We are using various krigging methods available in ArcGIS to map engineering soil parameters in mountainous region. Usually points with uniform distances picked by making grids are used in plane regions but we were wondering if the same principles could be adopted for mountainous regions as well. And how is the variability patter of index soil properties different in case of Mountainous regions and plane regions.
Hi Manoj,
I'm not sure what you mean about uniform points; if your mountainous region is represented in a DEM, you've already got regularly-spaced samples with the pixels. If you need those as vector points, just use the Raster to Points tool to get the centroids. Another way to get uniform points is to use the Create Fishnet tool, and take the 'label points', which will be the rectangle (or square) centroids.
Just a word of caution - to be sure the points are actually regularly spaced, make sure you're using an appropriate map projection that adequately preserves relative distances (and areas). This matters especially the smaller scale (i.e., more zoomed out) you are.
Cheers :)
Hi. If you are using ArcGIS 9.3 you can use Hawth's Tools for generating either regular or random grid points. Than you can use them to extract your date (soil parameters and predictors). Theoretically, the grid of points should have higher densities in areas with more complex topography, in order to capture the higher spatial variability of terrain parameters. You could test the influence of different sampling strategies by comparing the quality parameters of interpolations (eg. RMSE for cross-validation or for independent samples) and chose the optimum pattern.
- Badges
- Science topic
- Similar topics
- Agricultural Science
- Soil Science
- Soil Analysis