Ensemble Model

[Shawn]

At present, I have come across a problem with respect to the ENSEMBLE trajectory model calculation: No matter the online version or the PC version, I found if the backward calculation was set as above mean sea level (amsl), the initial points of the 27 back trajectories would be at different levels from the surface to the high level; but when set as agl, they would start at the same altitude as I input. Actually, the station altitude should be 1344m amsl.Why is this happening? Are there any difference in calculating the backward trajectory in ensemble model with AGL and AMSL? How should I choose the correct model parameters?

[barbara.stunder]

Please post your CONTROL and SETUP files, so I can reproduce it.

[Shawn]

Thank you for your reply. Here are the two files and the level height should be 1344m above mean sea level in ENSEMBLE model. To make my question clearer, I've attached the pic of the result where the starting heights are at several points for different BTs though I only set one height for simulation.

[barbara.stunder]

This is a feature.  HYSPLIT's default is heights AGL.  As stated https://ready.arl.noaa.gov/hysplitusersguide/S611.htm in the section "Height Unit for Input", "Setting this parameter to "1" [for above mean sea level] forces the model to subtract the local terrain height from source input heights before further processing."  

If you look at the trump files for an msl and agl trajectory ensemble (backward or forward), at the top section where the 27 start time/lat/lon/heights are given, they both show the same height (in your case 1344).  However for the 1st endpoint in the main section of the file, for the agl case, it gives 1344 as height for all 27 trajectories, but for the msl case it varies from 0 to 1290 m because it's subtracting the terrain height as defined in the model (GDAS).  

The terrain varies a lot near your latitude, longitude, making the difference more pronounced.