2023 Brazil Floods

From OpenStreetMap Wiki
Jump to navigation Jump to search

General Information

Context of the Activation

The extratropical cyclone that hit the state of Rio Grande do Sul (RS) in the early hours of Monday (4/9) [1] caused flooding in several cities in the state. Several cities are in a state of alert. The affected areas are not completely mapped in OpenStreetMap. In order to complete the map and make this open geographic data available to organisations, groups or individuals in need or interested in it, we are activating the mapping for some of the affected municipalities. This project involves volunteers from the OpenStreetMap community, especially the Brazilian community, and is supported by UMBRAOSM, the Open Mapping Hub of Latin America and the Caribbean, the Brazilian chapters of YouthMappers.

Hashtags

#enchentesRS #OpenStreetMapBR #UMBRAOSM #HOThubLAC #YouthMappers

About This Disaster Activation

About HOT: To learn more about the Humanitarian OpenStreetMap Team (HOT), explore more of our wiki-pages (root: HOT) or our website hotosm.org. HOT is a global community, mostly of volunteers, and it is a US registered nonprofit able to contract with organizations (email info at hotosm.org to contact our staff).

About YouthMappers:

About UMBRAOSM:

History of this Activation

13 Set 2023 - After the event, the Open Mapping Hub Latin America and the Caribbean, the YouthMappers chapters in Brazil, and UMBRAOSM agreed to launch an activation to support the response efforts after discussions with OpenStreetMap Brazil members.

Further coordination on mapping priorities, their capacity to coordinate and engage, and the technical aspects of Tasking Manager projects took place via UMBRAOSM, YouthMappers Brasil, OpenStreetMap Brasil and OpenStreetMap Rio Grande do Sul telegram and whatsapp channels, and the HOT Slack workspace channel #disaster-mapping, bridged by Matrix.

The Secretaria Nacional de Defesa Civil (National Civil Defense Secretariat) of Brazil was consulted about the priority areas. They responded with the municipalities of Muçum, Encantado, Roca Sales, Arroio do Meio and Estrela.

The Mapeadores Livres UFPR (YouthMappers Chapters) evaluate the SHP files of the priority areas.

15 Oct 2023 - Four out of five mapping projects have been completed.

15 Nov 2023 - OpenStreetMap Rio Grande do Sul state members asked for clean-up of mapping.

15 Dec 2023 - Brazilian OpenStreetMap community was asked for data quality evaluation.

Coordination

Everton Bortolini (HOT / Hub de Mapeamento Aberto para a América Latina e Caribe) - everton.bortolini@hotosm.org, Nathan Damas (YouthMappers), Raphael Assis (UMBRAOSM).

For Mappers

How You Can Contribute

Learn to Map

  • Most of our volunteer needs are for remote OSM contributors, visit LearnOSM.org to get started.

Mapping Priority

  • Please choose from highest priority first
  • Experienced mappers are also asked to participate in validating completed tasks. Information on validating can be found here
Project Priority Name Organization Area What to map Imagery Source Task Mapping Status % Task Validation Status % Project Clean up Status
15493 Urgent BRASIL - ENCHENTE NA CIDADE DE MUÇUM - (RS) UMBRAOSM Municipality of Muçum - Urban Area Buildings Bing Imagery 100% Complete 100% Complete Done
15557 Urgent Brasil - Enchentes na cidade de Encantado (RS) Open Mapping Hub for Latin America and Caribbean Municipality of Encantado - Urban Area Buildings Bing Imagery 100% Complete 100% Complete Done
15523 Urgent Brasil - Enchentes na cidade de Roca Sales (RS) Open Mapping Hub for Latin America and Caribbean Municipality of Roca Sales - Urban Area Buildings Bing Imagery 100% Complete 100% Complete Done
15602 Urgent Brasil - Enchentes na cidade de Arroio do Meio (RS) Open Mapping Hub for Latin America and Caribbean Municipality of Arroio do Meio - Urban Area Buildings Bing Imagery 100% Complete 100% Complete Done
- - - - Municipality of Estrela - Urban Area - - - - -

For Validators

Data Quality

OSMCha

A specific OSMCha filter can be used to detect data quality issues across a wider area.

Overpass turbo

A specific Overpass consult can be used to detect amenity=* modified.

Before and After Map

Before (2022) and After (Now) of town of Muçum.

Report data quality issues

Report chancesets with data quality issues:

Description Changeset (link) Action Status
Breaking administrative division. https://www.openstreetmap.org/changeset/142053572 Revert changeset Done
Deletion points of interest without replacement (eg: telephone c142196646 , city town hall of the municipality (Encantado Townhall) c141685445 https://www.openstreetmap.org/changeset/141685445 Recover all features victims of harmful Done
Data contaminated by incompatible licenses (Google Maps and Google Street View). (eg: building=yes, hospitals c141785695 and veterinary and shops c141793590 ). https://www.openstreetmap.org/changeset/141593785https://www.openstreetmap.org/changeset/141785695https://www.openstreetmap.org/changeset/141793590 Remove, revert or verify the source of features Done
Data with none or with misleading sources (Eg: Parks [c141913953] and named tourism attractions [c141890288] and amenity=fuel c142054380 ). https://www.openstreetmap.org/changeset/141913953https://www.openstreetmap.org/changeset/141890288https://www.openstreetmap.org/changeset/142054380 Revert features Done
Feature created by the local mapper, deleted by participants and created as a new element https://www.openstreetmap.org/changeset/142734654 (which POIs were deleted? note that deleting node and drawing accurate area would be fine) Recover POI added by mappers which are victims of delete-then-recreate Need confirm this action
Any significant change over the recovered data (other than imagery alignment) must be explained. https://www.openstreetmap.org/changeset/142196646https://www.openstreetmap.org/changeset/142196646 Recover all buildings previously added by mappers (local knowledge/survey) that were victims of delete-then-recreate by armchair mapping Need confirm this action

Report data quality issues by project:

Name Data quality issue Observations
BRASIL - ENCHENTE NA CIDADE DE MUÇUM - (RS)
Brasil - Enchentes na cidade de Encantado (RS)
  • Unsquared buildings / non-circular buildings and overlapping buildings using Mapathoner Plugin: corrected these one by one. There are still buildings that have not been mapped yet, but at least the ones already mapped are OK now.
  • Classification of highways for consistency (found some wrongly assigned highway=track, service, unclassified)
  • Mapped waterways (here I also spotted odd streams mapped by Maria Paula Simonette).
Brasil - Enchentes na cidade de Roca Sales (RS)
Brasil - Enchentes na cidade de Arroio do Meio (RS)
  • Mapping was not too bad. Only in some rare cases, the building contours did not match in the size/shape;
  • In some cases highway classification was incorrect, e.g. usage of living_street instead of residential or highways that were marked as unclassified were driveways;
  • No major issues at all: overlappings/nodes without properties (maybe someone cleaned this up already);
  • Tried to correct the positions of buildings on the highways using Strava HeatMap. Relatively difficult to create a consistent representation of the area. Applied different offsets to the imagery;
  • Corrected the tagging of building=house -> now building=yes in cases where these objects were created in the project and very likely by remote contributors;
  • Added "fixme=verifica que esse objeto é um edifício" to a few building I had doubts about their existence;
  • Primary highway ERS-130 (https://www.openstreetmap.org/way/176547223) was modified several times during this project which is logical as new highways connecting to it were drawn, but it might be that someone repositioned sections so that does not match perfectly to STRAVA / gpx traces (did not correct this issue);
  • One mapper did not stay within the borders of the task but had mapped also outside the project area - mostly non-squared buildings;
  • Widespread usage of buiding=house instead of building=yes from non-local mappers;
  • Mapping of land uses (wood/farmland). Nevertheless, it was not the topic of this project. The result of odd shapes that represent the shape of the task, but do not have any representation with the reality. Fixed these by joining areas and drawing the polygons in their shape using aerial imagery;
  • Wrong usage of highway=unclassified and highway=track for access highways to private properties - should have been tagged as highway=service;
  • The mapping quality differed a lot throughout the tasks. Some had a high ratio of non-squared buildings.

Results Report

Community result evaluation

The text was write on "community objections" by user EmericusPietra here , but was copied to this section. It needs validation from the Brazilian community.

  • Provide a public copy of instructions directed for mappers and validators as they might expose poor suggested tagging practices that need to be changed in future. Evidence: do exist content not already public description on Tasking Manager and large enough not fit a changeset reply Confirmation by global validator [c143333801].
  • Have a public commitment written in a archivable medium to adhere to minimal standards recommended even in the Tasking Manager documentation for the mapper level in any future mapping for the same organizers, such as never create an activity which allow 0-edit beginners edit areas with previous content from local mappers. Evidence of ignoring internal feedback: previous complains from one HOT global validator in public on previous organized editing TM13594 archived page TM13594, section Questions and Comments, reinforced by another global validator on the ongoing project, which, in a public changeset, admitted complain in private to the organizers and peers from HOT about the poor quality of the mapped tasks [c143333801]
  • Create Organized Editing Activity for each city (the first one, Muçum, was even different organizers and far less problematic result) or, if want to group the cities , the logical approach is use Vale do Taquari region and list the differences. The current title of the page "2023 Brazil Floods" makes no sense and even the hashtag "#EnchentesRS" implies the Rio Grande do Sul province.
    • This also means that recent attempts since 2022-12-15 Telegram message telegram message archived to move the discussion from active mappers from Rio Grande do Sul  (and even here, more priority to the city-level mappers if they engage) to any outside group to outnumber decision will be ignored. Even the organizers/validators which initially said to be from the province, did armchair mapping, with one of them, admiring using Google Street view to validate (and even without this, the map still has something similar to Copyright Easter Eggs easy to check).
  • Make public, written and archivable (such as on OSM Wiki to reference in future) commitment of any new organized editing by the same organizers will follow the Organized Editing Guidelines, and, based on the lack of awareness of the OEG in the previous weeks, MUST mention:
    • Make a commitment to start an explicit document in the instructions for armchair mappers and validators to respect license requirements to added data to OpenStreetMap. Don't copy from Google Maps / Google Street View!
    • Explicitly acknowledge the "no less than two weeks before the activity is started"
      • If the call for activity ask for any type of "urgency" to reduce this time or ignore any other part of the OEG, then it must faithfully respond to questions asking to prove the relationship between the organizers and the targeted data users without "move discussion to private channels"
    • The organizers MUST take responsibility to actively watch and respond to changeset comments when your contributors fail to do; however, if every changeset has a direct link to the OSM Wiki page of the activity, then organizers could just watch the wiki
    • Acknowledge the "within two working days while the activity is ongoing" to respond to feedback to any future organized editing.
      • Context: in addition to changeset comments, even several weeks with errors in points of interest and messages with direct link to the exact problem, feedback from organizers ranged from I am a volunteer/I have more to do than OpenStreetMap to 'please stop complaining again while more data is added without fix the previous previous problems
Retrieved from "https://wiki.openstreetmap.org/w/index.php?title=2023_Brazil_Floods&oldid=2679019"