Remote sensing using continuous spatial built-up area indices

Keywords: built-up area, continuous indices, remote sensing, suitability
Abstract Type: Paper Abstract

Abstract

Urban remote sensing makes the identification and mapping of built-up areas more efficient. In current urban greening studies, major ways of urban built-up area (BUA) classification are using land use land cover (LULC) maps, nighttime light data, and built-up indices. The complexity of BUA requires identification and ranking methods to be more precise and comprehensive. However, previous methods are either categorical, binary, static, or indirect for BUA identification, which ignores mixed spectral reflectance within each satellite imagery pixel and the spatial-temporal dynamics of built-up magnitudes. To address these problems, this study proposes an integrated method using continuous data that is more informative in describing built-up magnitudes and dynamics. The continuous vegetation dataset from Landsat-8 makes weighted spatial analysis available. The proposed descriptive indices cover spectral diversity, seasonal vegetation change, and road patterns related to BUA. Looking into Miami, Atlanta, and Denver, this study investigates 1) the seasonal variation of the BUA index using the Normalized Difference Vegetation Index (NDVI), Simpson diversity of NDVI, and road density buffers for analysis, 2) suitability analysis for urban greening interventions, and 3) transect analysis that visualizes the difference between BUA and agricultural land in areas with low NDVI values. The methodological framework and results of this study will add a new perspective of BUA that informs future decision-making on urban greening and urban development.