This resource contains the data and code used in MS thesis “Comparing Scaling Approaches To Estimate Unimpaired Streamflow Timeseries And Seasonal Flow Metrics At Ungauged Streams” by Karl Christensen.

ABSTRACT

In the winter rainfall driven Mediterranean-montane climate regions of California with high seasonal and interannual variability in precipitation, river ecosystems are controlled in part by natural variation in the flow regime. Streamflow alterations that impair these natural variations often have negative impacts for aquatic species related to changes in habitat. Calculating streamflow metrics capturing specific attributes related to the timing, duration, magnitude and rate of change of the unimpaired flow regime at a daily time-step can inform environmental water management goals to maintain or restore river ecosystems by providing instream flow targets/objectives associated with specific geomorphic or ecological processes. However, unimpaired daily streamflow data, which is a necessary input of these methods, is often not readily available due to limited gauging stations and anthropogenic alteration. Process-based hydrologic modeling approaches can be used to simulate unimpaired daily streamflows but require significant parameterization and resources. Alternatively, statistical scaling approaches allow for the estimation of unimpaired streamflow time series and associated flow metrics at ungauged locations based on readily available data.
This study evaluates a suite of statistical time series scaling approaches for their ability to predict daily unimpaired flow metrics that were previously linked to key ecological functions of rivers in winter rainfall driven Mediterranean climate regions of California. Established monthly scaling methods including the drainage area ratio and standardization by means were evaluated at a daily time-step and compared with alternative scaling approaches based on dimensionless reference hydrographs and modeled monthly flows. Performance of alternative scaling approaches was evaluated across hydrologic settings and climate conditions in terms of the simulated daily streamflow time series and flow metrics calculated from these time series. When these approaches were generally applied across the State of California, results demonstrated the utility of hydrologic and water-year-type-stratification to improve statistical scaling performance and indicated that different scaling approaches are better suited to estimate different flow metrics. Aggregated dimensionless reference hydrographs accounted for spatial and inter-annual variability better than a single reference site for improved representation across large regions. This study is the first known example of combining hydrologic classifications and stream class stratified reference hydrographs to refine scaling relationships and better capture streamflow timing patterns across a large heterogeneous region. This study is intended to inform selecting the best scaling approach for a specific study region or hydrologic setting based on the specific flow metrics of interest, reference site density, and distribution, and by better prediction of unimpaired daily flow metrics, facilitate environmental water management.