An Introduction to
Physical Parameterization Techniques Used in
Atmospheric Models
J. J. Hack
National Center for Atmospheric Research
Boulder, Colorado USA
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Outline
Frame broader scientific problem
Hierarchy of atmospheric modeling strategies
Scale interaction problem
concept of resolvable and unresolvable scales of motion
Parameterization of physical processes
approaches rooted in budgets of conserved variables
Some science drivers
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Top of Atmosphere Radiation Component Fluxes
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Top of Atmosphere Net Radiation Budget and
Implied Meridional Energy Transport
Zhang and Rossow (1997)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Total Versus Ocean Energy Transport
Required
by ocean
Simulated total
Observationally derived total
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Principles of Atmospheric Modeling
Scientific basis for atmospheric simulation
rooted in laws of classical mechanics/thermodynamics
developed during 18th and 19th centuries (see Thompson, 1978)
early mathematical model described by Arrhenius (1896)
surface energy balance model
Two modeling approaches developed over last century
based on energy balance requirements
dynamical models (e.g., explicit transports)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Atmospheric modeling hierarchy
Understanding has been aided by a hierarchy of approaches
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Atmospheric modeling hierarchy
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Atmospheric General Circulation Models
and Climate Simulation
Reduced models of the climate system
apply averaging operator to governing equations
Atmospheric General Circulation Models (AGCMs)
simulate detailed weather fluctuations in the fluid system
day-to-day solution details are non-deterministic (Lorenz, 1962)
apply averaging operator to detailed solution sequence
utility lies in prediction of statistical properties of the fluid system
chronological sequence of intermediate states unimportant
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Example of State of the Art Global Model Simulation
Precipitable Water (gray scale) and Precipitation Rate (orange)
Animation courtesy of NCAR SCD Visualization and Enabling Technologies Section
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Modeling the Atmospheric General Circulation
Economic and social relevance
agriculture and food production
transportation
energy
water resources
policy and planning
Understanding of climate and global scale dynamics
atmospheric predictability/basic fluid dynamics
physics/dynamics of phase change
radiative transfer (aerosols, chemical constituents, etc.)
atmospheric chemistry (trace gas sources/sinks, acid rain, etc.)
interactions between the atmosphere and ocean (e.g., El Nino, etc.)
solar physics (solar-terrestrial interactions, solar dynamics, etc.)
impacts of anthropogenic and other biological activity
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Climate System Components
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Atmospheric Energy Transport
Synoptic-scale mechanisms
hurricanes
[Link]
extratropical storms
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Clouds are a fundamental component of largerscale organized energy transport mechanisms
[Link]
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Other Energy Budget Impacts From Clouds
[Link]
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Other Energy Budget Impacts From Clouds
[Link]
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Energy Budget Impacts of Atmospheric Aerosol
[Link]
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Energy Budget Impacts of Atmospheric Aerosol
[Link]
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Scales of Atmospheric Motions
Anthes et al. (1975)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Examples of Global Model Resolution
Typical Climate Application
Next Generation Climate
Applications
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Global Modeling and Horizontal Resolution
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Capturing Principle Phenomenological Scales
of Motion in Global Models
Simulation of Tropical Cyclone Impacts on Climate
Courtesy, Raymond Zehr, NOAA CIRA
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
High-Resolution Global Modeling
Simulation of Tropical Cyclone Impacts on Climate
~500 km
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
High-Resolution Global Modeling
Still a Need to Treat Subgrid-Scale Processes
Panama
T42 Grid
T85
T170
Grid
Galapagos
Islands
Satellite Image Courtesy, NASA GSFC Scientific Visualization Studio
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Meteorological Primitive Equations
Applicable to wide scale of motions; > 1hour, >100km
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Numerical Approximations
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Global Climate Model Physics
Terms F, Q, and Sq represent physical processes
Equations of motion, F
turbulent transport, generation, and dissipation of momentum
Thermodynamic energy equation, Q
convective-scale transport of heat
convective-scale sources/sinks of heat (phase change)
radiative sources/sinks of heat
Water vapor mass continuity equation
convective-scale transport of water substance
convective-scale water sources/sinks (phase change)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Physical Parameterization
To close the governing equations, it is necessary to incorporate
the effects of physical processes that occur on scales below the
numerical truncation limit
Physical parameterization
express unresolved physical processes in terms of resolved processes
generally empirical techniques
Examples of parameterized physics
dry and moist convection
cloud amount/cloud optical properties
radiative transfer
planetary boundary layer transports
surface energy exchanges
horizontal and vertical dissipation processes
...
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Cumulus Convection
If the atmosphere is buoyantly unstable to small vertical
displacements, it can be said to be convectively unstable
Convective overturning
with or without phase change
space scale ~ 1-10km; time scale ~ 1 hour
Moist convection
most common and energetically important
affects the general circulation on wide range of time scales
provides fundamental coupling of dynamics and hydrological cycle
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Diagnostic Budget Studies (heat & moisture)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Diagnostic Budget Studies (heat & moisture)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Diagnostic Budget Studies (heat & moisture)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Diagnostic Budget Studies (heat & moisture)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Diagnostic Budget Studies (heat & moisture)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Diagnostic Budget Studies (heat & moisture)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Cumulus Parameterization Example A
No Process Model
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Process Models and Parameterization
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Parameterization of Cumulus Convection
To extract the details of how the observed Q1 and Q2 forcing
is maintained by moist convection, it is necessary to use an
abstraction for the collective behavior of convective motions
Convective mass flux
how much overturning is associated with convective activity
Breakdown of total diabatic forcing
where is the water condensing and/or raining out
what role do the convective eddy transports play
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Parameterization of Cumulus Convection
Assumptions associated with the introduction of a cloud model
Cumulus clouds in statistical equilibrium with large-scale
exploit idealized characterization for cloud elements
mass, energy, and water budgets can be derived
Dry Mass
Energy
Water Mass
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Parameterization of Cumulus Convection
Closure assumption: Incorporate process regulating convection
Neutral or specified thermodynamic stratification
moist convective adjustment
Manabe et al. (1965), Kurihara (1973)
thermodynamic adjustment schemes
Betts (1986), Emanuel (1991)
Moisture convergence
Kuo (1965; 1974), Anthes (1977)
Simple bouyancy closures
Gregory and Roundtree (1990), Kreitzberg-Perkey (1976),
Quasi-equilibrium
Arakawa and Schubert (1974)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Cumulus Parameterization Example B
Hack et al. (1984)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Does Process Formulation Matter?
Changes to the parameterization of deep convection
incorporation of precipitation evaporation mechanism
Simulated response to evaporation of falling condensate
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Diabatic Forcing of the Deep Tropics
Longwave Radiation
Moist Convection
Shortwave
Radiation
CAM2
Boundary
Layer
Total diabatic forcing (sum of components) is almost identical
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Heat and Moisture Budget Equations
But, the moisture budget is closely coupled to the total diabatic heating!
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Diabatic Forcing of the Deep Tropics
Convection
Radiation
TOTAL
CAM2
Boundary
Layer
CAM2
Convective component change results in significant moisture budget change
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Examples of Science Drivers
Climate Sensitivity: the final frontier
what is the real climate sensitivity to anthropogenic forcing?
Clouds!!
Separating the signal from the noise
non-deterministic system with large natural variability
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
IPCC 1995: Climate Model Projections
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
CMIP 2001: Temperature and Precipitation
Covey et al. (2001)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Some Fundamental Physical Sources of Uncertainty
IPCC Working Group I (2001)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Parameterization of Clouds
Another major source of model projection uncertainty
Cess et al. (1990)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Parameterization of Clouds
Cloud amount (fraction) as simulated by 25 atmospheric GCMs
Weare and Mokhov (1995)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Parameterization of Clouds
Seasonal cycle of cloud fraction as simulated by 25 AGCMs
Weare and Mokhov (1995)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Evaluating parameterization/simulation quality
Continue to compare long term mean climatology
average mass, energy, and momentum balances
tells you where the physical approximations take you
but you dont necessarily know how you get there!
Must also consider dominant modes of variability
provides the opportunity to evaluate climate sensitivity
response of the climate system to a specific forcing factor
evaluate modeled response on a hierarchy of time scales
exploit natural forcing factors to test model response
diurnal and seasonal cycles, El Nio Southern Oscillation, intraseasonal
variability, solar variability, volcanic aerosol loading
High-resolution, limited area, process modeling
explicitly resolve additional scales of motion (several more orders of magnitude)
proper definition of experimental framework of central importance
solutions appear to be strongly constrained by boundary conditions
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
El Nio Composite
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Testing AGCM Sensitivity
Pacific SST Anomalies and ENSO
Hack (1998)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Testing AGCM Sensitivity
OLR Anomalies and ENSO
Hack (1998)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Improving simulation quality
Examine role of parameterization techniques on transient behavior
oversimplifications playing a role in inadequate variability?
Inadequate process formulation; deficiencies in closure assumptions, ?
Understand role of scale interaction on transient and mean state
ITCZ behavior as function of horizontal resolution
Williamson, Kiehl, and Hack (1995)
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
Summary
Global Climate Modeling
complex and evolving scientific problem
parameterization of physical processes pacing progress
observational limitations pacing process understanding
Parameterization of physical processes
opportunities to explore alternative formulations
exploit higher-order statistical relationships in physical process formulations?
stochastic strategies
exploration of scale interactions using modeling and observation
high-resolution process modeling to supplement observations
e.g., identify optimal truncation strategies for capturing major scale interactions
provide better characterization of statistical relationships between resolved
and unresolved scales
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003
The End
J. J. Hack/SAMSI Multiscale Workshop 2 February 2003