Understand Back Trajectory and the Trajectory Endpoint File

[masisneros]

I was wondering if you could assist me. I ran the Hysplit Trajectory Model (Backtrajectory) Job #12473. I've looked thru your Q&A and the handbook and trainings and I'm not sure what idea behind the backtrajectory I'm missing.

My question/problem is when I run this particular job thru the online vs pc version my trajectory files are different. On my online run, my trajectory endpoint files and if I look at the control file as well, it appears that the trajectory goes past the 24 hours I had indicated the model to run, it goes like 48 hours or so. When I run it on PC version, the model stops at 24 hours.Why is the online version going past the 24 hours I had indicated for the model to run from my start time? It appears that it is running each trajectory for 24 hours. Now is this supposed to happen and if so what aspect of back trajectory could i be missing. And which run is the correct one, the online version or my pc version?

My second question is, how does one determine how many trajectories are useful to run per each height level that you are starting at (i.e. 10m, 100m, 500m, etc) up to the 6 levels that I can run in the pc version. Also how do you decide how often to start each trajection for example, after 1hr vs 4hrs. My problem is I'm trying to do a backtrajectory to see what is going on at an air monitoring station, so I don't understand the theory or thinking behind how many trajectories and how often to start them.

[glenn.rolph]

Looking at your SETUP.CFG file, it looks like you set up the web run to have 24-hour back trajectories start every 24 hours and when looking at the cross-section at the bottom of your plot (attached) that is what is happening. Therefore, I would guess that you put 4 into the box that asks "Start a new trajectory every:", you also set the model to run for a total of 24 hours and the maximum number of trajectories to create as 6 to get the values that are in the SETUP.CFG. If this is not what you want than set the "start a new trajectory every:" to 0 and only one back trajectory will be created.

The answer to your second question is not as easy. The number of trajectories needed to define source receptor issues will depend on the complexity of the meteorological situation near your sampler. You should look at the vertical shear by running several trajectories in the vertical or else run the option on the PC to choose a percent of the mixed layer as the starting location. You should also run trajectories several hours before and after your sample to see if there was a synoptic change in the meteorology that will give some uncertainty to your result. You can find more answer to this question on our FAQ at:

http://www.arl.noaa.gov/HySPLIT_FAQ.php

SETUP.CFG:
&SETUP
KMSL=0,
tm_terr=1,
khmax=24,
nstr=4,
mhrs=24,
/

[masisneros]

Would it be possible to get a more detail explanation of what is meant by: You should also run trajectories several hours before and after your sample to see if there was a synoptic change in the meteorology that will give some uncertainty to your result. You can find more answer to this question on our FAQ at:

http://www.arl.noaa.gov/HySPLIT_FAQ.php (I wasn't sure which question applied, sorry)

Also what is meant by a synoptic change in meterology...I guess what is it I'm looking for or what you have a good reference I could look at that may talk about this.

Maria