John Roads
Scripps Experimental Climate Prediction Center
UCSD, 0224
La Jolla, CA
jroads@ucsd.edu
858-534-2099
2nd International RSM Workshop participants. Front Row: Kuo, Shiao, Wu, Chen, Feser, Chen, Kodama, Roads, Fujioka. Back Row: Zhang, Anderson, Chambers, Brenner, Benoit, Hong, Cui, Sun, Vergara, Han, Kanamitsu, Juang, Smith, Misra, Stevens
The 2nd international Regional Spectral Model (RSM) workshop, which was hosted by MHPCC, the Scripps Experimental Climate Prediction Center (ECPC), the Hawaii Weather and Climate Modeling Ohana (HWCMO), the US Forest Service (USFS), NCEP, the International Research Institute (IRI), NOAA?s office of global programs (OGP) and COMPAQ Computer Corporation, was held Jul. 17-21, 2000 at the Maui High Performance Computer Center (MHPCC), Kihei, Maui, Hawaii.
The
major purpose of the workshop was to discuss the RSM, which was originally
developed at the National Centers for Environmental Prediction (NCEP) to provide
regional details for the NCEP global spectral model (GSM; see Juang and
Kanamitsu, 1994),).? Since the RSM uses
the same physics and code structure as the NCEP GSM, the workshop not only
discussed the RSM but also the parent GSM.
A
growing number of users are using the GSM and RSM to simulate and forecast
regional climate in a variety of regions. These regional simulations and
forecasts are thus helping the atmospheric modeling community to better connect
to various application communities, which need the highest resolution possible.
Our goal at this workshop was to attempt to better organize and communicate
some of the features of our individual GSM and RSM model setups, biases, and
numerical techniques for various regional climate (and weather) simulations.
The
workshop began with a tutorial on the RSM by the NCEP participants. M.
Kanamitsu provided an overview of the global and regional modeling system.
Henry Juang discussed the numerical techniques and nonhydrostatic version of
the model. S. -Y. Hong discussed the RSM and GSM model physics. Jongil Han
discussed simulations with the new RSM, which will soon become available to the
community. A standard RSM user manual has now been developed. This manual is
available on the web at /RSM/manual. This manual will be
updated by the RSM model master as the model is updated.
Model
resolutions ranging from 2 kms to 200 kms were discussed at the workshop. As
was discussed by Roads, the ECPC () uses both the GSM and
the RSM to make experimental climate forecasts and simulations. For that
reason, Reichler compared some simulations with the GSM using the older
simplified Arakawa-Schubert parameterization (SAS) with the current GSM, which
uses a relaxed Arakawa-Schubert parameterization (RAS). Results are still
somewhat inconclusive as to which parameterization provides a better simulation
since different variables are affected differently and further experiments with
other convective parameterizations are still being carried out at NCEP.
Roads,
Anderson, Stevens, Zhang, discussed various aspects of Hawaii RSM Forecasts by
the weather service and HWCMO.? The
HWCMO currently makes near real-time forecasts out to 48 hours and displays a
number of parameters on its web site (http://www.mhpcc.edu/~wswx/) and makes
this data available to the Pacific Disaster Center (PDC) and other Hawaii
agencies. Of special interest are the fire danger indices, which are being
increasingly used by PDC and various civil defense agencies to describe fire
danger in Hawaii. As was true for a number of ongoing studies, sufficient
numbers of forecasts are now beginning to be archived for evaluation purposes.
There
were a number of US RSM simulations that were discussed. Besides the ECPC
forecasting effort for the US, CA, Southwest, and Brazil, Roads discussed some
of the water and energy budgets simulated in a continuous climate simulation
for the US. Anderson discussed characteristics of the southwest monsoon
simulation developed by the RSM. Hong discussed characteristic features of the
Oklahoma-Texas drought during 1999. Although the workshop stressed climate
simulations and forecasts, Du discussed how NCEP is now attempting to develop
short-range ensemble forecasts with the RSM as well as the Eta model.
Chen,
Sun, and Misra discussed characteristics of Brazil forecasts. The IRI is
currently sponsoring an intercomparison project over Brazil, which will soon be
joined by the participants from the Project for Intercomparison of Regional
Climate Simulations. This workshop provided a first glimpse of the RSM
characteristics and biases for these simulations. In particular, Chen discussed
a drying feature of the RSM soil moisture that needs to be corrected for
climate simulations. A number of possible solutions were mentioned but no
definite resolution is yet available. Developing appropriate soil moisture for
these regional models is a serious problem, which is also being tackled by the
international land data assimilation system (LDAS) groups associated with the
Global Energy and Water-Cycle Experiment (GEWEX).
A
number of talks addressed the use of the RSM for Asia forecasts. Shiao
discussed simulations over eastern Asia, which were developed to understand
appropriate domain size and resolution. Kuo discussed development of a regional
model to study the eastern Asian Summer Monsoon and mentioned that there were
problems in simulating intraseasonal variability with the GSM and RSM. Chen
discussed some CO2 simulations focused on Taiwan. Wu then discussed how these
RSM forecasts could be used to drive a hydrologic model for water resource
applications. Apparently there are still a number of biases that need to first
be resolved before these models can be used to accurately simulate variations
in regional climate. Chen suggested a coarse resolution RSM should be run to
first remove systematic biases.
Regional
simulations for a number of other regions were also discussed. Vergara
discussed some of the problems with developing forecasts for Chile that
properly resolved the Andes and thus provided adequate snow and precipitation
in this region. Brenner discussed the use of the GSM and RSM for Mediterranean
forecasts. These forecasts are aimed toward eventually producing wave
forecasts, which may provide another validation tool for the surface wind and
heat flux forcings. Feser discussed a wave damping technique that was developed
to combine forecasts from a global analysis and a regional climate model. Lin
provided a general overview of a coupled hydrometeorological Canadian model and
possible components that may be useful for future hydrometeorological
development of the RSM.
On the
last day, a general discussion was held. ECPC, in collaboration with NCEP, is
now hosting the RSM, as well as providing an RSM model master, Dr. Jongil Han,
to answer questions about the model (jhan@ecpc.ucsd.edu) and provide onsite
help if requested. NCEP and Dr. Han?s current major goal is to develop an MPI
version of the RSM, which will be similar to the current MPI version of the GSM.
The
workshop then ended with a tentative discussion of the 3rd
international workshop, which will be held somewhere in Taiwan around July
2001. In summary, the RSM community is alive and well and welcomes additional
participants.
Roads, J.O. 2000:
The Second International RSM Workshop: Meeting
Summary. Bull. Amer. Meteor. Soc. Vol. 81 (12) p. 2979-2980.
Juang, H. -M. H., and M. Kanamitsu, 1994: The NMC
nested regional spectral model. Mon. Wea. Rev., 122, 3-26.