Developing perceptual models of hydrologic behavior across the North American continent

The North American continent is home to a wide range of different hydro-climates. A key research gap is that there is currently limited understanding on the spatial variability of dominant hydrologic processes across these different hydro-climates. This limited understanding makes it difficult to select computational models that faithfully represent the hydrologic processes across such large domains, yet faithful representation of the different hydro-climatic behaviors is critical for accurate numerical prediction.

Here we present progress on a synthesis of dominant hydrologic processes under different combinations of climate-terrain-human forcings, engaging the broader community of catchment and Critical Zone scientists. The product from this research will be a continental “Hydrologic Mosaic”, with each landscape in the mosaic described by a set of perceptual and conceptual models. In this first step, we produce a continental map of hydrologic landscapes defined through the juxtaposition of hydroclimate, terrain and geology, and vegetation, land use, and management. We will define hydrologically meaningful indicators of terrestrial hydrology that concisely describe a location’s (i) hydroclimate (e.g., aridity, snow fraction, energy/water seasonality), (ii) topography and geology (e.g. depth to bedrock, soil porosity, topographic slope), and (iii) vegetation, land use and management (e.g., vegetation type, agricultural drainage, reservoir size), and calculate values for these indicators for each location on the continent. We then use clustering analysis to create a manageable number of representative hydrologic landscapes.

This work functions as a starting point in a wider project, where these initial hydrologic landscapes will be refined through interactions with regional experts. Together, we will develop perceptual (sketches and descriptions) and conceptual (box-and-arrow diagrams) of the dominant processes in each hydrologic landscape. These conceptual diagrams will contribute to large-domain modeling efforts by allowing targeted model selection and comparison efforts for each hydrologic landscape.