Using spatially explicit indicators to investigate watershed characteristics and stream temperature relationships

• Published in: The Science of the total environment 2016-05, Vol.551-552, p.376-386
• Main Authors: Grabowski, Zbigniew J., Watson, Eric, Chang, Heejun
• Format: Article
• Language: English
• Subjects: Distance-weighted; Geographic information system; Landscape; Spatial analysis; Stream temperature; Urban ecosystem
• ISSN: 0048-9697; 1879-1026
• DOI: 10.1016/j.scitotenv.2016.02.042

We generate a series of novel indicators of spatially explicit watershed permeability and runoff characteristics to examine the relationship between land cover and water temperature parameters in a rapidly urbanizing watershed. Our framework provides a readily adaptable method to examine the thermal sensitivity of streams based upon the underlying geomorphological and surface characteristics of drainage basins. Using four model groups each using a different landscape characteristic weighting scheme (Model Group 1: areal averages; Model Group 2: inverse distance by total flow length; Model Group 3: overland distance to stream network and distance squared; Model Group 4: proportional flow accumulation), we examined the predictive capacity of 19 variables, including combinations of simplified land cover, elevation, slope, and flow accumulation, on five stream thermal properties: seven day moving average of daily minimum and maximum, seasonal mean temperature, a novel metric of thermal ‘flashiness’, and total days with maximum temperature exceeding 17.8°C. We find that the use of spatially explicit landscape indicators combining watershed processes improves the performance of regressions for predicting a number of ecologically relevant stream temperature variables. Improved indicators of watershed condition lend themselves for rapid investigation of the relationship between stream thermal conditions and landscape characteristics in watersheds modified by human land uses, ultimately providing a more hydrologically meaningful indicator for the impacts of landscape change. [Display omitted] •We created novel inverse distance and flow accumulation weighted landscape metrics.•We studied the relationship between stream thermal properties and landscape metrics.•Using distance and flow accumulation weighted metrics improved model performance.•Flow accumulation weighted metrics best explained the variation in temperature.