Now that I know how to make a map overlay for OsmAnd, I wanted a land ownership overlay. When we're hiking, we often wonder whether we're on Forest Service, BLM, or NPS land, or private land, or Indian land. It's not easy to tell.
Finding Land Ownership Data
The first trick was finding the data. The New Mexico State Land Office has an interactive New Mexico Land Status map, but that's no help when walking around, and their downloadable GIS files only cover the lands administered by the state land office, which mostly doesn't include any areas where we hike. They do have some detailed PDF maps of New Mexico Lands if you have a printer capable of printing enormous pages, which most of us don't.
In theory I could download their 11" x 17" Land Status PDF, convert it to a raster file, and georeference it as I described in the earlier article; but since they obviously have the GIS data (used for the interactive map) I'd much rather download the data and save myself all that extra work.
Eventually I found New Mexico ownership data at UNM's RGIS page, which has an excellent collection of GIS data available for download. Click on Boundaries, then download Surface Land Ownership. It's available in a variety of formats; I chose the geojson format because I find it the most readable and the easiest to parse with Python, though ESRI shapefiles arguably might have been easier in QGIS.
Colorizing Polygons in QGIS
You can run qgis on a geojson file directly. When it loads it shows the boundaries, and you can use the Info tool to click on a polygon and see its metadata -- ownership might be BLM, DOE, FS, I, or whatever. But they're all the same color, so it's hard to get a sense of land ownership just clicking around.
To colorize the polygons differently, right-click on the layer name and choose Properties. For Style, choose Categorized. For Column, pick the attribute you want to use to choose colors: for this dataset, it's "own", for ownership.
Color ramp is initially set to random. Click Classify to generate an initial color ramp, then click Apply to see what it looks like on the map.
Then you can customize the colors by doubleclicking on specific color swatches. For instance, by unstated convention most maps show Forest Service land as green, BLM and Indian land as various shades of brown. Click Apply as you change colors, until you're happy with the result.
Exporting to GeoTIFF
You can export the colored layer to GeoTIFF using QGIS' confusing and poorly documented Print Composer. Create one with: Project > New Print Composer, which will open with a blank white canvas.
Zoom and pan in the QGIS window so the full extent of the image you want to export is visible. Then, in the Print Composer, Layout > Add Map. Click and drag in the blank canvas, going from one corner to the opposite corner, and some portion of the map should appear.
There doesn't seem to be any way to Print Composer to import your whole map automatically, or for you to control what portion of the map from the QGIS window will show up in the Print Composer when you drag. If you guess wrong and don't get all of your map, hit Delete, switch to the QGIS window and drag and/or zoom your map a little, then switch back to Print Composer and try adding it again.
You can also make adjustments by changing the Extents in the Item Properties tab, and clicking the Set to map canvas extent button in that tab will enlarge your extents to cover approximately what's currently showing in the QGIS window.
It's a fiddly process and there's not much control, but when you
decide it's close enough, Composer > Export as Image... and
choose TIFF format. (Print Composer offers both TIFF and TIF; I don't
know if there's a difference. I only tried TIFF with two effs.) That
should write a GeoTIFF format; to verify that, go to a terminal and
run gdalinfo on the saved TIFF file and make sure it says it's
Load into OsmAnd
Finally, load the image into OsmAnd's tiles folder as discussed in the previous article, then bring up the Configure map menu and enable the overlay.
I found that the black lines dividing the various pieces of land are
a bit thicker than I'd like. You can't get that super accurate "I'm
standing with one foot in USFS land and the other foot in BLM land"
feeling because of the thick black DMZ dividing them. But that's
probably just as well: I suspect the data doesn't have pinpoint
accuracy either. I'm sure there's a way to reduce the thickness of
the black line or eliminate it entirely, but for now, I'm happy with
what I have.
[ 18:13 Apr 15, 2019 More mapping | permalink to this entry | comments ]