GEOMORPHIC IMPLICATIONS OF THE OROGRAPHIC TRANSITION FROM SNOWMELT TO RAINFALL TRIGGERED EXTREME EVENTS IN THE COLORADO FRONT RANGE

The Colorado Front Range is known for its proclivity to generate large, destructive floods. Prior hydrologic and paleo-hydrologic studies show that the largest floods are typically generated at low to intermediate elevations (below ~2300 m). This orographic transition in flood frequency corresponds to a switch from rainfall-triggered extremes at lower elevations to snowmelt-triggered extremes at higher elevations. While the detailed meteorological conditions of historic floods are well known, landscape evolution models require generalized hydrologic rules that relate runoff intensity, event duration, event location, and spatial scale to the rare floods that generate erosion. To address this in the Colorado Front Range, we analyze modern hydro-climatic data to determine how well sub-daily (i.e., peak rainfall / snowmelt rates, event durations) and spatial (i.e., size of rainfall / snowmelt source areas) attributes of runoff-producing events correlate with extreme floods (measured as the shape of the right tails of daily streamflow distributions). We find that the relative importance of hydro-meteorological attributes is directly linked to erosion process (e.g., landsliding, rilling, fluvial incision) and illustrate this using examples from the Betasso catchment in the Boulder Creek Critical Zone Observatory. Modern hydro-climate is extended to landscape evolution timescales using a stochastic-threshold model of river incision based on stream power. We find that spatial patterns in mainstem and tributary values of channel steepness are better explained when orographic gradients in runoff variability are included. This work highlights challenges to using simple algorithms in which drainage area is used as a proxy for the geomorphically effective event, particularly when the characteristic spatial scales of runoff events are smaller than the watershed area. However, we show how this limitation can be overcome by modifying river incision models to incorporate elevation-dependent stochastic parameters of storm location, spatial footprint, and runoff intensity.