Transcript CANSAC MECHANICS - Desert Research Institute

CANSAC MECHANICS
Julide Kahyaoglu-Koracin
CANSAC Workshop
16 February 2006
Sacramento, CA
CANSAC Operations
• Program for Climate
Ecosystem and Fire
Applications (CEFA) at
Desert Research Institute
(DRI) in Reno, NV, facilitates
the operational component
of CANSAC.
• Operational functions of the
center can be viewed under
two groups:
– Realtime numerical
weather predictions.
– Air quality and fire
danger assessment
forecasts.
Hardware Base
• SGI Altix 3700, Linux OS
• 32 Itanium processors
each at 1.3 GHz
• 80 GB RAM
• 2x1.3 TB Raid
• 4TB Tape drive (~.5TB
monthly data archive)
• Intel Fortran compilers
Numerical Weather Predictions
• The Fifth Generation Penn
Sate/NCAR Mesoscale
Model (MM5) is used as the
realtime numerical model.
• It is a community model that
can be applied to realtime
and historical studies of a
large spectrum of weather
events: mainly mesoscale
convective systems, fronts,
land-sea breeze and
mountain-valley circulations.
Physics Options in MM5: cumulus
parameterization
•
•
•
•
•
•
•
1. None – grid sizes < 5-10 km.
2. Anthes-Kuo based on moisture convergence (larger grid sizes > 30 km).
3. Grell based on rate of destabilization or quasi-equilibrium, simple single-cloud scheme
(smaller grid sizes 10-30 km)
4. Arakawa-Schubert multi-cloud scheme similar to Grell scheme (larger scales > 30 km).
5. Fritsch-Chappell based on relaxation to a profile due to updraft, downdraft and subsidence region
properties (20-30 km scales)
6. Kain-Fritsch similar to Fritsch-Chappell, but using a sophisticated cloud-mixing scheme to
determine entrainment/detrainment.
7. Betts-Miller based on relaxation adjustment to a reference post-convective thermodynamic profile
over a given period (scales > 30 km).
Physics Options in MM5: PBL schemes
1. Bulk PBL - suitable for coarse vertical resolution in boundary layer, e.g. > 250 m
vertical grid sizes. Two stability regimes.
2. High-resolution Blackadar PBL suitable for high resolution PBL, four stability regimes, including free convective
mixed layer.
3. Burk-Thompson PBL predicts turbulence kinetic energy for use in vertical mixing, based on MellorYamada formulas.
4. Eta PBL the Mellor-Yamada scheme as used in the Eta model. It predicts TKE and has local
vertical mixing.
5. MRF PBL or Hong-Pan PBL, suitable for high-resolution in PBL (as for Blackadar scheme).
Efficient scheme based on Troen-Mahrt representation of countergradient term and
K profile in the well mixed PBL.
6. Gayno-Seaman PBL this is also based on Mellor-Yamada TKE prediction. It is distinguished from others
by the use of liquid-water potential temperature as a conserved variable.
Physics Options in MM5 (cont.)
• Explicit moisture schemes:
– 8 schemes including warm rain, simple ice,
and mixed phase (Dudia)
• Radiation schemes:
– None, simple cooling, surface
radiation,cloud radiation, CCM2, RRTM.
• Surface schemes:
– None, Blackadar schemes, Five-Layer Soil
model, Noah Land-Surface model, PhleimXiu Land-Surface Model
CANSAC MM5 Configuration
•
•
•
•
•
•
Coarsest domain 36 km, nested domain 12
km, and innermost (Nevada/California)
domain 4 km horizontal grid spacing and 32
sigma (vertical) levels.
Physics:
– PBL scheme: ETA
– Cumulus: Grell
– Moisture: Simple ice
– Radiation: cloud
– Soil: Five-Layer soil model
Initialized with ETA forecasts (40 km Grid212)
LDM Unidata observations:
– SYNOP
– AIREP
– METAR
– SHIPS
– PILOT
Initialized at 00Z and 12Z
Forecast length: 72 hr for domains 1 and 2,
and 60 hr for domain 3
CANSAC BlueSky System
Fire Characteristics
Area burned
Fuel moisture
Fuel loadings
Fire location
Meteorology
CANSAC MM5 outputs
Winds/Temps/Moisture
12 and 4 km domains
72 and 60 hour (-12 hr spin up) forecast
Fire ignition time
Emissions
EPM emissions model
Fuel consumption
Variable rate emissions
Smoke
Dispersion&Transport
CALPUFF/CALMET modeling system
PM2.5 concentrations
PM10, PM2.5,CO, CO2, CH4
Display
PAVE visualization package
NCL images (in progress)
Loops and hourly concentrations of PM2.5
Emissions
• Emissions are computed using
Consume/EPM v1.03 which calculates
the heat release rate and emissions for
particulate matter and carbon
compounds as a function of time since
fire ignition.
• Fuel characteristics input this model are
derived from the Fuel Characteristic
Classification System (FCCS).
CALPUFF/CALMET DISPERSION MODELLING
SYSTEM
•
CALPUFF is a multi-layer, multi-species non-steady state Lagrangian
puff dispersion model which can simulate the time and space varying
pollutant transport, transformation and removal.
•
It contains algorithms to account for
–
–
–
–
–
–
–
–
–
–
Dispersion
Plume rise
Building downwash
Partial plume penetration
Overwater and coastal dispersion
Complex terrain
Dry deposition
Chemical transformation
Wet removal
Odor and visibility modeling.
CALPUFF/CALMET
• CALMET is a diagnostic meteorological model which calculates
the three dimensional winds and temperatures along with
microphysical parameters such as surface characteristics,
dispersion parameters, and mixing heights to be used by
CALPUFF dispersion model.
• It can also incorporate the output of prognostic weather models
such as MM5. In this case, prognostic fields can be used as
initial fields, step 1, or the observations.
– BlueSky forecasts use MM5 outputs as the pure
observations/objective analysis in CALMET.
• Wind fields processing includes kinematic effects of terrain,
slope flows, blocking effects, divergence minimization.
CALPUFF/CALMET
• Dispersion
– Pollutant puffs travel along a Lagrangian trajectory and are
sampled at every receptor. Each puff is treated as Gausian
plume. Dispersion coefficients are calculated based on
observations, PG stability classes, or computed internally.
• Puff splitting
– Puff splitting is considered due to vertical wind shear.
• Puff stretching (horizontal convergence)
• Plume rise
– Plume buoyancy, atmospheric strafication, partial plume
penetration into a stable layer, area source plume rise
(parameters passed from EPM for fire sources), line source
plume rise (Briggs (1975) equations).
BlueSky Runs
• Currently the system
runs using
WILDFIRE209 reports.
• These reports are
received daily around
midnight.
• The model run starts
after the burn data
download.
• 60 and 48 hour
forecasts for 12 and 4
km domains,
respectively.
Forecast Verification
ETA model
Initial hour
surface
temperature
&
winds
MM5 surface
temperature
and winds
for the
corresponding
hour
• Temperatures can be uneresdtimated up to 5˚C.
• Implications: soundings, stability, PBL depth,
diagnostic parameters such as air quality and
fire danger indices, etc…..
Forecast Verification: case study 20
July 2005
• Extreme surface temperatures and deep atmospheric layers
during a period of 14-20 July 2005 (Lewis et al., 2005).
• Despite the increased temperatures, pollution concentrations
were low and diluted.
• Low Haines index predictions.
Forecast Verification: case study 20
July 2005
• ETA PBL
• RRTM
• Five-layer
LSM
• Simple-ice
• Gayno
Seaman PBL
• RRTM
• Five-layer
LSM
• Simple-ice
• MRF PBL
• RRTM
• Five-Layer
LSM
• Simple-ice
•
•
•
•
ETA PBL
RRTM
Noah LSM
Reisner 2
Forecast Verification: case study 20
July 2005
Future Products….
• Cross section plots
–
–
–
–
–
–
Redding
Sacramento
San Francisco
Los Angeles
Bakersfield
Fresno
Future Products (cont.)
• 3K ft - ground level temperature difference.
Future Products (cont.)
• Time-height plots at the current selected
CANSAC point locations (RAWS).
• Point verifications.
Future Products (cont.)
User Requests !!!!