search

cemac / mosaic_traj

Python package for reading and plotting MOSAiC trajectories
BSD 3-Clause "New" or "Revised" License
0 stars 1 forks source link

Handle data with multiple clusters #14

Open tdjames1 opened 2 years ago

tdjames1 commented 2 years ago

From @sarnoldlds:

The clusters don’t necessarily have to be on different pressure levels (could also be different location etc).

In the trajectory file header there is a field called “CLUSTER POINTERS”. These are the trajectory numbers that correspond to the first trajectory in each cluster. In this case, they are just sequential consecutive numbers as there is only one trajectory per cluster. In other examples, there may be more than one per cluster and these numbers give the first in each cluster.

The usual way the model and the clusters would work is that the trajectories within a single file would have the same time of release (in some whole reanalysis interval -in our case a 3 hour UT time point within the day to align with the reanalysis data times). The clusters would then be at different pressure or location releases at that same time. This is the case for the example ALPACA files I sent you.

The release at different times is a specific case for the model which is not actually the standard operation. I don’t think I have so far used clusters with timed release but it could be an option.

tdjames1 commented 1 year ago

Commit 1d2b287 includes initial work to handle clustered data structure in trajectory plotting code, but this needs to be extended to implement plotting for data with multiple clusters. Need to decide how this will be done e.g. could use different colours for clusters.