search

jfaghm / OceanEddies

A collection of algorithms to autonomously identify and track mesoscale ocean eddies in sea surface height (SSH) satellite data
92 stars 51 forks source link

Rotational speed #14

Open claudiofgcardoso opened 7 years ago

claudiofgcardoso commented 7 years ago

Hello,

First of all, thank you for making this tool freely available. I'm using MatLab's version to track and characterize the eddy field around Cape Verde. Everything worked fine! However, I'm getting seemingly incorrect geographical patterns for the rotational speed.

eddies_inside_rot_speed_30days_v2

I posted a similar result in a comment related with the same issue. What bothers me the most is the close to zero speed values... Is this problem related with the "tolerance_track_lnn" or the "process_eddies_and_tracks_tolerance"? Or am I doing something wrong?

Greetings, Cláudio Cardoso

lematt1991 commented 7 years ago

Which detection (v1, v2, or hybrid) and tracking (lnn or mha) algorithms are you using? You might try comparing a few other approaches. I'll defer to @jfaghm for an official response though.

claudiofgcardoso commented 7 years ago

I used the already processed eddy features (Global mesoscale ocean eddy features) available at http://datadryad.org/resource/doi:10.5061/dryad.gp40h. For the tracking, I used the lnn algorithims with a 3-day tolerance.

Is it possible that the problem is related with this dataset?

claudiofgcardoso commented 6 years ago

Hello again,

@ml9951 Following your suggestion, I scanned the original AVISO SLA dataset for eddies using both v1 and v2 detection methods. Unfortunately there is no improvement in the final results...

image

This image was created by computing the mean geostrophic speed within the daily .mat files generated with the function "scan_multi.m". So, the problem must be in the calculation of the mean geostrophic speed and not in the tracking (lnn or mha) algorithims.