Transcript 支配方程式
A Next-Generation Atmospheric General
Circulation Modeling
Frontier Research Center for Global Change
Hirofumi TOMITA
Masaki SATOH
Tomoe NASUNO
Shi-ichi IGA
Hiroaki MIURA
Contents
Motivation of our new modeling
Global cloud resolving model
•
To avoid the ambiguity of cumulus parameterization
Model description
Quasi-uniform grid in the horizontal direction
•
Icosahedral grid
Nonhydrostatic framework
•
Suitable for climate simulation
Aqua Planet Experiment
The first attempt of global cloud resolving in a long term
Summary and Future plan
Motivation (1)
General problem for current AGCMs
Cumulus parameterization
•
•
One of ambiguous factors
Statistical closure of cumulus convections
Future AGCM
Explicit treatment of each cloud
•
•
Cumulus parameterization
Large scale condensation scheme : not used!
Cloud microphysics : used!
Explicit treatment of multi-scale interactions
•
Each cloud scale meso-scale planetary scale
Global Cloud Resolving Model
Motivation (2)
Target resolutions
5 km or less in the horizontal direction
Several 100 m in the vertical
Strategy of dycore development
Quasi-uniform grid
•
Spectral method :
not efficient in high resolution simulations.
– Legendre transformation
– Massive data transfer between computer nodes
•
Latitude-longitude grid :
the pole problem.
– Severe limitation of time interval by the CFL condition.
•
The icosahedral grid:
homogeneous grid over the sphere
– To avoid the pole problem.
Non-hydrostatic equations system
•
Very high resolution in horizontal direction.
Current Status of Our Model
Model name : NICAM(Nonhydrostatic Icosahedral Atmospheric Model)
Model feature
Governing equations
Full compressible non-hydrostatic system
including acoustic wave
Spatial discretization
Horizontal grid configuration
Vertical grid configuration
Topography
Finite Volume Method
Icosahedral grid
Lorenz grid
Terrain-following coordinate
Conservation
Total mass, total energy
Temporal scheme
Slow mode - explicit scheme (RK2)
Fast mode - Horizontal Explicit Vertical Implicit
scheme ( HEVI )
Physical parameterization
Almost completed ( turbulence, radiation, cloud
physics, surface flux )
Computational tuning
Vectorization
Well tuned for NEC SX6 architecture
Parallelization
2D decompostion,
Flexible configuration against load imbalance
Target machine
WS-cluster, Linux-cluster, Earth Simulator
Grid Generation Method
(0) grid division level 0
(1) grid division level 1
Grid generation
(2) grid division level 2
(3) grid division level 3
1. Start from the spherical
icosahedron.
(glevel-0)
2. Connection of the midpoints of the geodesic arc
4 sub-triangle
(glevel-1)
3. Iteration of this process
A finer grid structure
(glevel-n)
# of gridpoints
11 interations are requried
to obtain the 3.5km grid
interval.
Aqua-Planet Experiment
Past reseaches
Hayashi & Sumi (1986), Swinbank et al.(1988)
•
Behaivior of MJO etc.
Gotswami et al.(1984), Numaguchi(1995)
•
Formation and intensity of Hadley circulation
APE as a standard test case
Neale & Hoskins(2001)
•
AMIP-like model intercomparison
experimental setup
•
•
•
Fixed zonal-symetric SST
Prescribed Ozone distribution
Equinoctial solar radiation
investigate the dependency of cumulus param.
on the results
Our approach:
to perform the APE by a cloud resolving model
resolution (15km~3.5km)
ONE REFERENCE RESULTS against other parameterization models
Series of experiment by NICAM
Analized term
Spin-up time NICAM
0 day
60 day
14km grid
model
90 day
30days
Interpolation
7km grid
model
30days
Interpolation
3.5km grid
model
10days
Initial condition: appropriate climatology
of a conventional GCM
( CCSR/NIES/FRCGC AGCM ver 5.7)
OLR(1S-1S平均)
Precipitation rate [mm/day] at day 85 : log-scale
by NICAM-3.5km model
Super cloud cluster
Mid-latitude cyclone
OLR (7km-model) during 60-90 day
A typical Super Cloud Cluster
Super cloud cluster : ~1000km
Westerly wind burst
Cloud cluster :~100km
Convectively-Coupled Kelvin Wave
High pressure
Low pressure
Hovmoller diagrams of OLR ( 2S-2N )
NICAM-14km
NICAM-7km
Westward moving of CC
Lifetime of 2days
Eastward propagation of SCC
NICAM-14km:
20~25 days
fast propagation
NICAM-7km, 3.5km :
25-40 days
corresponding to MJO
also well organized
rather than NICAM-14km.
NICAM-3.5km
Histograms of diurnal cycle for precipitation
Peak : midnight
Peak : early morning
LT [hr]
Consistent with the obs.
in open ocean
LT [hr]
Summary
We have developed a global CRM in order to avoid
the ambiguity on the cumulus parameterization.
Nonhydrostatic system
Icosahedral grid
As the first attempt of GCRM, we performed an
Aqua-Planet-Experiment.
Hierarchical structure of cloud convections
MJO-like signal with realistic phase speed
Diurnal cycle of precipitation
We confirm that the GCRM approach becomes the
one of major approach in the climate research field
in the near future.
The Global Cloud Resolving Climate Simulation is not a dream!
Example of stretched grid
Default grid : glevel-6
120km grid intv.
•
Homogenious
Stretched grid
After the transformation
•
Grid interval :
–
120km 12km
Reduction of earth radius : 1/10
1.2km grid interval