What Do Climate Models Require from Aerosol Researchers: My Views

1
What Do Climate Models Require from Aerosol
Researchers My Views

• Warren M. Washington
• National Center for Atmospheric Research
• Boulder, Colorado
• October 2005

2
Outline of Talk

• My perspective will be on the 20th and 21st
  century climate change.
• What is in the Community Climate System Model
  (CCSM) with respect treatment of aerosols?
• How can the aerosol community help make climate
  model aerosol effects better?

3
Timeline of Climate Model Development
4
Horizontal Grid Resolutions
(300 km)
(150 km)
(75) km
(37 km)
Next IPCC
5
Chemistry Interactions in CCSM
From Bill Collins
6
Shortwave Formulation in Standard CCSM

• ?-Eddington, solar spectrum divided in 19
  spectral intervals, 1 visible, 7 H2O, 3 CO2, 1
  near infrared, 7 in UV for ozone heating (Collins
  et al.,1998)
• Molecular scattering, scattering and absorption
  by cloud droplets and aerosol included
• Aerosol species-3D time dependent distribution
  sea salt, soil dust, carbonaceous, sulfate,
  volcanic sulfuric acid
• Method allows for both direct and indirect
  effects. Most simulations have indirect turned
  off. e.g. IPCC simulationswhyuncertainties!

7
More

• Three optical properties for each aerosol type
  specific extinction, single scattering albedo,
  and an asymmetry parameter.
• Hygroscopic growth included.
• More details are included in http//www.ccsm.ucar.
  edu/models/atm-cam/docs/description/

8
Some Numerical Details with Respect to Aerosol
Transport

• Finite volume transport approach which is volume
  conservative and positive definitenot trouble
  free e.g. requires polar smoothing, may be too
  diffusive. Key is transport the plumes more
  accurately both vertically and horizontally.
  Connection to ARM.
• Note the difference between conservation and
  transport accuracy! This presents a serious
  problem because aerosol properties evolve in the
  process of transport.

9
Title slide
Mt Pinatubo eruption in the Philippines, June 15,
1991. Gases and solids injected 20 km into the
stratosphere.
From Church, White, Arblaster
10
(No Transcript)
11
Krakatau (1883)
Sometimes referred to as Krakatoa
12
PCM 2XCO2 2.1C TCR
1.32C CCSM3 2XCO2 2.7C TCR 1.46C
13
Sulfate (direct effect)
14
Change of Extremes

• Heat waves, cold snaps
• Floods, droughts
• First freeze dates, hard freeze frequency
• Precipitation intensity
• Diurnal temperature

Note all of these are sensitive to regional
aspects of aerosol distribution
15
CCSM Climate Change Working Group Plans, 2005-2006
Further analysis of the 20th century and future
climate change simulations from CCSM3, as well as
the IPCC multi-model dataset (e.g. climate
variability and change, vertical structure of
temperature change, detection/attribution
studies) Analyses of climate sensitivity Large
ensembles and climate change signals (collaborate
with CVWG) Address role of aerosols in 20th
century and future climate change (e.g. sulfate
direct and indirect, dust, black carbon,
etc.) Determine respective roles of natural
forcing ( solar and volcanic) and anthropogenic
forcing (changes in GHGs, aerosols also land
use/land cover change (LULCC)collaborate with
LMWG)
16
2006-2008 Quantify possible sea level rise
scenarios based on hypotheses of ice sheet
de-stabilization (include dynamic ice sheet
model) Perform targeted climate change
mitigation/adaptation simulations with CCSM3 with
coupled carbon cycle (collaborate with BGCWG and
LMWG) initial use of WG3 scenarios to precede
fully coupled carbon cycle
17
More on CCSM Radiation Physics

• Cloud geometrical overlap
• New water vapor treatment
• Prognostic sulfur cycle for prediction of sulfate
  aerosolsthis is in reasonably good shape.
• Prescribed distribution of sulfate, soil dust,
  carbonaceous species, sea salt for direct effects
  on radiative fluxes and heating rates
• Volcanic aerosols
• Indirect effects of aerosols on cloud albedo and
  cloud lifetime are not yet included in CAM3.

18
What Do We Need From the Aerosol Community?

• Better observational based parameterizations of
  optical depths of specific aerosols. (Is what we
  are using the best?)
• MODIS and other satellite system are valuable
  instruments but it has problems with
  uncertainties in several areas contamination
  clouds and over various surfaces, accuracy
  degrades with high and low optical depths, need
  for more in situ validation, sun glint, etc. The
  data we can use is mostly totalsbut we are happy
  with it.
• How should we deal with various size
  distributions? i.e. is the log normal good
  enough? Or should we go with, say, 10 size ranges
  in a grid cell?
• Do we have enough solar bands?
• Methods for incorporating indirect effects.

19
More

• Uncertainties in emissions, especially carbon
  aerosols-anthropogenic, biomass burninglighting
  caused forest fires. Models tend give low
  concentrationscompared to observationsreason
  not known. (Rasch,2005)
• Removal mechanisms are 20 years oldare there new
  parameterizations coming out future research?
  (Rasch, 2005)

20
More

• If possible, we need more computationally
  efficient parameterizations (i.e. stop using
  those exponents and power approximations when
  table lookup or series approximations will do).
  However, real computational cost is the transport
  of aerosols and associated chemistry.

21
Optical Depths base on aerosol chemistry
transport model
Industrial sulfur emissions
Sulfate
Mostly near the Sahara
Mineral Dust
Near strong wind areas
Sea Salt
Biomass burning
Organic and Black Carbon
From Bill Collins et al
22
Clear Sky Only (W/m2)
Top of Atmosphere
Surface
Atmosphere
23
What types of aerosol observations do we need?

• Surface, aircraft, and satellite (What is DOEs
  role? Is it to integrate the various sources?)
• MODIS clear sky observations TOA forcing and
  other details on amount and type of aerosols
• Use satellite spectral dependence of cloud
  reflectivity to obtain effective radius of clouds
  dropletstie to indirect effect.

24
Are we heading down a slippery road of increased
complexity?

• Computer capability will increase by factors of
  five every two years or so. This limits climate
  model and BGC complexity!
• How many species should be keep track of in a
  climate model?
• What does DOE want special emphasis on? e.g.
  carbon cycle, effects of carbon and sulfate
  aerosols from fossil fuel use, etc. To sort out
  role of carbon we need all major aerosol types.

25
The Next IPCC Simulations (1870-2100)

• Some moderate resolution climate model
  simulations with fully interactive BGC model,
  with emphasis on carbon cycle and aerosols. Note
  this should be DOE priority.
• Higher resolution (T170) simulations with
  specified or transport BGC but no feedback on
  climate system
• Scenarios need improved treatment of aerosols for
  20th century and prediction of future emissions
  or concentrations.

26
The End