|
Source/Receptor Concentration Matrix
Using Hysplit4
Roland Draxler
Revised Method, June 2001
Background
Frequently it is necessary to determine which pollutant
sources may be affecting specific receptor (or sampling)
locations. One simple approach, the computation of back
trajectories from the receptor location, gives an indication
of the origin of the pollution, but it does not account
for the complexities of atmospheric dispersion. An alternate
approach is to compute the transport and dispersion from
all sources to compute the contribution of each source
to a specific receptor. The latter approach is followed
here.
Modeling Details
First the Hysplit4
model was reconfigured to run multiple sources simultaneously.
Corresponding with each sample averaging period, air
concentrations are output on a grid for each source location.
One unit of pollutant per hour is emitted from each source
for a 24 hour period at the beginning of the simulation
cycle. There is one simulation cycle per day. That is
at the beginning of each day, a new cycle (24 hour emission)
is started. However the pollutant particles from yesterday's
and all previous emission cycles continue to be advected
and dispersed until they pass out of the computational
domain. Each emission cycle is maintained as an element
in the matrix, and therefore it is possible to examine
the source regions for a receptor independently for each
previous day's emission cycle.
Pollutant Sources
Pollutant sources are defined at each 1 degree latitude/longitude
intersection over the eastern half of the United States
(25N-50N and 100W to 66W). The concentration grid is defined
at a resolution of 0.5 degrees over the whole continential
US and the results are averaged over 24 hours from 0000
to 2400 UTC each day. The configuration results in the
computation of transport and dispersion from 710 independent
sources (points over the Atlantic Ocean excluded). The
model is run once each day, based upon the 0000 UTC NCEP
ETA forecast.
Interpreting Results
The model's output file
is examined by selecting a latitude and longitude at either
a source or receptor location. A selection of a source point
will result in a map of air concentrations at all receptor
locations (the concentration grid; this is the same as running
a concentration run of Hysplit on READY).
The selection of a receptor point will result in a map of
normalized air concentrations as contributed by each source
location to the receptor point. This normalized concentration
can be multiplied by the emission (if known) from each source
location to obtain an estimate of the receptor air concentration.
Alternately, the dispersion factor can be divided into any
measurements made at the receptor to obtain an estimate of
the emissions from a particular source location. For instance,
if the dispersion factor from source A to receptor B is indicated
to be 1.0E-15 and the model simulation emission rate was
1 unit/hr (total = 24 units), then if the pollutant emission
rate at A is really 1.0E+12 ug/hr, then the air concentration
at B from A must be 1000 ug/m3.
Contact Information
Roland R. Draxler
e-mail: roland.draxler@noaa.gov
|
