Climate Models: Past to Present

1
Climate Models Past to Present

• V.Vidyunmala
• Ph.D Student
• Students Seminar Series

2
Plan of Talk

• Climate System
• Scales of Climate System
• Evolution of models
• Comparison of AMIP-CMIP results
• Model Sensitivities

3
Climate System
Climate system consists of atmosphere,
hydrosphere, cryosphere, lithosphere and biosphere
4
Feedbacks mechanisms in Climate
Climate processes are mainly forced by external
and internal forcing mechanism
External Forcing Solar
forcing Internal Forcing Determined
by the existence of feedbacks

• Positive Feedback
• Greenhouse effect of water vapour and atmospheric
  temperature
• Albedo of snow and ice cover and atmospheric
  temperature
• Greenhouse effect of CO2 and the surface air
  temperature

• Negative Feedback
• Equator-to-pole temperature difference and the
  meridional heat transport
• Soil humidity and albedo of land surface
• Air temperature and cloudiness

5
Scales of temporal Variability of the Climate
System
Climate system oscillation are found to have
large temporal variability

• Small scale oscillations fraction of
  second to several minutes
• Meso-scale oscillations several minutes
  to several hours
• Synoptic scale oscillations several hours to
  several days in atmosphere and weeks in
  ocean
• Global variations weeks to
  months
• Seasonal Variations annual periods
  
• Inter-annual Variations periods of
  several years
• Intra-centennial variations periods of tens of
  years
• Inter-centennial Variations periods of several
  centuries
• Long-Period Oscillations periods of tens of
  thousands of years

6
Equilibration Time of the Climate System
Equilibration Time is the response time (or)
relaxation time of the components of the climate
system
7
What are models ? And why do we use them?
A climate model is a mathematical representation
of the physical processes that determine climate
To gain quantitative insights into the behavior
of the earth system.
Models are natural extensions of theory
Theory Analytic solutions Models
Numerical Solutions
8
Are these models Complex or Simple
Box (1976) .all models are wrong some are
useful. Accepting this principle, the job is not
so much in search of the true model but to select
a model that is appropriate for the problem in
hand.
A useful model is not the one which is true but
the one that is informative Feldstein, 1982
Models are simplifications of the complexity of
the nature
Lastly, Models should be as simple as possible
but not simpler - Einstein
9
Components of Climate Model
The main components of a climate model are
Radiation Dynamics
Surface Processes
The way in which the input and absorption of
solar radiation and the emission of infrared
radiation are handled The movement of energy
around the globe (from low latitude to high
latitude) and vertical movements(
convection) Inclusion of land/ocean/ice and the
resultant change in albedo, emissivity, and
surface-atmosphere exchanges.
10
Types of Climate models
Energy balance models (EBMs) are one-dimensional
models predicting the variation of surface
temperature with latitude.
One-dimensional Radiative-Convective models (RCs)
they compute the vertical temperature profile by
explicit modelling of the radiative processes and
a convective-adjustment which reestablishes a
predetermined lapse rate
Two-dimensional Statistical dynamical models
(2D-SDs), they deal with surface processes and
dynamics in a zonally averaged framework and have
a vertically resolved atmosphere
General Circulation Models (GCMs), these are
three dimensional nature models of
atmosphere/ocean with all the physics and
dynamics included
11
Historical Evolution of Climate Models
12
Energy Balance Models
Simple Energy Balance Model If we consider each
latitude zone independently
Disadvantages of 1-D EBM
They are sensitive to changes in solar constant
and albedo.
13

• There are two other types of EBMs
• A simple box model of the ocean-atmosphere In
  this it is a system of four components, two
  atmosphere over ocean and land and a oceanic
  mixed layer and deeper different ocean.

The heating of the mixed layer is calculated by
assuming a constant depth in which the
temperature difference changes due to thermal
forcing, atmospheric feedback and leakage of
energy permitted into underlying water.
2. A coupled atmosphere,land and ocean energy
balance box model It includes the polar sinking
of oceanic water into deep oceans.
Disadvantages Hemispherically averaged cloud
fraction is prescribed as seasonal varying
feature. We cannot incorporate temperature of
surface albedo feedback since land is
hemispherically averaged
14
Radiative-Convective Models
They resolve many layers in the atmosphere and
seek to compute the atmospheric and surface
temperature.

• Three main assumptions are made here
• There is no reflection of upward traveling
  short-wave radiation by cloud.
• The surface emissivity has been set equal to
  unity.
• Cloud/dust absorption in infrared wavelength is
  equal to epsilon.

Model sensitivity

• Radiation field determines the temperature when
  saturation does not occur, otherwise moist
  adiabatic lapse rate is used.
• Cloud coverage is fixed to 50, and there are
  separate calculations for clear and cloudy sky.
  The radiation fluxes are then weighted by cloud
  cover to yield final temperature

15
Two-Dimensional Statistical Dynamical Models
General circulation in this case is assumed to be
composed mainly of cellular flow between
latitudes which is characterized by using
empirical and theoretical formulations a set
of statistics summarizes wind speed and
direction an eddy diffusion coefficient is
used which govern EBM transport.

• Difference between this and GCM is that all the
  variables of interest are zonally averaged.
• Advantages of 2-D SD models
• Detecting the signal produced by small changes in
  case of GCM will take many years of climate
  simulation.
• Slight disturbances in climatic state do not
  negate the assumption inherent in the formulation
  of two-dimensional models.
• Model sensitivity
• Instead of using a simple convective adjustment
  scheme as RCs, a more complex empirically derived
  scheme for onset of convection and cloud process
  is required.

16
General Circulation Models
Primitive equations are used in Atmospheric and
Oceanic GCMs The following processes are
represented

• Conservation of mass
• Conservation of momentum
• Conservation of energy
• Conservation of water vapor
• Equation of state

17
Parameterization
The representation of the sub-grid scale
phenomena as functions of variables that are
represented on the model grid.
What processes are parameterized?

• Atmospheric radiative transfer (short-wave and
  long-wave radiation).
• Moist convective processes/ Mesoscale convection
  of eddies in ocean.
• Planetary boundary layer/ Mixed layer in oceans.
• Cloud formation and radiative interactions
• Mechanical dissipation of kinetic energy/eddy
  resolving models.

18
Initial and Boundary Conditions for Atmospheric/
Oceanic GCMs

• Oceanic GCM
• The thermodynamic sea-ice model is an integral
  part of the OGCM in some cases.
• Initial condition given to the oceanic GCMs are
  surface temperature, sea-ice extent, surface
  albedo over ice-covered and ice-free regions,
  sea-surface salinity, partial pressure of CO2,
  wind stress at surface and fluxes at surface
  etc.,
• The topography of oceanic basins is very
  important for getting the coastal circulation
  properly.

• Atmospheric GCM
• Land surface models are treated as integral
  components of the atmospheric model.
• Initial condition given to drive the atmospheric
  GCMs are winds, temperature profile, specific
  humidity, orography, radiative fluxes at surface
  and top of the atmosphere, land-sea mask,
  hydrological parameters, albedo, snow-ice extent
  etc.,
• Boundary forcing given to the atmospheric models
  are through SST( Sea surface temperature).

19
Coupled Ocean Atmosphere Models
Hierarchy of Coupled ocean atmosphere models
20
Spin up and Flux Adjustments
Time scales for atmosphere are order of weeks,
but those for land surface and upper ocean extend
to seasons, while those for deep oceans it takes
thousands of years.
The experimental strategy is to spin-up
separately the component atmosphere and ocean
models before coupling.
21
Flux Adjustment

• Surface heat flux is modified for ocean model by
  the addition of heat flux parameter that
  depends on the mean calculated by the separate
  ocean and atmosphere models in the spin-up phase.
• Similarly, adjustments can be applied to fresh
  water flux, surface wind stress and surface
  temperature as seen by the GCM.

22
Comparison of AMIP-CMIP Results
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
Model Sensitivities

1. Sensitivity to representation of water vapor
   Distribution of water vapor is characterized by
   large values near surface and low values in upper
   troposphere. This leads to overshoots and
   undershoots.
2. Sensitivity to Model resolution Strong
   dependence of a models parameterization on
   resolution makes it difficult to separate the
   purely dynamical and physical effects on
   resolution changes. Ocean models are found to be
   more sensitive to horizontal resolution in case
   of coupled models.
3. Sensitivity to Convection and Clouds
   Mitchell(89), Boer(93), Meehl (96) have shown the
   sensitivity of choice of cumulus parameterization
   scheme. Hence the consensus on which is better is
   yet not known.
4. Sensitivity to Land-Surface Processes
   Thomson(96) showed that Land-Surface Transfer
   scheme produces smaller variations of soil
   moisture in greenhouse gases simulation than in
   bucket model.
5. Sensitivity to initial and boundary conditions.

28
References

• McGuffie and Henderson-Sellers , 2001 FORTY
  years of Numerical Climate Modelling, Int. J.
  Climatol, 21, 1067-1109.
• Houghton, Callander, Harris and Maskell, 1996
  The Science of Climate Change, The Second
  Assessment Report of IPCC.
• Ocean-atmosphere interaction and Climate
  modelling Boris A. Kagan, Cambridge Atmospheric
  and Space Science Series, 1995
• Climate System Modeling Kevin E. Trenberth,
  Cambridge University Press, 1992.

29
Thank you..