Greg Cocks Avoid Backtracking And Burn Your Inputs - CONUS-Scale Watershed Delineation Using OpenMP
--
https://doi.org/10.1016/j.envsoft.2024.106244 <-- shared paper
--
“HIGHLIGHTS
• A memory-efficient watershed delineation algorithm was introduced.
• The new algorithm uses a node-skipping depth-first search to save memory.
• Both input and output data are stored in a shared matrix to reduce required memory.
• It performed 95% faster than its CPU benchmark algorithm using 33% less memory.
• It can solve 50% larger problems than what the CPU benchmark algorithm can handle..."
#GIS #spatial #mapping #hydrology #water #code #algorithm #watershed #delineation #OpenMP #opensource #memory #efficient #CPU #benchmark #MESHED #model #modeling #largescale #catchments
![figure - Illustration of self-destructiveness. Numbers indicate the discovery order of each cell from [paper's] Fig. 1(c).
W is the watershed ID for the outlet. See [paper's] Fig. 1(a) for cell IDs.](https://files.techhub.social/media_attachments/files/113/386/211/345/872/578/original/b54310bb8a3d084f.jpg "figure - Illustration of self-destructiveness. Numbers indicate the discovery order of each cell from [paper's] Fig. 1(c).
W is the watershed ID for the outlet. See [paper's] Fig. 1(a) for cell IDs.")
