Integrate over time series

Topics about the HYSPLIT dispersion model.
sle20
Posts: 10
Joined: August 4th, 2016, 10:51 am
Registered HYSPLIT User: Yes

Integrate over time series

Post by sle20 »

Hello,

I was wondering if you could tell me how to integrate over a time period, so it gets the probability of where the air has come from. As supposed to at the moment it seems to produce a picture every 12 hours as a seperate image.

Also I have been workign my way through the tutorial and there is no folder called Tutorial/batch

Many thanks,
Stephen
sle20
Posts: 10
Joined: August 4th, 2016, 10:51 am
Registered HYSPLIT User: Yes

Re: Integrate over time series

Post by sle20 »

Think found what i want in Ensemble Probability display (under concentration) but it won't run due to a missing file:

forrtl: severe (29): file not found, unit 101, file c:\hysplit4\working\fort.101





Any idesas what this file contains or how to obtain it? Couldn't see anything about this on the tutorials
ariel.stein
Posts: 660
Joined: November 7th, 2012, 3:14 pm
Registered HYSPLIT User: Yes

Re: Integrate over time series

Post by ariel.stein »

Please check the following information
https://ready.arl.noaa.gov/documents/Tu ... _freq.html
We will post a more extended answer shortly.
sle20
Posts: 10
Joined: August 4th, 2016, 10:51 am
Registered HYSPLIT User: Yes

Re: Integrate over time series

Post by sle20 »

I think that is the kind of thing I am after (will have to check with supervisor) but this is for a forward trajectory, I followed the same steps again with backward but on the Trajectory>Special Runs>Daily is this running backwards as start day cant be after the end day.
sle20
Posts: 10
Joined: August 4th, 2016, 10:51 am
Registered HYSPLIT User: Yes

Re: Integrate over time series

Post by sle20 »

Thank you very much, but is there a way to do it with concentrations as supposed to trajectories? As in lots of particles?
MarkCohen
Posts: 19
Joined: August 5th, 2016, 11:56 am
Registered HYSPLIT User: Yes

Re: Integrate over time series

Post by MarkCohen »

Not exactly sure what you are trying to do, but I think maybe you could try to do a "backward dispersion" run, where you run the concentration model, but run in backwards in time, like a back-trajectory. In such a simulation, you are in effect doing lots of back-trajectories (i.e., one for each computational point particle that you are including in the dispersion simulation).

When you do this simulation, you can specify the output grid(s) and time averaging parameters in the Concentration / Setup Run / Grid Definition

This concentration grid set up (for forward or backward dispersion) is described in the HYSPLIT User Guide (http://www.arl.noaa.gov/documents/repor ... _guide.pdf) in the Concentration / Grid Definition section.

So, for example, if you do the backward dispersion simulation for 1 week (-168 hrs), you can set the output concentration grid to, say, give you just one output at the end, or once every 24 hours (or whatever you want), and you can set whether you want snapshots at the output times, or averages over the output periods.

Here's the relevant section on setting up the concentration output parameters from the User Guide:

25(10)- Sampling interval: type hour minute

Default: 0 24 0

Each grid may have its own sampling or averaging interval. The interval can be of three different types: averaging (type=0), snapshot (type=1), or maximum (type=2). Averaging will produce output averaged over the specified interval. For instance, you may want to define a concentration grid that produces 24-hour average air concentrations for the duration of the simulation, which in the case of a 2-day simulation will result in 2 output maps, one for each day. Each defined grid can have a different output type and interval. Snapshot (or now) will give the instantaneous output at the output interval, and maximum will save the maximum concentration at each grid point over the duration of the output interval. Therefore, when a maximum concentration grid is defined, it is also required to define an identical snapshot or average grid over which the maximum will be computed. There is also the special case when the type value is less than zero. In that case the value represents the averaging time in hours and the output interval time represents the interval at which the average concentration is output. For instance, a setting of {-1 6 0} would output a one-hour average concentration every six hours.
sle20
Posts: 10
Joined: August 4th, 2016, 10:51 am
Registered HYSPLIT User: Yes

Re: Integrate over time series

Post by sle20 »

Mark,

I had seen that bit of the documentation but it wasn't till you said it that I saw how to change the grid settings! So thank you very much. Do you have any idea what would cause the plot to have huge gaps-see attached file.

I also attached my expected output in my other question (which is same as this one)

Many thanks for your help

Stephen

p.s
Also about half of the time when I run this, the loading percentage, after pressing run just doesn't move and works again after reinstalling?
Attachments
Why is there gaps.PNG
MarkCohen
Posts: 19
Joined: August 5th, 2016, 11:56 am
Registered HYSPLIT User: Yes

Re: Integrate over time series

Post by MarkCohen »

Hi Stephen,

Couple of thoughts I can offer re: the output you showed.

First, the concentration layer that you are collecting data on is from 0-100m which is relatively "thin". You might want to increase this to 200m or 500m or something to see what sort of difference this might make. The particles appear to be making it inside the 0-100m layer only sometimes, and so, you see a sort of discontinuity. You could write out and examine a "Pardump" output, and you could look in the Tutorial on how to do that. But in the meantime, see what happens if you pick, say, 500m as the thickness of your concentration layer.

Second, it seems that the grid size of your concentration grid may be really small. If you are trying to show transport over 100's or 1000's of km, then you don't necessarily need a super fine concentration grid. Maybe try 1 x 1 degree or something? Alternatively or in addition, you might want to increase the number of particles in the calculation, by increasing numpar and maxpar in the Advanced configuration setup. If there aren't enough particles to reasonably populate the concentration grid, you can get odd looking results. In other words, the dispersion may be in the area, but there aren't enough particles to populate all the grid squares so it seems like some are empty.

Regarding the freezing up of the program, I'm not sure what might be happening. You really shouldn't have to be reinstalling. When you say "half the time", it sounds like something very bizarre is happening that corrupts the installation. And then you have to reinstall to create a fresh, uncorrupted version. I'm not sure what could cause this, as haven't ever seen anything like it. But, note that when you start a run, it will keep going until it ends or crashes. Just exiting out of the GUI will not kill the run. You'd have to kill the process from the Windows Task Manager (if you are using Windows) or some equivalent method in other operating systems.

Finally, if possible, could you share the CONTROL, SETUP.CFG, and MESSAGE files for the simulation that is giving you troubles? Then whoever is trying to help can see better what you might be doing.
sle20
Posts: 10
Joined: August 4th, 2016, 10:51 am
Registered HYSPLIT User: Yes

Re: Integrate over time series

Post by sle20 »

Many thanks,

Goign to try that, quick Q:

I've worked out how to change the initial speed but how do you changet he heights as you sugegsted to do?
MarkCohen
Posts: 19
Joined: August 5th, 2016, 11:56 am
Registered HYSPLIT User: Yes

Re: Integrate over time series

Post by MarkCohen »

Hi Stephen,

The concentration grid layer levels are specified when you specify the concentration grid. The image below shows the steps you'd take to do this in the GUI:
Picture1.png
Post Reply

Return to “Dispersion Model”