Assessing land use change in Ninh Binh province using a landscape approach

Abstract

This study assesses land use changes in Ninh Bình Province from 2015 to 2020 using a landscape approach, integrating Landsat remote sensing imagery and GIS analysis. The Random Forest algorithm on Google Earth Engine was applied for image classification, land use change detection, and landscape structure analysis over time. The results indicate significant transitions among land use types. Residential and built-up land experienced the most substantial increase, expanding by 2,652.9 ha, primarily due to the conversion of agricultural land. In contrast, agricultural land declined by 5,918.5 ha as a result of urbanization. Meanwhile, aquaculture land decreased by 1,615 ha, and forested areas experienced a slight reduction of 404.6 ha due to exploitation and land-use conversion. Landscape analysis indicates an increasing degree of land use fragmentation and diversity. The SDI index rose from 1.75 to 1.89, reflecting greater diversification of land use types. Residential and construction land in Ninh Bình city showed the highest increase in CA, NumP, and ED indices, highlighting rapid urban expansion. Although agricultural land area decreased, the NumP index increased from 278 to 305, indicating fragmentation due to conversion to other land types. The study confirms that urbanization and industrialization trends are intensifying in Ninh Bình city and Tam Điệp city, while agricultural and aquaculture land is shrinking. The landscape approach provides a quantitative assessment of land use changes, supporting planning and sustainable development efforts.