HYSPLIT Forum: hysplitbbs.arl.noaa.gov

Dear sir/Madam,

I am doing some research by using HYSPLIT to simulate some particle deposition. In order to analyze the output, I mounted the points in the cdump.txt to the google earth, and at the same time ,the deposition value of each point displayed near the pointmark. I found that the deposition value varied from the small value to big value , then to small value. This trend is different with the measured trend in the field experiments. When we measure in the field experiments,the deposition value is the biggest near the emitting source , and long the distance from the source ,the deposition becomes small. So we wonder how can solve the problem.

I am so sorry about my English. In the attaching pic ,there is one deposition output by HYSPLIT.

Thank you very much.

How did you created your cdump.txt file?

I attach the control file for this running.

I would not be surprised to see differences with this very high resolution run. I do not think it is adequate using 12 km meteorological data for this 0.001x 0.001 degree (~100 m) resolution concentration grid. You will need a higher resolution meteorological data to drive the model.
Also, your runs will be very sensitive to the number of lagrangian particles. How many are you releasing for this run? (SETUP.CFG)

If we need this high resolution, what kind of meteorological files are possible to drive the model? The setup.cfg file has been attached. The numpar=25000, and maxpar=100000, is it ok?

Many Thanks.

You could use meteorological data from the WRF model but you will have to generate your own model runs.
Regarding the number of particles, I would increase the number until you see no change in the results. That means that your results are independent of the number of particles.

I wonder whether the wrf files could input the HYSPLITmodel directly?

We have a fortran program called arw2arl.exe that is located in your exec directory that will convert the WRF output to a format that is compatible with HYSPLIT.