global re-analysis

[trickl]

I am presently revising a manuscript on smoke plumes. The main concern of both reviewers was concerning my HYSPLIT calculations based on re-analysis data. They surprised me by pointing out that the spatial resolution is coarser than I had thought. I had never considered this because in more than eight years I have been able to assign even rather thin free-tropospheric layers in our observations to reasonal sources. Most importantly layers descending from the stratosphere could be nicely traced over many days.

The reviewers suggest to use the GDAS model. However, I have found sometimes major discrepancies with respect to observational data. In particular, in a case discussed in my paper, a pronounced fire area in Canada was perfectly hit by the re-analysis trajectories, but Canada was even completely missed with GDAS.

I would appreciate if somebody could help me in this matter.

With best wishes

Thomas Trickl

[ariel.stein]

In general, using higher resolution data should give you a better result. However for your particular case you can argue that the meteorological data that you are using, even though it is coarser than other datasets available, shows better performance. Therefore, it has a better depiction of the phenomena you want to describe. Having said that you will need to prove that the renalysis is better than the GDAS through a comparison of independent variables such as wind speed and direction and temperature (for instance).