How to use the HYSPLIT smoke dispersal tool in detail
Click on this link here.
Under Release Type select Unknown Material (Generic Mass, <24 hrs)
Under Meteorology select GFS (0.25 degree, global, 06/2019-present)
In most internet browsers, directly below the Meteorology line, a map centred in the USA should be visible. If the map is visible follow part (i), if the map is not visible, go to (ii):
(i) On the map below use the mouse to drag the map to your desired location and then click on the location. A blue marker will appear and the Decimal Degrees Latitude and Longitude boxes below the map will automatically be filled in. You can zoom in the map to pick a precise location if desired.
(ii) Enter the latitude and longitude of the location interest in Decimal Degrees Latitude and Longitude boxes.
5. Click Next>>
6. On the next page Meteorology File and other configuration information we select the forecast for the day we want to model.
(i) In the dropdown box next to Choose an archived meteorologicalfile: select the day of interest. Each file corresponds to the forecast for a single day and the naming convention is YYYYMMDD_gfs0p25 such that:
1st January 2024 = 20240101_gfs0p25
The suffix _gfs0p25 tells us the forecast source (GFS) and resolution (0.25º). Forecasts are available up to 2-3 days in advance (e.g. on 1st January, the forecast for 3rd January is available).
(ii) Set Deposition and Advanced Options to “No”.
(iii) Click Next>>
On the next page Model Run Details we select the details of the model run we want to use. Apply the following settings:
(i) Dispersion direction: Forward
(ii) Release starting time (UTC): select the start date and time for the burn. For example, a burn starting at 9 am on 30th March 2024 would have the following:
Year: 24 Month: 03 Day: 30 Hour: 9 Minute: 0
Note all times at UTC (= GMT) so if operating during BST (British Summer Time) the modelled time will be one hour before the actual time (e.g. if simulating a burn starting at 10 am BST, set Hour to be 9). You only choose a time in the same day as the day of the meteorologicalfile chosen on the previous page.
(iii) Source Latitude and Longitude should already be filled in but if they are not, put in the latitude and longitude of your burn location.
(iv) Release Top and Release Bottom: we recommend zero.
(v) Release Quantity should be 1.
(vi) Release duration should be approximately the length of time you expect to burn for.
(vii) Total duration. This is length of time for which the model simulates pollution dispersion. It should be at least the same length as the release duration and probably a couple of hours more.
(viii) Averaging period/Output interval: we recommend 1 hour.
(ix)Top of averaged layer: we recommend 100m.
(x) Compute center-of-mass trajectory: no
(xi) GIS output of contours?: Google Earth (kmz)
(xii) Plot resolution (dpi): 96
(xiii) Zoom factor: 0
(xiv) Distance circle overlay: None
(xv) U.S. county borders: No (unless looking at US)
(xvi) Postscript file: No
(xvii) Create PDF file of graphics?: yes
Click Request Dispersion Run>> to run the model
You will be taken to a new webpage with a blue Model Status box. It may take a few minutes to generate the output; the longer the duration time, the longer it will take. See next page for interpreting output.
Interpreting the output
Once the model has finished running, results will be available in two areas. First, a map will appear below the Model Status box. Second, below the map and the HYSPLIT Model Simulation Details box, there will be a table called More Results.
Map
The map provides the clearest indication of dispersion, showing the concentration of pollution in the lowest 100 m of the atmosphere. It is interactive so you can move laterally and zoom in/out. HYSPLIT shows dispersion using coloured regions where the colour of the regions indicates the minimum pollution concentration in that region.
The small red region close to the burn site shows the region with the highest concentration of pollution. The orange, yellow, blue, green, and grey regions have progressively lower concentrations with the grey region in this case having a minimum concentration 300x smaller than the red. It is important to note that the absolute values of the concentrations (e.g. 1E-10 mass / m3) should largely be ignored as we have set an arbitrary amount of material to be released (Release Quantity). The spatial dispersion of the pollution and relative concentrations are much more informative.
You can also move forward in time using the control bar in the bottom left to see the spread of pollution evolve. The default map shows the dispersion of pollution 1 averaging period after the start.
For example, if your averaging period is 1 hour (recommended on page 1) and you started your burn at 9 am and ran for 6 hours, the map will initially show the dispersion at 10 am (see timestamp in white box). Moving the blue marker left to right advances the model by 1 averaging period (i.e. 1 hour), so in the above example you would see the pollution dispersion at 11 am, 12 am and so on until 3 pm. (Pressing the > the right of the bar cycles through all the hours). You should see the plume move over time. The two plots below show the dispersion of pollution 1 hour (left) and 2 hours (right) after the start of a burn. Unsurprisingly, there is a greater dispersion of pollution 2 hours after the start of the burn.
More Results Table
The More Results table offers further data visualisation, and we recommend you explore it. The PDF Plots of the Concentration Grid 1 (below left) provide still images of the map at every increment of the averaging period (e.g. every hour) while the PDF plots of the Particle Position (below right) shows both the lateral distribution of the particles used to simulate the pollution (upper panel) but also their vertical motion (lower panel). If more of the pollution gets higher up in the atmosphere, its impact on local air quality will be lower but it could affect a wider region.