Classification of California streams using combined deductive and inductive approaches: Setting the foundation for analysis of hydrologic alteration

摘要: Regional classification of streams is an early step in the Ecological Limits Hydrologic Alteration framework. Many stream classifications are based on inductive approach using hydrologic data from minimally disturbed basins, but this may underrepresent heavily basins or sparsely gaged arid regions. An alternative a deductive approach, watershed climate, land use, and geomorphology to classify streams, miss important hydrological characteristics streams. We classified all reaches California both approaches. First, we used Bayesian hierarchical clustering according characteristics. Streams were clustered into seven classes elevation, sedimentary rock, winter precipitation. Permutation-based analysis variance random forest analyses determine which variables best separate their respective classes. Stream typology (i.e., class that reach assigned to) shaped mainly by patterns high mean flow behavior within stream's landscape context. Additionally, was reference non-reference each In contrast typology, deviation conditions more difficult detect largely defined changes low-flow variables, average daily flow, duration flow. Our combined deductive/inductive allows us estimate under compare measured order change.