search

schwilklab / skyisland-climate

Climate data and code for Sky Island project
2 stars 2 forks source link

DEM-derived variables descriptions --- clarifications #33

Closed dschwilk closed 8 years ago

dschwilk commented 8 years ago

I'm a bit confused regarding a couple of variables, especially z_valley and zdist_valley (and probably the ridge versions as well once we solve the ridge definition).

Regarding zdist_valley: I think negative numbers mean point is above a valley. Positive values mean the point is below? OK, so that is possible because valleys end at the map edges, so there are low elevations for which the closest valley really is off the map, so really we'd want to define all of those as NA. I think we can replace positive values with NA since a point cannot really be below the nearest valley. Is this right?

Additionally, there are still a couple of gaps in the methods document: https://github.com/schwilklab/skyisland-climate/blob/master/methods/topo_grid_methods.md

dschwilk commented 8 years ago

@hpoulos Can you check these descriptions? I corrected some names (adefdb3), but still some descriptions don't seem to match method. For example, I cannot find explanation of the smoothing for ridges that you implemented.

hpoulos commented 8 years ago

There was and updated document in the push from yesterday with information about that smoothing step. The smooth raster operation is an intermediary step, so it is described as part of the flow_accum raster creation step. I saved it separately as a toolbox, .jpg, and python script called smooth raster because it is a useful "stand alone" tool that could be used for this or other operations. However, those steps are also embedded in the flow_accum toolbox and described in the updated document I sent in that step because that grid and the flow_dir grids then serve as the input raster for manyof the subsequent derivative layers.

dschwilk commented 8 years ago

Ok, I must have missed that. If https://github.com/schwilklab/skyisland-climate/tree/master/GIS is all good, I am closing this issue.