JA:Import/Catalogue/JP MAFF farmland import

From OpenStreetMap Wiki
Jump to navigation Jump to search

Japan_MAFF_Farmland_import is an import of Fude Polygon from MAFF (Ministry of Agriculture, Forestry and Fisheries. Japan) dataset which is of type farmland/landuse covering (broad location in Japan).

The import is currently (as of 2019/June) at the planning stage.

Goals

Import the Fude Polygon from MAFF.

Schedule

TBD

Import Data

Will be published in July 2019 from MAFF.

Spec of the data

This dataset is made by MAFF.

Traced by hands of MAFF staff using 50cm resolution imagery. It takes 5 years of making for whole Japan, each year they can renew 1/5 of that.

Iwasaki-san creates geojson data from MAFF sample, so you can check geometric issues with that. Basemap is GSI imagery.

The polygons does not implies cadastral (law) boundaries.

Background

Data source site: http://www.maff.go.jp/j/tokei/porigon/index.html
Data license: http://www.maff.go.jp/j/tokei/porigon/attach/pdf/index-23.pdf
Type of license (if applicable): MAFF original. It is based on Japan government Open Data License, but less strict. It requests attribution similar to CC-BY. Link to permission (if required): user:nyampire got a special permission from MAFF officer to import their Fude data. By F2F conversation and mail.
OSM attribution (if required): http://wiki.openstreetmap.org/wiki/Contributors#MAFF
ODbL Compliance verified: yes

OSM Data Files

TBD

Import Type

This is one-time import.

But MAFF is planning to update their publication each year. So commutative update is possible option.

Method TBD (Data is generated by Municipality unit, Using Tasking Manager to handle, etc...)

Data Preparation

Data Reduction & Simplification

The principle regarding existing data should be discussed and defined in Japanese community.

  • How to map farmland (Each agricultural block? or should be united if polygons are so closer)
  • How to handle existing farmland data.
  • How to handle existing roads or forest data which duplicate with MAFF dataset.

Tagging Plans

Each polygon has following column.

- 耕地の種類 (feature type): 田 (rice field) or 畑 (other crops)

- ID: Unique ID for each polygons

- check_date: original data has a timestamp information on file name. All the data has "2019-5" (generated in 2019 May). This attribution could be used on future update.

Original Field Possible Value OSM Taggingfor the Polygon
耕地の種類 (Type of farmland) 田 (Rice field) landuse = farmland

crop = rice

畑 (Other crop) landuse = farmland
ID "Unique ID for polygons" "Unique ID for polygons"

(keep original value)

- check_date check_date = 2019-5

Changeset Tags

Following tags will be added in changeset tag.

Data Transformation

Original dataset is shapefile with EPSG: Japan Plane Rectangular.

The shapefile has attribute table with 2byte header (shift-jis encoding). So translation step could be...

  • convert original shapefile (EPSG:245N, Shift-jis) to geojson (EPSG:3857, UTF-8) using ogr2ogr.
  • replace 2 byte characters on attribute table header using sed.
  • convert the geojson to shapefile (EPSG:3857, UTF-8) once again.
  • translate the shapefile into .osm file using ogr2osm.

And should be divided by some units. (500m Mesh or district boundaries?)

Data Transformation Results

TBD

Data Merge Workflow

Team Approach

Import Team is organized in OSMFJ Github project.

This project is hosted in OSMFJ (OpenStreetMap Foundation Japan) repository. But it is not a part of the activity of OSMFJ itself. OSMFJ provides this repository space to organize

References

TBD. List all factors that will be evaluated in the import.

Workflow

  • Step by Step instruction: details TBD
  1. Set prioritized area
  2. Data Preparation (Fude data conversion, put them into Github repository)
  3. Data Preparation (existing data )
  4. Data import work

Changeset size policy: If each changeset has 15000 or more polygons, it should be divided less.

Revert Plan: Team has by week meeting and will report situation to community. If something bad is reported, the changeset will be reverted by @nyampire

Tasks are organized in Github project.

Conflation

WIll be discussed in Mailing list first.

QA

We will reduce errors using JOSM validation feature.

See also

TBD

The email to the Imports mailing list was sent on YYYY-MM-DD and can be found in the archives of the mailing list at [1].