SARNr: Efficiently extract actively-flowing drainage networks from RS imagery

SARNr is an R implementation of a modified version of the SARN algorithm. The original SARN algorithm is an Arcpy script developed by Xin Lu and Kang Yang at Nanjing University to map actively flowing drainage networks by combining Sentinel-2 water masks and a DEM-based river network. See the below paper for more.

Lu, X., Yang, K., Lu, Y., Gleason, C.J., Smith, L.C., & Li, M. (2020). Small Arctic rivers mapped from Sentinel-2 satellite imagery and ArcticDEM. Journal of Hydrology, 584, 124689

SARNr takes this algorithm and rewrites it in R such that you 1) don't need an ArcGIS license, 2) you can use any satellite imagery of any resolution, and 3) processing can easily be farmed out to parallel workers on an HPC and dramatically increase processing speeds.

We also make some minor changes to the algorithm itself, namely 1) some pre-processing tasks, 2) water classification method, and 3) automated procedures for topological cleaning and network connectivity. This 3rd one is the most important, as it facilitates river routing directly on the SARNr output and opens the door for really interesting research! However, the core theory behind the SARN algorithm (trimming the network to the RS rivers) is maintained.

To install

Note that one of the package dependencies (sfnetworks) requires R 3.6.0.

# First get devtools package
if (!require("devtools")) {
  install.packages("devtools")
  library("devtools")
}

#install SARNr
Sys.setenv(TAR = "/bin/tar") #if using a Linux machine (apparently)...
devtools::install_github("craigbrinkerhoff/SARNr", ref='main', force=TRUE)

Note that the terra package auto loads sp and raster libraries into the namespace. Depending on your setup, you might need to manually specify these.

To use

Check the Introduction vignette at vignette('introduction', package="SARNr').