thermokarstdetection

The package thermokarstdetection provides a function to easily detect thermokarst features from a single digital terrain model (ground surface elevation raster). It identifies cells which are at lower elevation than the surrounding landscape by using a neighborhood function to calculate the average (either median or mean) elevation in one or more user-defined neighborhoods.

Installation

You can install the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("HRodenhizer/thermokarstdetection")

Example

This is a basic example which shows you how to detect thermokarst features using the default options:

library(thermokarstdetection)
library(raster)
#> Loading required package: sp

### Inspect some elevation data
tk_ponds
#> class      : RasterBrick 
#> dimensions : 100, 100, 10000, 2  (nrow, ncol, ncell, nlayers)
#> resolution : 1, 1  (x, y)
#> extent     : 389060, 389160, 7085440, 7085540  (xmin, xmax, ymin, ymax)
#> crs        : +proj=utm +zone=6 +datum=WGS84 +units=m +no_defs 
#> source     : memory
#> names      : NEON_DTM_2017, NEON_DTM_2018 
#> min values :        669.92,        670.06 
#> max values :        671.55,        671.72
plot(tk_ponds[[1]])

### Use tk_detect() to find thermokarst features
tk <- tk_detect(tk_ponds)

The main output of tk_detect() is a categorical raster of thermokarst class:

# inspect cropped thermokarst output
plot(tk$thermokarst[[1]][[1]])

However, it also returns intermediate data products. elev.crop is a version of the input elevation raster which has been cropped to the maximum possible output extent of the neighborhood function output:

plot(tk$elev.crop[[1]])

avg.elev is the average neighborhood elevation (either median or mean, as specified by the user):

plot(tk$avg.elev[[1]][[1]])

microtopography is a raster of the deviation in elevation from the average elevation, defined as elev.crop - avg.elev:

plot(tk$microtopography[[1]][[1]])

The output of tk_detect() reports the input parameters (radii, fun, cutoff) which were used, as well. radii can be either a single value or a vector of radii which determine(s) the neighborhood size when calculation the average elevation. fun is the function to use when calculating the neighborhood average elevation (either median or mean). cutoff is the value used to classify microtopography as thermokarst.

tk$radii
#> [1] 15
tk$fun
#> [1] "median"
tk$cutoff
#> [1] 0