Remote sensing of within-class change in urban neighborhood structures

Two conceptual models for urban neighborhood change and urban landscape ecology are used to devise an integrative methodology for measuring temporal changes in the urban morphology of Los Angeles, California, based on the techniques of multiple endmember spectral mixture analysis (MESMA), landscape metrics, and fuzzy logic. To illustrate an application of the methodology, the research utilizes two Landsat Thematic Mapper images acquired in 1990 and 2000. Through a wall-to-wall exercise, the paper discusses: (1) how the spatially continuous character of the urban morphology in Los Angeles is analyzed through the use of MESMA technique to capture and quantify within-class changes in land cover at the sub-pixel level; (2) how the magnitude of land cover changes is assessed through fuzzy logic; and (3) how landscape metrics are applied to quantify patterns of change in land cover at the census tract level. The results of the case study indicate that the integration of MESMA, fuzzy logic and landscape metrics can provide a systematic way to derive comparable measures of change in urban neighborhoods in Los Angeles. Further research is however needed to develop more rigorous methods for assessing the accuracy of land cover fractions generated by MESMA. The paper concludes with a discussion of how the proposed methodology may present a step towards a closer integration of remote sensing technology into the reformation of planning policies concerning urban sustainable development.