Verkehrswende-Meetup/Parkraum/Mapping Kampagne Xhain/QA-public-transport-stops-GTFS

From OpenStreetMap Wiki
Jump to navigation Jump to search

Mapping-Projekt

>>> https://maproulette.org/browse/challenges/#TODO <<<

Ziel

Für die Parkraumauswertung sind präzise Daten zu Bushaltestellen entlang der Straße wichtig. Mit diesem Datensatz der Qualitätsanalyse (QA) wollen wir sicherstellen, dass die Daten zu Bushaltestellen in OSM präzise und aktuell sind.

Mehr über die Stichprobenanalyse unter https://parkraum.osm-verkehrswende.org/posts/2022-09-09-stichprobenpruefung

Vorgehen

Mappen:

  1. TODO

Datenaufbereitung

Stops Data:

  1. Download the GTFS OpenData from https://daten.berlin.de/datensaetze/vbb-fahrplandaten-gtfs (My file was from 29.09.2022.)
  2. Pick the stops.txt, use https://placemark.io/ to create a GeoJSON file from it
  3. Download the GeoJSON from Placemark

Bezirksgrenzen Data:

Download "Bezirke" from https://daten.odis-berlin.de/de/dataset/bezirksgrenzen/, https://tsb-opendata.s3.eu-central-1.amazonaws.com/bezirksgrenzen/bezirksgrenzen.geojson

QGIS OSM Data:

Use QuickOSM to load OSM data

  • public_transport=platform
  • public_transport=stop_position

We need both, since the GTFS position does not match either tags in OSM, so we look at a buffered version of both.

QGIS filter and extract Data:

  1. Both files: Export/import the data as EPSG:3857 to allow buffers in meter
  2. Both files: Create a buffer of 5 meter – since precision is one goal for the parking analysis, this buffer is quite small
  3. Merge the buffered files to become one layer
  4. Dissolve this merged layer
  5. Split dissolved layer into separate areas (one merged group of buffers is one feature) using "Multipart to singlepart"
  6. Use "Extract by location" to only keep the GTFS points that are _not_ (disjoint) inside the area of one of our buffers
  7. Use "Join attributes by location" to transfer the "Gemeinde_name" from "Bezirksgrenzen" to our filtered Dataset
  8. Filter our dataset some more with "Extract by attribute" and use "Gemeinde_name" "Friedrichshain-Kreuzberg" and "Neukölln"
  9. Export those two files as GeoJSON with EPSG:4326

Those are files we can use in MapRoulette to check each location.