[GRASS-user] Re-organizing Project

[Rich Shepard]

On Tue, 22 Dec 2009, Daniel Victoria wrote:

> I'm not understanding why you need / want different mapsets. I always see
> mapsets as a group of maps related to each other that I would like to
> separate in groups to keep things organized. For instance, I use different
> mapsets for different climate simulation scenarios. You could use
> different mapsets for different satellite images (years, sensors etc). But
> keep in mind that you might not need several mapsets and certainly, one
> mapset for each map is an overkill. Now, the PERMANENT mapset is a special
> one that can be accessed from within any other mapset. So that's the place
> to put maps like boundaries, counties, roads etc, that will be accessed by
> all.
>
> Another way of organizing is by creating different mapsets for different
> models and in PERMANENT you can put all common data to the models (DEM,
> streams, etc)
>
> This page helped me a lot in understanding what mapsets are for:
> http://grass.osgeo.org/grass64/manuals/html64_user/helptext.html

   Yes, the latter URL repeated the information in the book. But, there seems
to be many different preferences for organizing data. I'm open to all
opinions on this.

   I have source data for the following:
 	DEM 10m, northwest portion of the state (raster); projection: Oregon LCC
 	streams/rivers in the the state at 24K; projection: Oregon LCC
 	lakes/reservoirs in the state at 24K; projection: Oregon LCC
 	dams in the state at 24K; projection: Oregon LCC
 	hydrologic units (basins) in the state at 24K; projection: Oregon LCC
 	soils (map units, line and point features) at 24K in one county;
 		projection: Oregon LCC
 	roads/highways in the state at 24K; projection: Oregon LCC
 	state and county boundaries (not needed for this project, but to
 		establish largest region if necessary)
 	key point features specific to this project; projection: Oregon LCC

   Originally, each map was in a separate location. Now I want to put them
all in a single, statewide location. I don't care if they're all
imported/re-projected into the PERMANENT map set or individual ones.

   There's also the project-specific location which is a small subset of one
county. After getting all the source data correctly in their base
location/mapset(s), I want to define the region of this location or mapset
as slightly larger than the drainage basin I need to model. I plan to define
this by viewing the stream of interest with the HUC basins overlaid.
Running r.watershed on the DEM should match this area closely (I hope). This
will be the project's working directory. I was hoping to be ready to model
by now, but am high-centered on both organizing all source data and
moving/re-projecting it to a common place.

   There will be several (or more) projects in this and other states where I
will need to run GRASS models so defining an organizational structure and
being able to correctly move maps around is vital. I would like to establish
a standard now, use it for this project, and also use it for future
projects.

   Everyone's free to suggest organization. Interestingly, both the book (2nd
ed) and above URL focus more on mapsets per user, but show they can be used
per map theme.

Thanks,

Rich