Ruben
Description:
I am a recreational fisherman in Tarragona, Spain, and I maintain a small
personal, non-commercial project that analyses fishing activity along my
local stretch of coast. Nothing is published or monetised; it runs on a
Raspberry Pi at home and the outputs are for my own use.
I already hold an API token and I am using the Vessels, Events and 4Wings
APIs successfully. Thank you for making them available - the purse seine
analysis has worked very well.
REQUEST
dataset: public-global-all-tracks:latest
endpoint: GET /v3/vessels/{vesselId}/tracks
This currently returns 403 "You do not have permission to access this
dataset". The request appears otherwise well formed: four different
parameter combinations all return the same 403 rather than a 422.
WHAT I AM TRYING TO UNDERSTAND
I am studying the relationship between where the trawl fleet works and the
morphology of the seabed in a small area (roughly 40.3-41.5 N, 0.4-2.1 E).
Trawlers cannot work rough or rocky ground, so their spatial behaviour is an
indirect record of the bottom - useful both for understanding the fishery and
for choosing where to fish with rod and line from a small boat.
WHY THE AGGREGATED GRID IS NOT ENOUGH
I first tried this with the 4Wings apparent fishing effort grid, looking for
unworked cells surrounded by heavily worked ones. It does not work, and I
think the reason is instructive: at HIGH spatial resolution the cells are
about 1.1 x 0.8 km, while the seabed features that stop a trawl are 80-300 m
across - one to two orders of magnitude smaller. A gap in the grid is
therefore not evidence of an obstruction.
I tested this properly rather than assuming. I took three of the strongest
candidates out to sea and surveyed them with a depth sounder and side scan
sonar. There was nothing at any of the three, nor around them. The method
fails, and it fails for a resolution reason rather than a tuning one.
WHY TRACKS WOULD ANSWER IT
With position-level tracks I could look for the signature that actually
indicates a fixed obstruction: vessels altering course to steer around the
same point, repeatedly, across many independent trips and over a long period.
That is a minute-scale pattern and it simply is not visible in a
kilometre-scale grid, at any level of filtering.
SCOPE AND USE
About 150 trawlers, one year of history, in the area above. I would fetch
this once and cache it locally rather than polling. Non-commercial use only,
under your terms, with attribution. I am happy to share back what I find
about the limits of the effort grid at this scale if that is of any use to
you.
Thank you for considering it.