Clustering Trajecteroies

[Souri_RS]

Hello All,
I am a newbie to this fascinating model. I run the model to retrieve back-trajectories for a month using "Daily" special feature. I want to have more months; thus I run the model for another month. Now I am willing to cluster these trajectories for a source during many months (e.g. a season); however it seems that the program will only let one month observations to be clustered. Am I right? Is it possible to classify a lot of back trajectories for a source during a season for example? Or even, it it possible to get frequency map for a season?
Thank you in advance,

[barbara.stunder]

Yes it is possible to both run the cluster analysis and get a frequency map for a season's worth of trajectories. All of the trajectories need to be in the same directory (folder). Information about the cluster analysis and frequency analysis are in the Workshop materials at http://www.arl.noaa.gov/HYSPLIT_workshop.php

[sachilleos]

Hello,
I am running trajectories cluster analysis with 72 hrs to cluster but I get the error message below. However, it runs for 48 and 60 hr. Any suggestions?
thanks

ERROR - origin lat-lon for all trajectories must be identical. Differs in C:/hysplit4/cluster/endpts/.....

[barbara.stunder]

If the trajectory cluster program works for a given set of trajectories using durations of 36 and 48 hours, it should also work for the same set of trajectories for 72-hours, assuming the trajectories are at least 72-h long. Please send the set of trajectories to Barbara.Stunder@noaa.gov or tell me how I can ftp them. Thanks.

[shakir]

My work is almost similar to this post but i am using online HYSPLIT, can i do it?

[glenn.rolph]

Clustering cannot be done using the online version.

Glenn

[shakir]

Thanks for your response glenn.rolph .