image/svg+xml (a) (b) (c) Generates results & figures for a case study storm and for the full loss exceedance curve. Generates simulated realizations of damage ratio and expected loss for each building in the AOI. (a) Defines terrain & location information for the hydrodynamic software LISFLOOD;(b) calibrates environmental parameters to produce a best-fit model for Sonoma County; and (c) fits a low-dimensional surrogate model to reduce computational demand. Generates simulatedrealizations of precipitation, antecedentsoil moisture, peakstreamflow, and time to peak streamflow. Defines an area of interest (AOI) andcreates a catalog of historic ARs that haveoccurred within that area. 1. collect site-specific information 2. fit hazard distributions 3. fit inundation distribution 4. fit damage & loss distributions 5. run PARRA framework generate_soilmoisture: Fits a lognormal distribution to recorded soil moisture values in the historic catalog generated by catalog.Rmd and simulates new values based on this distribution. generate_runoff: Calculates runoff as a function of precipitation agenerate_soilmoisture: Fits a lognormal distribution to recorded soil moisture values in the historic catalog generated by catalog.Rmd and simulates new values based on this distribution. generate_runoff: Calculates runoff as a function of precipitation and soil moisture using the curve number method. generate_hydrograph: Fits a linear regression to peak streamflow, Qp, as a function of precipitation, runoff, and soil moisture. Fits a lognormal distribution to time to peak streamflow, tp, based on values in the historic catalog. Simulates new values of Qp and tp. RNFF.R Finds best-fit coefficients for a quantile regression and simulates new precipitation values as a function of AR characteristics.PRCP.R Generates figures illustrating the area of the case study at the Russian River, Sonoma, California. sonoma.Rmd Creates a catalog of historic AR events based on grid_catalog.Rmd. Calculates hazard variables and other relevant information for each AR event.catalog.Rmd Downloads IVT, duration, & AR identification information from Rutz et al. (2014)grid_catalog.Rmd Formats the digital elevation model (DEM) for use in LISFLOOD. Creates rasters for river width and floodplain roughness. Defines coordinates of the study area inlet and outlet.lisflood.Rmd Creates a raster file of the FEMA NFHL to serve as the "true" value for comparison.NFHL.Rmd Generates Latin hypercube samples of the environmental parametesr of interest and produces LISFLOOD input files for each sample realization.generate_files.sh Produces a LISFLOOD inundation map for each of the sample realizations.run_lisflood.sh Performs parameter sensitivity testing, determines best-fit environmental parameters, and calculates accuracy metrics for the "true" (NFHL) vs. simulated inundation maps.rp100.Rmd Generates Latin hypercube samples for the hydrograph parameters of interest and produces LISFLOOD input files for each sample realization.generate_files.sh Produces a LISFLOOD inundation map for each of the sample realizations.run_lisflood.sh Calculates best-fit values for surrogate model hyperparameters.fit_npalpha.sh Estimates surrogate model error compared to LISFLOOD model.surrogatemodel.Rmd Calculates a conservative buffer around the maximum inundated area to speed up Monte Carlo simulations.nonzero.Rmd Identifies single-family residential buildings within the bounds of the area defined by nonzero.Rmd and creates a dataset of building locations. Attaches RESA damage information and valuation information from the Sonoma tax assessor roll to each building.buildings.Rmd Estimates the distribution of foundation types by census tract.foundations.Rmd Collates depth-damage curves from multiple sources and formats them for consistent application.depthdamage.Rmd Randomly assigns foundation heights to buildings and determines water depths. Converts these depths to expected damage ratios for each building.DM.R Uses the valuation information from buildings.Rmd and the damage information from DM.R to estimate repair cost for each building.DV.R Steps through a component-by-component case study, using validation data from a severe AR event affecting Sonoma County in 2019. componentmodels.Rmd Runs the PARRA framework from ARs to impacts for a given number of Monte Carlo realizations.PARRA.sbatch Creates a loss exceedance curve for AR-induced fluvial flood losses to residential buildings in Sonoma County. Creates a mitigated loss exceedance curve after raising all residential buildings about the 100-year floodplain and compares estimated losses. lossexceedance.Rmd legend: Defines the AOI as a rectangular sf (spatial) object in R.aoi.Rmd output files (start here) data creation scripts Sherlock (Linux) bash/batch scripts component modelimplementations Implementation WorkflowPerformance-based Atmospheric River Analysis (PARRA) Framework Created for the paper "A Performance-Based Approach to Quantifying Atmospheric River Flood Risk" (doi:XX)Last edited by Corinne Bowers, 8/9/2021 This graphic is not built for mobile and will not display correctly. Descriptive text is available if you hover your mouse and pause over each filename.