摘要
探究粮食作物种植结构时空分布特征,有利于探究轮作制度实施情况,有利于进一步优化生产格局,提高经济效益,保障粮食安全。选取玉米和大豆为研究对象,以黑龙江省北安市为研究区,采用高分一号影像作为数据源,计算植被指数中的比值植被指数(RVI)、差值植被指数(DVI)、转换型植被指数(TVI)、改进型叶绿素吸收比率指数(MCARI)和灰度共生矩阵(GLCM)纹理特征中的二阶矩阵(the second moment)和熵(entropy)作为最优特征级,并结合随机森林分类算法对实验区玉米大豆进行分类。结果表明:①选取最优特征结合随机森林算法分类的总体分类精度以及单类地物分类精度能够平均达90%,Kappa系数能达到0.9;②2013-2016年大豆种植面积上涨,由2013年的582.81 km^(2)上涨到2016年的2029.42 km^(2);2017年基本与2016年持平,2018年大豆种植面积下降到925.31 km^(2);2020-2022年大豆种植面积稳定在1100 km^(2);③2013-2014年北安市玉米种植面积增大;2014-2016年玉米种植面积减少,由2014年的1687.46 km^(2)缩减到2016年的1031.7 km^(2);2017年和2018年基本与2016年持平,2020年下降到551.1 km^(2);2021年上涨到1386.77 km^(2),2022年略有下降。指出了2016年黑龙江省轮作制度实施后北安市大豆种植面积整体上呈现上涨趋势;2013-2016年北安市玉米种植面积受价格变动等因素影响较小,说明玉米的固有需求指数要高于大豆;玉米大豆复合种植带逐步形成。
The objective of this study is to explore the spatial and temporal distribution characteristics of grain crop planting structure,which is conducive to exploring the implementation of crop rotation system,further optimizing the production pattern,improving economic benefits and ensuring food security.In this paper,corn and soybean were selected as the research objects,and Bei'an City,Heilong Province was selected as the study area,and the GF-1 image was used as the data source to calculate the ratio vegetation index(RVI),differential vegetation index(DVI),transformed vegetation index(TVI),improved chlorophyll uptake ratio index(MCARI)and gray co-occurrence matrix(GLCM)texture features in the vegetation index(Entropy)was used as the optimal feature level.Combined with the random forest classification algorithm,the maize and soybean in the experimental area were classified.The results show that①The overall classification accuracy of random forest algorithm and the classification accuracy of single types of ground objects can reach an average of 90%,and the Kappa coefficient can reach 0.9.②The soybean planting area increased from 582.81 km^(2) in 2013 to 2029.42 km^(2) in 2016.In 2017,it was basically the same as in 2016,and the soybean planting area decreased to 925.31 km^(2) in 2018,and the soybean planting area stabilized at 1100 km^(2) from 2020 to 2022.③From 2013 to 2014,the maize planting area in Bei'an City increased.From 2014 to 2016,the maize planting area decreased from 1687.46 km^(2) in 2014 to 1031.7 km^(2) in 2016.In 2017 and 2018,it was basically the same as that in 2016,and decreased to 551.1 km^(2) in 2020.In 2021,it rose to 1386.77km^(2) and decreased slightly in 2022.It is concluded that after the implementation of the crop rotation system in Heilongjiang Province in 2016,the soybean planting area in Bei'an City showed an upward trend as a whole,and the corn planting area in Bei'an City was less affected by price changes and other factors from 2013 to 2016,indicating that the inherent demand index of maize was higher than that of soybean,and the maize-soybean compound planting belt was gradually formed.
作者
张卜戈
魏东岚
Zhang Buge;Wei Donglan(School of Geographical Sciences,Liaoning Normal University,Dalian 116029,Liaoning,China)
出处
《绿色科技》
2024年第9期212-216,共5页
Journal of Green Science and Technology
关键词
遥感
玉米大豆种植结构演变
农作物分类
GF-1
remote sensing
evolution of maize and soybean planting structure
crop classification
GF-1