Multi-scale analysis of moisture patterns in urbanizing landscapes

The urbanization of ecosystems is a phenomenon of increasing importance, with ecological implications that are poorly characterized. A necessary first step in improving the characterization of these systems' behavior is developing a better understanding of how terrestrial hydrological processes function in a landscape modified by urbanization. The strong spatial heterogeneity of the landscape necessitates examining how terrestrial hydrological processes operate as functions of catchment terrain structure, land-use/land cover composition, and with respect to natural and anthropogenic moisture forcings.; Investigating the patterns and dynamics of spatially-distributed near surface soil moisture, a key measurable intermediate quantity in the hillslope hydrology portion of the hydrologic cycle, has allowed the development of a more spatially-explicit understanding of the terrestrial hydrology of urbanizing landscapes. Soil moisture sampling was conducted in study sites at the Baltimore Ecosystem Study (BES) urban Long Term Ecological Research (LTER) site, taking advantage of its field data collection efforts that are focused on urban ecosystem characterization. Land use/land cover sampling focused on creating a landscape representation designed to capture those elements of the urbanizing landscape that significantly modify process dynamics has been conducted. This sampling effort was supported through land-use/land cover analyses that incorporated remotely sensed data at multiple resolutions, in conjunction with other sources of spatial data. By applying transect sampling to digital orthophotography, information was extracted to describe urban land cover at a fine scale, and was used to develop relationships between the land use/land cover objects in the urbanizing landscape and landscape position as determined by statistics derived from digital terrain analysis. Scaling up was accomplished through the use of a measure of surface moisture conditions derived from multispectral satellite data at moderate resolutions. By modeling this remotely sensed measure of surface wetness using a radar rainfall derived antecedent moisture condition and land use/land cover information, the contributions of different components of the urbanizing landscape to surface moisture conditions have been characterized spatially at this broader scale. Together, these efforts improve our understanding of terrestrial hydrologic processes in urbanizing landscapes, taking significant steps towards supporting process-based hydroecological simulation modeling of these landscapes in the future.