中国农业信息2025，Vol.37Issue(1)：1-17,17.DOI:10.12105/j.issn.1672-0423.20250101
山东省耕地质量遥感评价及预测研究
Evaluation and prediction for farmland quality based on remote sensing in Shandong Province
摘要
Abstract
[Purpose]Shandong Province is a core region for agricultural production in China.Conducting farmland quality monitoring in Shandong Province is of great significance for ensuring national food security and promoting the sustainable development of agriculture.[Method]This study selected Shandong Province as the study area,integrated Sentinel-2 remote sensing satellite imagery with topographic datasets to establish a farmland quality assessment index system based on the Pressure-State-Response(PSR)framework,and employed both random forest and classification and regression tree(CART)algorithms to develop predictive models at the prefecture-level city scale.[Result](1)Farmland quality was significantly negatively correlated with slope gradient.It was positively correlated with multispectral indices such as NDVI and REP.The correlation of the indicators was influenced by spectral characteristics.(2)The spatial distribution patterns of farmland quality was consistent with soil fertility.(3)Analysis of NDVI spatial autocorrelation showed that adjacent pixels with significant quality differences and farmland boundary areas exhibited high values,indicating spatial heterogeneity.NDVI had high potential in identifying farmland quality and land types.(4)Random forest models had higher prediction accuracy than CART algorithms,and cross-validation was effective.Indicators such as slope,SAVI,and NDWI had high importance,with their contributions related to mechanisms such as soil erosion and water supply.(5)Over the past five years,the overall distribution pattern of farmland quality remained stable.Grade 1 farmland was concentrated in the central and eastern regions of Shandong Province,while Grade 3 farmland was mostly distributed in the western and southern regions.[Conclusion]The integration of remote sensing technology and machine learning algorithms enhances the efficiency and accuracy of farmland quality assessment.This study offers a scientific basis for decision-making in agricultural production and ecological conservation.关键词
耕地质量/遥感评价指标/空间分析/预测模型Key words
farmland quality/remote sensing evaluation indicators/spatial analysis/prediction model引用本文复制引用
郑义,李威,赵增锋,张天伟..山东省耕地质量遥感评价及预测研究[J].中国农业信息,2025,37(1):1-17,17.基金项目
宁夏高等学校一流学科建设(水利工程)项目(NXYLXK2021A03) （水利工程）
