as the background data of maize resource allocation is of great strategic significance to ensure food security and scientifically guide the allocation of food resources.In this study,based on Liaoning counties and districts,statistical yield estimation was selected as the background of the method,and the growth and development characteristics of spring maize were taken as the theoretical breakthrough.Around the technical means of statistical data,meteorological data,remote sensing data integration suitability evaluation,and spectral characteristics analysis,the yield of trend was separated according to historical long-period changes,and the time series data of climate suitability and vegetation spectral index in multi-growth periods were increased.The coupling multi-source information maize yield estimation model driven by statistical prediction and crop physiological characteristics was constructed to effectively improve the accuracy of yield estimation and application stability.The main work of this study is as follows:(1)Considering the influence of chance fluctuation factors and historical trends on yield,we established a trend model by separating trend yield and fluctuation yield with the 5-point sliding average method,and searched for the development pattern of spring maize yield,the results showed that the R~2 of the trend model in each county and region reached above 0.85 and nearly0.9,and the change of maize yield increase and decrease rate in different stages echoed the current agricultural development status,indicating that the trend yield in each county and region of Liaoning can reflect the current social development trend.(2)Taking spring maize yield estimation as the anchor point,we set up a model for the suitability of important climate factors in Liaoning region,and analyzed that the suitability of precipitation,temperature and sunshine in Liaoning region has interannual spatial variability,with the characteristics of temporal and spatial variables,and shows high correlation with fluctuating yield of spring maize,with the suitability of precipitation at the tasseling stage being the most sensitive,and the suitability of temperature at the emergence stage and the suitability of sunshine at the nodulation stage being prominent,ready for use as The yield estimation factors introduced climatic information to the model.(3)Time series spectral index data set was established to measure the correlation between fluctuation yield and maize yield.Combined with maize growth mechanism,the relationship between fluctuation yield and maize yield was discussed.GNDVI at tasseling stage and EVI at filling stage,which were sensitive to canopy nitrogen content and chlorophyll content and had the highest correlation coefficient,were selected as important information to describe crop biomass and productivity.(4)The coefficients and constant terms of the coupled multi-source information yield estimation model for each county and region were determined by multiple linear regression,and compared with the simulation results of the trend model and trend+climate model,the results showed that the R~2 of the coupled yield estimation model was about 0.87 on average,and the effect was improved overall,especially in areas where agricultural disaster weather occurred.This indicates that the yield estimation model with coupled multi-source information variables constructed in this study has higher accuracy and stability,and is more applicable in practice.The work related to this study takes Liaoning as a typical microcosm to construct a semi-empirical and semi-mechanical maize yield estimation model and achieve considerable accuracy and stability,which is expected to provide others with a research basis for maize yield estimation and new thinking of yield estimation models.The paper has 25 pictures,9 tables,and 84 references. |
PDF Full Text Request