How to define areas where to go mapping?

[kadry55]

One of the first questions when asking a company/club/organization to go mapping is where should we go? This could be an extra task for country coordinators/admins to define the areas manually. Or this could be something automated.

Example userstory: Invitation-> website with a map and areas(app.worldcleanupday.com)10 options areas (based on how far you are, team size, area not mapped, priority)-> area booked-> guideline -> start -> updates on map -> manual check if trashpoints were added-> Area mapped by "..."

Questions: Where to get the administrative areas? How to confirm that area actually was mapped? How to solve it if we won't get the areas from Google? How to prioritize areas? Should dangerous areas be left out? How to define them? Issue #75 How to leave out areas that doesn't need mapping? When should a user log in? If and where should the "Mapped by ... " be shown?

[kadry55]

@krishaamer @KristiinaKerge @kristiinapalu - this is what we are going to discuss also on Friday.

[lynxlynxlynx]

Here are my inputs:

1. Administrative areas can be extracted from OSM. There is also the GADM project, but OSM likely has better data. Our approach: municipalities are often too big, so we used the smaller cadastral municipalities and just grouped them where necessary.
2. We relied on trust. Also, if you have no reports in the area, it's already a yellow flag, unless the areas are very small.
3. Huh?
4. Proximity to settlements is a very good first predictor. In our case we didn't bother with it. Later data mining showed we could have magically focused only on 2% of the area and got at least 75% of the reports in.
5. Discussed elsewhere.
6. Mark them as completed.
7. Last step possible.
8. Just another data point when a (logged-in) user clicks on an area. If the original user opted-in.

[kadry55]

User story for 2 audiences:

1. Companies who we invite to have a team event as mapping waste and who would like to have suggestions on where to go.
2. Waste index input (areas will be defined differently based on the input from Knowledge Team)

Area priorities: Teams who are cleaning up nation wide (light green and dark green countries in: letsdoitworld.org) Knowledge team areas (for approach 2) - defined later Proximity to settlements - based on lynxlynxlynx comment Map not changeable - based on one map provider (China needs another map as Google doesn't work there)

Only for those who we send the link - every team will get a separate hashed link.

Who can use this marketing approach: LDIF mapping team LDIF partners team Country teams on local level

Marketing approach: direct

Month when needed: May Development should be done: February, March

User story:

1. An e-mail is sent to the company HR/events department which asks people to come mapping. It describes the event as one-day event, gives team participant limits (if more than 100 people then the same steps to be done more than once), describes how areas should be selected, mentions that area should be mapped in 3-4h, link to select an area, link of a map with trashpoints (World Waste Map).

2. When user click on "Select area for mapping event" then next landing page is a selection box: Enter location around which you'd like to see unmapped areas. Max distance from selected location (1...x km) Team size (5...100 persons) Start and end date (during 1 month timeframe) Enter

3. A map divided into areas opens up where suggested area is highlighted.

4. If there are more people in the team than the area needs for mapping then several areas (next to each other) are highlighted. Notification: considering the amount of people, please select at least ... areas. (We allow user to select less areas than recommended)

5. If there are no areas in this distance then comes a notification: Please change the distance from location.

6. User clicks on the selected area/areas on the map.

7. Another selection box pops up: Complete your booking: Company/Organization name ... Your e-mail: ...

8. Next user receives an e-mail: confirm you registration

9. User clicks on: confirm

10. User receives an automated e-mail which shows: what he/she selected, area booked, more information on how to do mapping: guideline, what to consider, admin contact(country level) in case he/she wants to cancel/change the booking.

11. Selected area turns gray - not selectable until the day after the mapping event.

12. Team event

13. On the day of the event there should be an automatic check regarding trash points. a) If there are new trashpoints created on the area during that day - area stays gray b) If there was no new trashpoints created in the area - area changes back to be selectable.

END OF STORY

Country admin: Can see on map view areas in his/her country in admin web.

Superadmin: Can see on map view all areas that are created in admin web.

Super/country admin: Can see if new trashpoints were created during the day of the event - how many? Clicking on area that is booked, opens a sidebar about the booking information. Buttons: release booking/ block e-mail link re-use

User e-mail is asked for:

1. Sending automatic email in step 6
2. To block spammers

Extra details: area map doesn't have to be refreshed online - map doesn't show trash points

[kadry55]

Is there a booking system already existing that could be used?

Is the cost of the development dependant on how many areas we use?

[kristiinapalu]

Time evaluation according to the user stories: Can be developed with any developer

1. 2 h
2. Landing page with the selection - 4h (development), probably will need design as well 3 - 6 could take 1-2 weeks // 40h-..
3. 4h 8 - 10. 3-4h 11 - 13. 8 - 16h

[KristiinaKerge]

can we use Google support to get selectable areas https://www.google.com/nonprofits/products/google-earth-outreach.html What should we ask from them?

[lynxlynxlynx]

I talked to Kadri a bit about this and more, so here are my thoughts:

• it's about the Google Earth Engine: https://earthengine.google.com/
• looked into eligibility: doesn't seem linked to the general google for work or nonprofits
• it's a computing environment, so it has upsides and downsides
• critically, it seems to have satellite imagery and derivatives only, no vectors, count(r)y borders https://earthengine.google.com/datasets/

Overall, I don't see a value in it for this particular bug.

• Oceans can be excluded — OSM has the data — from the area splitting.
• Beside the country level, lower level divisions can be generated automatically, if they're not available (in OSM or GADM), just through some area (as in km^2) criterium.
• Ranking areas by priority for mapping we could guestimate automatically, but it's a lot of data to run analysis on. Vector data, so I don't see GEE helping. And the criteria would first have to be defined better, which would require some reading.
  • But as I always say, divide et impera and the locals know best.
  • Priorities would be good if we could stand behind them firmly and then be able to estimate the extent of overall coverage ("mappedness") in a weighted way
  • if teams can be handled nationally/locally great, otherwise we almost have to do the arduous analysis or just rely on their choice by proximity
• if someone takes over an area to map and finds no waste, that's a miracle — the app is for mapping litter and higher, so this scenario is unlikely. If it happens, we just have to trust them.
  • when they find waste, great, they were there
  • I can't think of any way to verify if mapping was partial or full
  • the app could track users while mapping and we'd get excellent coverage data, but that's a privacy no-no
  • extra data can be gathered from the user when he finishes mapping an area — when they mark it as done — which can then help estimate thoroughness.