HYSPLIT
ensemble for smoke from fires - Experimental Products
The ensemble products are created by driving HYSPLIT with members of the National Weather Service (NWS), short range ensemble forecast (SREF).
There are 26 SREF members. These are experimental products.
Link to SREF page
HYSPLIT is run 26 times over the entire country using each of 26 SREF meteorological forecasts as inputs.
One hour average output maps of primary PM2.5 air concentration between the ground and 5km are produced using
the actual fire locations from the previous day. The dispersion simulation consists of two parts: 1) a 24 h analysis simulation
run for the previous day, and 2) a 48 h forecast simulation. The smoke particle positions at the end of each analysis period
are used to initialize the next days' analysis simulation.
Output from the 26 runs is combined to create products which characterize the uncertainty in the HYSPLIT model predictions of smoke
concentrations which arise from uncertainties in the NWP model output used as input into HYSPLIT.
Animation showing surface PM 2.5 average ensemble mean concentration (ug/m3).
Request other products below
No current products available at this time
Request Products
Ensemble mean shows the mean concentration over all the ensemble members at each location. APL : Agreement in percentile level. At each point the concentration for
which N percent of the members predict values below that concentration.
For example, the concentration at the 50th percentile level is
the concentration for which half the members predict values below that
concentration and half predict values above. The concentration at the 90th percentile level
is the concentration such that 90 percent of the members predict concentrations
below that level. Note that the 50th percentile corresponds to the
median model; concentration at each point is the median of all the
ensemble outputs.
ATL : Agreement in threshold level. The percentage of ensemble members which
give a result greater than a defined thresh-hold.