Meteorological
Research at the Idaho National Laboratory Site: Improved
Atmospheric Dispersion Modeling for the Idaho National
Laboratory Site
Background
The Field Research Division of the National Oceanic and
Atmospheric Administration Air Resources Laboratory (NOAA
ARL-FRD) provides meteorological support to the INL Site. This
includes maintaining the INL Site meteorological tower network
(the INL Site Mesonet) and running atmospheric dispersion models
for emergency response applications. For many years NOAA ARL¬FRD
has used a locally developed dispersion model called MDIFF to
simulate potential hazardous releases from INL Site facilities.
However, MDIFF has not been significantly upgraded in many years
and is showing its age. Both INL Site and NOAA staff have
requested new modeling capabilities that are not present in
MDIFF. While MDIFF could in principal be upgraded, it is more
cost effective to adopt a newer model that already has many of
the requested capabilities.
The NOAA Air Resources Laboratory already has an advanced
dispersion model called HYSPLIT that contains many of the
features that have been requested for future INL Site
applications. HYSPLIT is used operationally within NOAA for
various applications including smoke forecasting from wildfires
and forecasting the movement of ash plumes from volcanic
eruptions. It is also used by the National Weather Service to
produce plume forecasts for toxic releases.
NOAA ARL-FRD recommended several years ago that HYSPLIT be
adopted for INL Site use, but funding restraints did not allow
any efforts along these lines until 2007. Adoption of HYSPLIT
benefits both INL Site and NOAA as a whole. INL Site benefits
because HYSPLIT is a more modern model with new modeling and
display capabilities. Unlike MDIFF, HYSPLIT also has broad
support within NOAA, so the limited resources available locally
can be leveraged through interactions with the broader HYSPLIT
community within NOAA. The benefit to NOAA is that any
improvements to HYSPLIT made for INL Site applications can feed
back to the wider HYSPLIT community.
Objectives
The objectives of this work are to:
Transition INL Site dispersion modeling from MDIFF to
HYSPLIT
Develop more realistic wind fields that account for the
local topography and changes in the wind with height
Improve the dispersion model output so it is more useful
to decision makers in the INL Site Emergency Operations
Center
Provide capability to forecast future plume movements
using gridded atmospheric models
Develop release scenarios for INL Site facilities that
are compatible with the HYSPLIT model inputs
Accomplishments Through 2007
Many of the capabilities required for using HYSPLIT at INL
Site actually coincide with the needs of other HYSPLIT users
within NOAA. NOAA ARL-FRD has therefore collaborated with these
other groups in adding many features. These include improving
the model output so that it more useful for decision makers and
adding a chemical database within HYSPLIT that is used in
generating contours. Output from the model can be provided in a
GIS format that allows it to be layered with other geographical
data. A basic radiological dose calculation algorithm has also
been added, although it is still undergoing testing.
NOAA ARL-FRD is running a version of the WRF gridded
atmospheric weather prediction system with a horizontal grid
spacing of 4 km over the INL Site region. The forecast winds
from WRF can be used in HYSPLIT to create forecasts of future
plume movements.
Results
HYSPLIT now has many of the capabilities that are needed for
INL Site dispersion modeling. The main obstacle remaining is the
development of a realistic three-dimensional wind field based on
the INL Site Mesonet data, and this is discussed below.
Plans for Continuation
Unlike MDIFF, HYSPLIT requires a three-dimensional wind field
that accounts for vertical changes in the winds and for terrain
effects. This can already be done using the output from gridded
models such as WRF, but it must be remembered that these gridded
models are only producing forecast winds and not actual winds.
These forecasts can fail just like the models used by the
National Weather Service in producing general weather forecasts.
For this reason, it is not advisable to rely solely on forecast
model winds when generating plume plots in an emergency-response
environment.
NOAA ARL-FRD therefore plans to develop a capability to generate
a three-dimensional HYSPLIT wind field based directly on the INL
Site Mesonet data. Such a wind field is a “nowcast” rather than a
forecast, since it involves current conditions rather than
future conditions. MDIFF currently creates two-dimensional
nowcasts simply by horizontally interpolating between Mesonet
towers. The situation is more complicated with HYSPLIT because
of the need to deal with vertical variations in the wind. In
addition, it is highly desirable to have the HYSPLIT winds flow
realistically over or around terrain obstacles and to obey mass
continuity. Simple interpolation does not produce such results.
NOAA ARL-FRD is therefore looking at either adopting or
developing a so-called diagnostic wind-field model that matches
the Mesonet observations as closely as possible while still
obeying physical constraints such as terrain blockage and mass
continuity. Once such a wind-field model is in place, the
transition from MDIFF to HYSPLIT can be initiated.
Investigators and Affiliations
Richard M. Eckman, Meteorologist,
National Oceanic and Atmospheric Administration, Air Resources
Laboratory, Field Research Division, Idaho Falls, Idaho Funding Sources
Funding
Sources
U.S.
Department of Energy Idaho Operations Office
National
Oceanic and Atmospheric Administration, Air Resources Laboratory
