Preliminary Assessment of Landslides Along the Florida River Downstream from Lemon Reservoir, La Plata County, Colorado

Abstract

Nearly two-dozen shallow landslides were active during spring 2005 on a hillside located along the east side of the Florida River about one kilometer downstream from Lemon Reservoir in La Plata County, southwestern Colorado. Landslides on the hillside directly threaten human safety, residential structures, a county roadway, utilities, and the Florida River, and indirectly threaten downstream areas and Lemon Dam. Most of the area where the landslides occurred was burned during the 2002 Missionary Ridge wildfire. We performed geologic mapping, subsurface exploration and sampling, radiocarbon dating, and shallow ground-water and ground-displacement monitoring to assess landslide activity. Active landslides during spring 2005 were as large as 35,000 m3 and confined to colluvium. Debris flows were mobilized from most of the landslides, were as large as 1,500 m3, and traveled as far as 250 m. Landslide activity was triggered by elevated ground-water pressures within the colluvium caused by infiltration of snowmelt. Landslide activity ceased as ground-water pressures dropped during the summer. Shallow landslides on the hillside appear to be much more likely following the Missionary Ridge fire because of the loss of tree root strength and evapotranspiration. We used monitoring data and observations to develop preliminary, approximate rainfall/snowmelt thresholds above which shallow landslide activity can be expected. Landslides triggered during spring 2005 occurred within a 1.97 x 107 m3 older landslide that extends, on average, about 40 m into bedrock. The south end of this older landslide appears to have experienced deep secondary landsliding. Radiocarbon dating of sediments at the head of the older landslide suggests that the landslide was active about 1,424-1,696 years ago. A relatively widespread wildfire may have preceded the older landslide, and the landslide may have occurred during a wetter time. The wetter climate and effects of the wildfire would likely have resulted in increased ground-water pressures, which may have triggered the older landslide. This landslide appears to have crossed the valley floor and been subsequently eroded from this area. We found no evidence that landslide debris across the valley floor formed an impoundment of the Florida River, although it is very likely. Erosion of buttressing landslide debris from the valley floor and the lower strength of the landslide basal shear zone relative to pre-slide strength created less stable conditions than were present prior to occurrence of the landslide. However, deep ground-water conditions largely control the stability of the slope and are unknown here; hence, the potential for future deep landsliding is unknown. Additional investigation could be undertaken to further characterize landslide hazards in the area. This investigation could include episodic surveying of monuments we installed across the older landslide, obtaining detailed topographic data and aerial photography, mapping landslide debris and lacustrine deposits related to the potential former landslide dam, mapping secondary landslides, obtaining additional ages of landslide activity, constructing deep boreholes and ground-water monitoring wells, laboratory testing of soil and rock strength and hydraulic properties, and ground-water and slope-stability modeling.