Spatial and temporal variability in channel surface flow across an elevation gradient on the Colorado Front Range
Date
2018
Authors
Martin, Caroline, author
Kampf, Stephanie, advisor
Rathburn, Sara, committee member
Falkowski, Michael, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Topographic indices such as Upslope Accumulation Area (UAA) and the Topographic Wetness Index (TWI) are commonly used in watershed analyses to derive channel networks. These indices work well for large rivers and streams, but they do not always produce stream locations that match those observed in the field for headwater streams, where geology and soils affect locations of surface channels. This study maps the actively flowing drainage network of four headwater watersheds across an elevation gradient in the Colorado Front Range and examines how these locations of flow relate to topography, geology, climate, and soils. The objectives are to 1) document and digitize the active stream networks in the field, 2) delineate stream network with topographic indices and evaluate how index-derived channel networks compare with observations, and 3) evaluate how geology, climate, and soils affect surface water flow paths. Study sites are small headwater watersheds (1.7 – 15.5 km2) that vary in elevation from 1780 m up to 4190 m. At each watershed, surveys of surface water locations were conducted twice during the summer about a month apart in order to capture temporal variation. Stream densities documented during these surveys ranged from 5.09 * 10-4 m-1 at highest elevation site (3494 m – 4192 m) to 1.83 * 10-3 m-1 at lowest elevation site (1781 m – 2322 m). The lowest elevation site had the largest change in stream density between surveys, decreasing 84%. A middle elevation site that was affected by forest fire had the least change in stream density with only a 5% decrease between visits. TWI and UAA methods for deriving channel networks from topography performed well relative to field observations, ranging from 73% to 91% accuracy at low and middle elevation sites. At high elevation sites, these methods had the poorest performance, with accuracy between 21% and 74%. Also, at high elevation sites TWI performed slightly better than UAA, with 6-25% increased overall accuracy. Comparing channel networks at the four catchments, stream densities generally decreased with elevation, whereas streamflow magnitude and duration increased with elevation. Although stream density decreased with elevation, it had no apparent relationship with precipitation. The soil and bedrock geology were linked to streamflow location; in some cases, streamflow was discontinuous or dried up quickly in areas with high bedrock/soil hydraulic conductivity. Streams also followed shear zones, faults, and bedding contacts, where rocks are "weak", whereas they diverted around less erodible pegmatites. Results suggest that topography is the primary factor controlling streamflow location; however, geology and soils explained some of the cases where topographic predictions of flow location were inaccurate. Future channel delineation methods could add in a parameter based on the hydraulic conductivities of underlying soil and bedrock to improve stream channel mapping.