A GIS-Based Comparison of Vector and Raster Data for Landscape Environmental Assessment

Abstract

Landscape environmental assessment using Geographic Information Systems (GIS) requires spatial data that accurately represent land-cover characteristics. However, differences in spatial resolution and vector-to-raster conversion methods can introduce quantitative and geometric distortions. This study compares vector and raster data representations for landscape environmental assessment in Tomsk City, Russia, focusing on the effects of conversion methods and spatial resolutions on land-cover area estimates. PlanetScope imagery with a native spatial resolution of 3 m, acquired on August 14, 2026, was classified using the Maximum Likelihood Classification (MLC) method in ArcGIS Pro into six land-cover classes: Built Area, Crops, Flooded Vegetation, Rangeland, Trees, and Water. The resulting vector data were converted into raster format using three cell-assignment methods: Cell Center, Maximum Area, and Maximum Combined Area, at spatial resolutions of 3 m, 30 m, and 300 m. The results show that the 30 m Cell Center method produced minimal area deviations, generally below 0.1% compared with the 3 m reference. In contrast, increasing the spatial resolution to 300 m generated substantial distortions, particularly for small and fragmented land-cover classes. Under the Maximum Area and Maximum Combined Area methods at 300 m, Flooded Vegetation decreased by 87.80%, Rangeland by 77.23%, Crops by 71.58%, and Built Area by 64.10%, whereas Water increased by 158.53%. These findings demonstrate that spatial resolution and cell-assignment methods strongly influence land-cover representation through class aggregation and the loss of small spatial features. The findings support SDG 11 and SDG 15 by informing spatially accurate urban planning and terrestrial land-cover monitoring, without directly measuring SDG performance.