Fix observation location using other observations

To quickly fix the location of an observation, select one or more observations, and automatically generate a location overlapping the locations of these observations.

1 observation selected --> take its location.
2 observations selected, or more --> disc overlapping the locations of these observations.

The selection of the other observations could be done by entering observation reference numbers or by making easy the selection on the map of one previous observation and one next observation. Ideally, the observer needs to the find and select the 1st previous observation and the 1st next observation whose locations are correct.

This would be vey useful when we have a long sequence of observations, half of them with correct locations, half of them with wrong locations.

you may want to look at: https://forum.inaturalist.org/t/interpolating-coordinates/5170. i’ve never used this functionality myself, but it seems like it might be useful in the situation you’re describing.

1 Like

Interesting. But we need to be able to choose the observations. Only the observer knows which ones have correct locations.

We may have 1 observation with correct location, followed by 4 with incorrect locations, followed with 1 observation with correct location. So, for each of these 4, the location needs to be fixed based on these 1 + 1 observations. Whatever the technical solution is, it is needed to first select the observations with correct locations.

See for instance this observation and the next ones, whose locations I fixed manually, indicating in the notes which observations had initially correct locations:
https://www.inaturalist.org/observations/35847445
Many wrong locations, with a few interspersed correct locations.

Related topic:
https://forum.inaturalist.org/t/add-interpolating-coordinates-to-gui-for-observations-without-location/31767

1 Like

This would only be reasonably accurate if you proceeded in a straight line and the time between observations is equal.
If you made 3 observations and the first and third were a mile apart, then the second would be estimated at 1/2 mile between when in fact it may only be a few feet from one or the other.

How do you get wrong locations in your data set? I can imagine not precise locations, but wrong ones?

How: I don’t know, it’s a fact, it comes from my smartphone.

See for instance (and check the locations directly from the photos):
https://www.inaturalist.org/observations/35847309
https://www.inaturalist.org/observations/35847445
https://www.inaturalist.org/observations/35847448
https://www.inaturalist.org/observations/35847483
https://www.inaturalist.org/observations/35847543
https://www.inaturalist.org/observations/35847550
https://www.inaturalist.org/observations/35847553
https://www.inaturalist.org/observations/35847614
https://www.inaturalist.org/observations/35849832
https://www.inaturalist.org/observations/35850076

It’s not the only location where it happened. When it happens, about half of the observations have all exactly the same wrong location, as far as I remember.

1 Like