Florida Landuse Import

Background

Data source site: https://geodata.dep.state.fl.us/datasets/FDEP::current-landuse-landscape-support-index-lsi/about
Data license: Public domain
ODbL Compliance verified: yes

The dataset was created in 2017, the data is thus pretty up-to-date. Some areas are even newer, with up to 2019.

Import Type

• I import the data from the processed .osm file on a countywide scale. For the objects in the table below with extra notes for review, I look at these with extra scrutiny. SherbetS (talk) 02:10, 30 September 2022 (UTC)

Tagging Plans

Non-imported objects

Public Comment and Testing Periods

Changeset got a lot of flak when Hiausirg initially posted to mailing list and OSMUS slack around February 2022, Approval was received on the mailing list August 2022.

Changeset Tags

• Comment: (County Name) County Florida Landuse Import
• Source: Florida Department of Environmental Protection
• Website: https://wiki.openstreetmap.org/wiki/Florida_Landuse_Import

Data Transformation

1. Download dataset.
2. Load shapefile into QGIS.
3. Split large shapefile into a smaller shapefile based on county boundaries using extracted OSM data. For coastal counties, modify your county boundary extract to use the coastline as the border.
4. Use the buffer tool to create a 0.00003 degree wide buffer around the boundary, and use the difference tool to crop the boundary buffer out of the landuse file.
5. Extract all highway=* in the target county.
6. Fix any geometry issues. Review all unnamed residential roads. Delete any invalid geometry. Upload fixes to OSM server.
7. Purge any tracks, footpaths, paths, service roads, cycleways, etc. that you won’t want to cut out. Go through and purge any unpaved roads, and roads that are too minor to create a buffer around.
8. Save as .geoJSON file and open in QGIS. create a buffer around the roads that is 0.000035 degrees in width. set the buffer settings to use square ends and bevel corners.
9. Use the difference tool to remove the road buffers from the landuse file.
10. Download all landuse=* natural=* tourism=* leisure=park leisure=golf_course amenity=* power=substation power=plant zoning areas from OSM into JOSM.
11. Filter out buildings, places of worship, bridges, gas stations, parking lots, etc. that you won’t cut out of the final landuse file.
12. Delete low quality polygons and polygons that don’t provide any useful metadata.
13. Upload the deleted objects, save the file, and load into QGIS
14. Create a buffer around your landuse extract that is 0.00003 degrees in width and run the difference tool to remove the overlapping landuse.
15. Load smaller county landuse shapefile into JOSM
16. Delete any objects with extremely small area size created as a result of cropping. (JOSM search function can do “areasize: -700” or similar)
17. Run the JOSM validator to iron out any overlaps and other errors.
18. Review the entire file by eye. Delete any landuse areas that weren’t translated well (long areas near roads, small water areas next to cropped water areas, etc.)
19. Correct the tags cut off by the shapefile. (residentia=* >> residential=* intermitte=* >> intermittent=*)
20. Review all cemeteries and reservoirs. delete any that you cropped during conflation.
21. Upload data to OSM server. Use import account + the changeset tags above.

Data Transformation Results

This is an old file and should not be seen as reflective of the quality of the rest of the import.

https://drive.google.com/file/d/1kw76oj9sWAqe6HXFow5yuo4btkdFWZvO/view?usp=sharing

Jefferson county example file:

https://cloud.disroot.org/s/jH5rEj89cFoTQgt

Taylor, Lafayette, Madison, Dixie, Jefferson, Wakulla, Liberty, Leon counties are uploaded now.