The natural flow regime is critical to the health of riverine ecosystems. Many hydrologic metrics (HMs) have been developed to describe natural flow regimes, quantify flow alteration, and provide the hydrologic foundation for the development of environmental flow standards. Many applications require the use of models to predict expected natural values of HMs from basin characteristics at sites with no observed records of unimpaired flows. However, the error associated with HM estimation has not been evaluated. The primary goal of our study was to provide guidance for river scientists and managers in the selection, use, and interpretation of HMs for stream classification and hydroecological investigations of river ecosystems. We evaluated the predictability of a broad suite of HMs for the conterminous USA based on random forest statistical models. We also examined how the predictability of metrics varied among unique components of the flow regime. Roughly 40% of 612 HMs we examined could be predicted reliably from basin characteristics. The predictable metrics were disproportionately represented in 5 flow components: asymmetry, seasonality, magnitude, variability, and average monthly flows. Most metrics that represent extreme hydrological events (i.e., high and low flows) could not be reliably predicted. Roughly ⅔ of the evaluated HMs were incalculable or highly biased at intermittent/ephemeral streams because of the need for logarithmic transformations or scaling by other HMs, such as mean flows or percentile flow thresholds. Scaling metrics by drainage area tended to improve predictability. We recommend that the predictability of HMs be given greater consideration in studies and applications in which they are used to characterize and assess alteration of streamflow regimes.
|Publication Subtype||Journal Article|
|Title||Predictability and selection of hydrologic metrics in riverine ecohydrology|
|Series title||Freshwater Science|
|Publisher||University of Chicago Press|
|Contributing office(s)||National Research Program - Eastern Branch|
|Google Analytic Metrics||Metrics page|