Multi-scale analysis of urban wetland changes using satellite remote sensing techniques
Metadata[+] Show full item record
This study investigates urban wetland-cover changes in the Kansas City metropolitan area with analyses at various spatial and temporal scales. Not many studies fully addressed multi-scale urban wetland-cover dynamics in both the temporal and spatial dimension. The objective was to understand how major driving factors - human disturbances and climate variation - impacted urban wetlands as determined by the scale effects of observing land-cover changes. To address this objective, multi-year and multi-season SPOT satellite images were acquired and digitally classified to generate wetland and related land-cover data over various temporal ranges. To detect long term changes of urban wetland, the study examined the landscape changes between 1992 and 2008. Furthermore, for a short term analysis over a period between 2008 and 2010, the study analyzed seasonal land-cover variation among the autumn, spring, and summer. These multi-temporal land-cover data were analyzed at various spatial scales - the metropolitan region, watersheds, sub-watersheds, specific wetland areas, and particular urban development zones. The results show that over the 16-year period, both wetland and impervious surfaces gained in area at the metropolitan level. However, the wetland change patterns were varied at other spatial scales of analysis, which were related to the dominant site-specific development activities. Further, the wetland change patterns differed if large surface water bodies (> 8ha) were excluded from the class of wetlands. The study also revealed that the seasonal change patterns of urban wetlands were likely correlated with short term precipitation conditions; but this effect may be varied depending on sampling area sizes. The study suggests that the effects of spatial and temporal scales should be considered in remote sensing detection of urban wetlands as they influence the interpretation of remotely sensed land-cover changes and correlation of driving factors. In conclusion, understanding the complex human-climate coupling factors affecting urban wetland-cover requires a multi-scale and multi-faceted analysis.
Table of Contents
Introduction -- Literature review -- Methodology -- Remote sensing analysis -- Quantifying land cover data using geospatial modelling -- Discussion and conclusion -- Appendix A. Lake inflows: historical and actual -- Appendix B. Source code for the geo-processing model -- Appendix C. Precipitation received before a satellite imaging dates -- Surface water cover analysis at Kansas City Metropolitan, watershed and sub-watershed scales