粤港澳大湾区2010-2020年耕地时空变化研究

Study on the Spatial and Temporal Changes of Cropland in the Guangdong-Hong Kong-Macao Greater Bay Area from 2010 to 2020

【目的】及时准确地提取耕地空间分布与变化信息是科学保护和有效管理耕地资源的重要技术手段,对保障区域粮食安全、制定耕地保护政策具有重要指导意义。研究粤港澳大湾区耕地时空变化,可以为粤港澳大湾区耕地保护和粮食安全保障提供数据支撑。【方法】以多时相的Landsat遥感影像为数据源,利用支持向量机方法进行土地覆盖分类,获取粤港澳大湾区2010、2015和2020年的耕地空间分布信息,分别从数量变化、空间转移和景观格局3个方面分析2010—2020年粤港澳大湾区耕地时空变化特征。【结果】数量变化上,2010—2020年粤港澳大湾区耕地表现为先少量减少后缓慢增加的变化趋势,从2010年的16155.56 km^(2)减少至2015年的15740.54 km^(2),再增加至2020年的16473.93 km^(2),其中江门耕地净增加量最多、为176.99 km^(2),东莞耕地净减少量最多、为74.68 km^(2)。空间转移上,新增耕地60.25%来源于林地,35.85%来源于水体;减少的耕地有49.53%转化为林地,43.51%转化为不透水面。景观格局上,耕地斑块密度表现为“增加—减少”的变化趋势,耕地聚集度指数表现为“下降—提高”的变化趋势。【结论】2010—2020年粤港澳大湾区耕地数量总体动态平衡;耕地增加主要来源于林地和水体,耕地减少的主要去向是转为林地和不透水面;耕地空间破碎度降低,耕地空间聚集程度提高。

【Objective】Quickly gathering information of accurate cropland distribution and spatial and temporal change is an important technical means for scientific protection and effective management of cropland resources,which is an important guiding significance for ensuring regional food security and formulating cropland protection policies.The study of spatial and temporal change of cropland in the Guangdong-Hong Kong-Macao Greater Bay Area(GBA)can provide data support for local cropland conservation and food security assurance.【Method】Based on the multi-temporal Landsat remote sensing image,the support vector machine method was used to obtain the land cover types and cropland distribution of the GBA from 2010 to 2020.Then,the characteristics of spatial and temporal changes of cropland in the GBA were analyzed from three aspects:quantitative changes,spatial transfer and landscape patterns.【Result】In terms of quantitative changes,from 2010 to 2020,the area of cropland in the GBA showed a trend of decreasing first and then increasing,from 16155.56 km^(2)in 2010 to 15740.54 km^(2)in 2015,and then increasing to 16473.93 km^(2)in 2020.At the municipal scale,the net increase of cropland in Jiangmen is the largest,with the net increase area of 176.99 km^(2),and the net decrease of cropland in Dongguan is the largest,with the net decrease area of 74.68 km^(2).In terms of spatial transfer,the area of 60.25%of the increased cropland were coming from forest,and the area of 35.85%of the increased cropland were coming from water body.Meanwhile,the area of 49.53%of the decrease cropland were converted to forest,and the area of the 43.51%of the decrease cropland were converted to impervious surface.In terms of landscape patterns,the patch density aggregation index of cropland in the GBA showed a trend of increasing and then decreasing.【Conclusion】In total,the change of cropland resources in the GBA maintained a dynamic balance from 2010 to 2020.The increased areas of cropland were mainly coming from forest and water body,and the decrease areas of cropland were converted to forest and impervious surface.Besides,landscape pattern of cropland generally tended to be less fragmented and more aggregated.