The impacts of spatial resolutions on global urban-related change analyses and modeling

Abstract

Various global land use/land cover (LULC) products have been developed to drive land-relevant climate and hydrological models for environmental assessments. However, systematic studies remain scarce that assess the uncertainties of using these products. By using a total of 16 commonly used global LULC products, we find a logarithm law of upscaling with the spatial resolution. The law reveals spatial details of urban features will be majorly distorted when using LULC products with coarser resolutions. A tipping point of the law around the 30-m resolution was identified by additional analysis of the 1-m and 10-m local land use dataset. Through the example of assessing crop production loss, we further find that most of these products will yield a significant underestimation of crop production losses, globally and locally. We conclude that the underestimated urban land rooted in most of these products would cause vital impacts on global change analyses and modeling.