Application of SIBERIA landscape evolution model to assess long-term erosion at sites within Los Alamos National Laboratory
Since October 2009, Portage, Inc., has performed modeling of long-term erosion over selected sites at the Los Alamos National Laboratory (LANL) using the SIBERIA landscape evolution model [Willgoose and Riley, 1998]. SIBERIA repeatedly applies a landscape-forming annual runoff event to a gridded model landscape, deforming the model topography in response to erosion and deposition. Where conventional erosion models predict sediment production from a site for a particular event, SIBERIA and other landscape evolution models incorporate natural feedback mechanisms as drainage networks and depositional features form over the long term.
SIBERIA was applied under the surface erosion component of the Material Disposal Area G Performance Assessment/Composite Analysis (PA/CA) of 2008, initiated in 2003 [French et al., 2008; Wilson et. al, 2005]. Material Disposal Area G (MDA G) is a site for subsurface disposal of low-level radioactive waste at LANL. Prior to joining Portage, Kelly Crowell, Ph.D., implemented the SIBERIA model for MDA G with LANL and United States Department of Agriculture-Agricultural Research Service personnel, modeling erosion over periods of 1000 years and longer. FY 2010 activities with Portage involved continuing previous work toward assessing improvements to the SIBERIA code, presenting the erosion modeling work to the Northern New Mexico Citizens' Advisory Board, and initiating an uncertainty analysis to assess the model’s sensitivity to changes in the primary SIBERIA parameters.
Material Disposal Area T (MDA T) is a site which received different types of waste from 1945-1986. Four absorption beds received treated liquid radioactive waste,
64 buried shafts received cement-treated mixed waste, and a removable waste storage area served as temporary storage for cement-treated radioactive waste (now removed). Two liquid waste treatment plants and associated subsurface infrastructure are or were associated with the site. A risk assessment for MDA T modeled after a Performance Assessment/Composite Analysis was initiated in late CY 2009. Portage performed the surface erosion component by applying the SIBERIA model to model 1000 years of erosion.
SIBERIA is a Fortran code that solves differential equations describing rates of change of elevation and potential for channelization. The primary inputs involve surface topography as a gridded digital elevation model (DEM) and a set of parameters relating to advective and diffusive sediment transport processes, material properties, and various mechanisms controlling the model execution. The sediment transport parameters are ideally developed from rainfall-runoff data taken at a site, but the model may also be calibrated against other models. Until data for the relevant LANL sites and proposed cover materials become available, the Hillslope Erosion Model [Lane et al., 2001] is used to generate reference sediment flux datasets for the PA/CA simulations.
Additional inputs may be provided to describe geometry and altered sediment
transport parameters through two- and three-dimensional regions, allowing the
modeling of, for example, a mobile cover surrounded by riprap or a mobile cover and soil mantle atop a bedrock mesa.
Model outputs are text files written at user-selected times, generally 20-year intervals for the LANL models. State variables describing the model landscape are written to restart files and, if requested, additional derivative outputs such as accumulated elevation change at a grid cell or in the cells draining through it.
Internal model states are written to a separate text file. When the three-dimensional layers capability is enabled, a file describing sediment layers at each grid cell is written.
Applying SIBERIA at a site requires at minimum creating properly formatted text files with the surface topography and model input parameters. In the full three-dimensional modeling case, text files are created to define two-dimensional regions and additional DEMs describing surfaces such as a bedrock layer, the upper surface of a body of waste, or some other surface at which a change in sediment transport parameters occurs or a tracer is desired.
Routines have been written in IDL (ITT Visual Information Solutions, Boulder, CO) to write SIBERIA input files based on geospatial datasets created in a GIS and to read output files for further analysis in the IDL environment or a GIS. An effort to reorganize the routines and port them to Python is in the embryonic stage. Maps of net erosion and deposition as well as depth to waste for a given modeling scenario at selected simulation times are typical products created in the GIS. IDL routines for working with the layers output file create geospatial datasets for depicting transport of tracers, which might represent cover material removed from over a waste disposal unit or contaminated material mobilized after excavation by erosion. Other derivative datasets provided to the risk assessor include time series of cover removal for use in biotic intrusion modeling and tracking potentially contaminated sediment from different waste disposal units of interest to particular reaches along neighboring streams.
Simulations run for the initial PA/CA were set up with two-dimensional regions which partitioned the closure cap from the mesa cliffs. The capability to track different sediment types and tracers in three dimensions was added in 2006. The Portage work with SIBERIA began with analyzing simulations run with the three-dimensional "layers" capability, comparing them to the original simulations, and documenting the differences [Crowell, 2010, submitted for LANL review]. Improved handling of sediment transport across boundaries between mobile material and riprap or bedrock generally increased the predicted sediment yield across areas with steeper slopes, while predicted yield decreased in low-slope regions. Predicted amounts of cover removed from over waste disposal units increased slightly on average due to increased gullying.
An analysis is underway to assess sensitivity of the MDA G application to variation in the primary advective transport parameters. Hillslope Erosion Model fluxes were generated based on slope, canopy cover, and ground cover data taken on a series of profiles in Area G and the equivalent SIBERIA parameters discovered through a downhill simplex optimization technique. Erosion was simulated for 1000 years for each set of parameters. Analysis of sensitivity of total sediment yield and rates of cover removal over waste disposal units is underway.
The MDA T Risk Assessment of FY 2010 was modeled after a PA/CA exercise and was intended to establish a baseline scenario for future remediation decisions. Assumptions underlying the MDA G PA/CA modeling were used in applying SIBERIA to MDA T, with unmodified current surface topography (minus structures) used to establish the model landscape. A set of erosion scenarios was run for 1000 years to produce erosion and cover removal maps and time series of cover removal.
Crowell, K.J., 2010, Updated Surface Erosion Modeling for Repository Waste Cover at Los Alamos National Laboratory Technical Area 54, Area G, submitted for Los Alamos National Laboratory review.
French, S., R.Shuman, G. Cole, C.J. Wilson, K.J.Crowell, M.S. Day, C.W. Gable, M.O. Gard, J.J. Whicker, D.G. Levitt, B.D. Newman, B.A. Robinson, E. Springer, and P.H. Stauffer, Performance Assessment and Composite Analysis for Los Alamos National Laboratory Technical Area 54, Area G, Revision 4, Los Alamos National Laboratory, Report LA-UR-08-06764.
Lane, L.J., M.H. Nichols, L.R. Levick, and M.R. Kidwell, 2001, A Simulation Model for Erosion and Sediment Yield at the Hillslope Scale, Chapter 8 (pp. 201-237) in Landscape Erosion and Evolution Modeling (Harmon, R.S. and Doe, W.W., III, eds.), Kluwer Academic/Plenum Publishers, New York.
Willgoose, G.R. and S. Riley, 1998, The Long Term Stability of Engineered Landforms of the Ranger Uranium Mine, Northern Territory, Australia:
Application of a Catchment Evolution Model, Earth Surface Processes and Landforms, Vol. 23, No. 3, pp. 237-259.
Wilson, C.J., K.J. Crowell, and L.J. Lane, 2005, Surface Erosion Modeling for the Repository Waste Cover at Los Alamos National Laboratory Technical Area 54, Material Disposal Area G, Los Alamos National Laboratory, Report LA-UR-05-7771.