Impacts of fires & floods on water infrastructure
The objectives of this project were to collect perishable data that would lead to understanding if, how, and where sediment delivered from burned and flooded basins may reduce the sustainability, defined by functional life, of reservoirs.
Project tasks involved:
A) Sediment generation rates: Estimating the relative and absolute yields of sediment throughout a wet water year following a wildfire in the Mad River basin, California, to establish sediment generation rates associated with fire;
B) Short-term sedimentation rates: Surveying bathymetry of Ruth Lake, the reservoir downstream of burned areas in the Mad River basin, to calculate short-term sedimentation rates associated with burned and unburned areas; and
C) Long-term sedimentation rates: integrate models of future sediment generation with models of trap efficiency and reservoir capacity to estimate long-term sedimentation rates due to precipitation alone and due to wildfires-flood scenarios.
A) Sediment generation rates: Sediment generation rates were developed using field observations and model simulations. Between December 2015-May 2016, we collected suspended sediment (SS) samples and made point estimates of discharge at eight tributaries to the Mad River surrounding Ruth Reservoir. The channel geometry and bed substrate was sampled for each tributary prior to and following the wet winter season. The SS samples were processed in the lab according to standard protocol. A hydrologic model (HBV) of the basin and tributaries was developed to simulate a continuous hydrologic record for converting SS samples to daily concentrations across the period of study. We also developed soil erosion models of the catchment above the dam and tributaries using RUSLE to estimate annual sediment loads for wet, dry, and post-burned years. Substantial post-processing of the field and modeled data has occurred, including developing and evaluating a new methodology for estimating sediment generation rates using a sampler (Phillips et al. 2000) that can be deployed in remote settings. In addition, we acquired SS concentration data collected by other researchers covering pre-burn years and are producing sediment generation rates for a subset of the same tributaries in a wet and a dry pre-burn year.
B) Short-term sedimentation rates: Bathymetric surveys were conducted for the entire reservoir the summer before and after the wet season. These boat-based surveys were used to develop and difference reservoir bathymetry for identifying the locations of and calculating the volume of storage lost to sedimentation in the 2016 water year. This sedimentation rate was compared with observed and modeled sediment generation rates from the tributaries and with historical sedimentation rates.
C) Long-term sedimentation rates: We integrated new components into an existing reservoir sedimentation model (Minear and Kondolf 2009) to enable analysis of impacts from wildfires and other episodic sediment events, including the ability to vary sediment load annually and the addition of decay models to reflect the exponential rate of decline in sediment production following wildfires and other major sediment disturbances (i.e. landslides, debris flows). The sedimentation model was then used to evaluate the impact of varying fire frequencies on storage capacity by introducing sediment loads associated with wildfire at a range of frequencies.
Figure 1: Representation of the percent soil burn severity class for each sub-basin, plotted with its corresponding sediment yield from the RUSLE model normalized for drainage area. Sub-basins are ordered from downstream to upstream and separated by generalized aspect.
Eco-Informatics student projects: Sediment movement across river networks following disturbances
Project sponsor and partners:
This project is funded by the National Science Foundation under award #1600016. The Humboldt Bay Municipal Water District is a project partner.