Spatio-temporal Analysis And Yield Forecast Of Soybean Production In Brazil From 2000 To 2019

Soybean is an important food crop,and its stable supply is related to national food security.However,due to the shortage of self-produced soybeans in my country,the domestic soybean consumption is highly dependent on foreign countries.This paper takes Brazil as the research object,based on the soybean yield data from 2000 to 2019,qualitatively and quantitatively analyzes the quantitative change characteristics of soybean yield in Brazil at different scales in the past 20 years,and uses the standard deviation ellipse method to study the production pattern of soybean yield at municipal level in Brazil.At the same time,using the spatial autocorrelation method to explore its spatial clustering and differentiation.Then,the grey prediction model,ARIMA model and combination model are used to forecast the soybean yield in Brazil in the next four years.The main research results are as follows:(1)In the past 20 years,Brazil’s soybean production has shown a fluctuating upward trend,but there are differences in soybean development at different scales.Among the five geographic regions,the central-western regions and southern regions are the key regions for soybean production in Brazil,and the southern region has the highest contribution to soybean production growth;The proportion of soybean yield in the northern and northeastern regions increased slightly,which is the potential area for the rapid development of soybeans;The soybean production situation in the southeast is relatively stable;Among the traditional planting areas,Mato Grosso has always maintained its status as the largest soybean state.The soybean production in the states of Pará and Piaui is increasing significantly.With the continuous expansion of their soybean planting areas,the pulling trend of soybean production in Brazil will become more and more obvious in the future;In the urban area,the number of middle-high and above production areas increased,while the number of no production areas decreased significantly.At the same time,in the past 20 years,China’s soybean output from Brazil has increased by nearly 20 times,which reflects that the development of Brazil’s soybean agriculture is in direct proportion to the demand for soybeans in my country’s soybean market.(2)In the past 20 years,soybean production in Brazil has shown a trend of spatial expansion,and there is an obvious spatial correlation between regions.Except for insignificant areas,the distribution areas are dominated by high-high aggregation areas and low-low aggregation areas.The yield distribution ellipse shows a north(easter)-south(westerly)distribution pattern,with the distribution center moving in the regions of Mato Grosso do Sul and Goias,and there is an evolutionary trend of migration to the northeast,which indicates that the spatial distribution of soybean yield tends to the central,western and southern regions,and the dominant regional pattern of soybean production is stable.At the same time,the growth rate of soybean yield in the advantageous regions has driven the development of other regions and improves the overall development level of soybean yield.Since the trade cooperation between my country and Brazil is based on the cooperation between domestic importers and soybean producers in various regions of Brazil,the spatio-temporal analysis of soybean production in various regions of Brazil is beneficial to provide more scientific and systematic import decision basis for domestic soybean importers.(3)Compared with GM(1,1)grey model,ARIMA(0,1,1)model and combination model,it is found that the combination model has the highest prediction accuracy.It predicts that Brazil’s soybean output from 2022 to 2025 will be: 138.3502 million tons,144.4670 million tons,150.6912 million tons and 157.0253 million tons.The future output will show a trend of continuous high yield,which indirectly proves the possibility of continuous trade between China and Brazil.