rwPerformance: Random walk runout path performance

In jngtz/runout.opt: GPP regional runout model optimization and validation in R

Description

Computes the area under the receiver operating characteristic curve (AUROC) for runout paths simuluated using the random walk model component of the GPP tool in SAGA-GIS. The AUROC compares a runout polygon to the simulated path.

Usage

rwPerformance(
  dem,
  slide_plys,
  slide_src,
  slide_id = 2,
  slp = 33,
  ex = 3,
  per = 2,
  gpp_iter = 1000,
  buffer_ext = 500,
  buffer_source = NULL,
  plot_eval = FALSE,
  saga_lib
)

Arguments

dem	A DEM as a RasterLayer object
slide_plys	Runout tracks as a SpatialPolygonsDataFrame
slide_src	Source points as a SpatialPointsDataFrame or source areas as a SpatialPolygonsDataFrame
slide_id	Selects a single runout polygon from slide_plys by row to run GPP model
slp	Random walk slope threshold - below lasteral spreading is modelled
ex	Random walk exponent controlling lateral spread
per	Random walk persistence factor to weight flow direction consistency
gpp_iter	Number of model iterations
buffer_ext	(Optional) Defines buffer distance (in meters) around runout polygon to crop source DEM. This helps to reduce computational time
buffer_source	(Optional) Can define a buffer distance (in meters) to extend source point to a source area
plot_eval	logical. If TRUE will plot random walk path and runout polygon
saga_lib	The initiated SAGA-GIS geoprocessor object

Details

Runout source can be either point or area.

Value

The area under the receiver operating characteristic

Examples

## Not run: 
# Initialize a saga object
saga <- Rsagacmd::saga_gis()

# Load elevation model (DEM)
dem <- raster::raster(system.file("extdata/elev_12_5m.tif", package="runout.opt"))

# Load runout polygons and source points
runout_plys <- rgdal::readOGR(system.file("extdata/dflow_runout_ply.shp", package="runout.opt"))
source_pnts <- rgdal::readOGR(system.file("extdata/dflow_source_pnt.shp", package="runout.opt"))

# Run GPP random walk model for a rounout polygon
rw <- rwPerformance(dem, slide_plys = runout_plys[1,], slide_src = source_pnts,
    slp = 30, ex = 3, per = 2,
    gpp_iter = 1000, buffer_ext = 500, buffer_source = 50,
    plot_eval = TRUE, saga_lib = saga)

rw # returns AUROC


## End(Not run)

jngtz/runout.opt documentation built on Sept. 8, 2021, 2:15 p.m.