Backward Trajectory Model Run

Topics about the HYSPLIT trajectory model.
Post Reply
Posts: 2
Joined: January 21st, 2014, 10:31 am
Registered HYSPLIT User: No

Backward Trajectory Model Run

Post by teoxu »

Hi there,

I am new to HYSPLIT and am using it for my thesis to run backward-trajectory analyses and have a few questions pertaining to the model. I have read through all the FAQs and help files regarding the questions below but have yet to find satisfactory solutions. I am currently using the web version as I do not yet have many backward-trajectories to run. I seek your advise on the following:

1.) Under "Model Run Details" of the model, does "Level 1 height" (together with Level 2 and 3 heights), represent the release height of the particle? I am trying to trace radon in the atmosphere as I already have previously collected radon concentration data from a 230m tall tower - the gas inlet is at 222m. The pressure readings have been in the range of approximately 850 - 980mb. The site is located approximately 400m above sea level. What height should I input into the field?

2.) What will the ideal number of hours to run the model for to minimise any sort of trajectory calculation error? I am trying to trace radon from within the UK and possibly slightly further afield in the Atlantic.

3.) Would you recommend running only 1 trajectory at a time at a single height?

I apologise in advance for any questions that may seem elementary in nature.
Thank you very much!

Posts: 451
Joined: November 9th, 2012, 4:23 pm
Registered HYSPLIT User: Yes

Re: Backward Trajectory Model Run

Post by barbara.stunder »

HYSPLIT trajectories are based on the mean wind, they do not account for turbulence and vertical mixing. A back-trajectory calculation from the height/location of the sampler will most likely not be representative of the actual path of air to the sampler because of turbulence and vertical mixing. An approximation would be to start the back-trajectory at a height in the middle of the boundary layer. There is some description of trajectory error in our workshop materials at It is better to run several trajectories (e.g. different start heights, times, and locations) to get an estimate of the uncertainty of the single trajectory of interest.

1.) Yes, “Level 1 height” is the release height of the particle, entered as meters AGL or AMSL as specified. The ground level is that in the meteorological model not actual ground level, which introduces uncertainty. Use a mid-boundary-layer height.

2.) There is no ideal number.

3.) As it states on the web under “More info” about the trajectory start height, there can be sensitivity to the time step for multiple-level runs. In general, it is good to run multiple levels since a single trajectory may be unrealistic of actual transport and dispersion. However you should look for any differences between multiple-level trajectories run in one run vs. running them separately. Differences depend on the 3-d wind field.
Post Reply