HYSPLIT Forum: hysplitbbs.arl.noaa.gov

Dear all,

I am doing a very simple concentration simulation with Hysplit. I have constant emissions from a single point source and the simulation duration is a week. I am driving the simulation using the HRRR data. I am generating min-by-min simulated concentrations at roughly 100 m x 100 m grid cell for a 5km x 5km domain. No chemistry, no deposition, very simple configuration. I am running Hysplit in pure puff mode and turn off puff split. The problem that I have is this simple simulation takes a hour or so to complete, whereas same simulation only takes about 10 min with Calpuff. So I am wondering what is causing this long simulation time with Hysplit. Is there any stupid mistake with my configuration?

One potential reason that I think may be related to the definition of the simulation domain. My interested domain is quite small, roughly 5 km x 5 km. How can I specify in Hysplit that when a released puff hits the boundary of this domain, it will be dropped from further simulation? Puff is released at every time step (1 min in my case). I am wondering is it because Hysplit is simulation lots of old puffs as the simulation goes on that is causing the long simulation time?

Thank you very much!
Ling

Puffs will continue to be simulated with HYSPLIT if it is outside of the concentration domain, but within the meteorological domain in case the puff is re-circulated back into the concentration domain. If you would like HYSPLIT to not simulate the transport of this puff after it is transported outside of a certain domain, then you can create a smaller meteorological ARL file using the xtrct_grid utility program that only covers your area of interest. xtrct_grid will extract a subgrid of a meteorological ARL file based on a user defined latitude and longitude. To run, from your working directory, type: ../exec/xtrct_grid and then follow the user prompts.

Thank you very much for the reply! That is extremely helpful!