; Constraining an atmospheric GCM by combining ground-based
Learning Center
Plans & pricing Sign in
Sign Out
Your Federal Quarterly Tax Payments are due April 15th Get Help Now >>

Constraining an atmospheric GCM by combining ground-based


  • pg 1
									 Constraining an atmospheric GCM by combining
  ground-based, satellite and reanalysis data for
improved estimation of the aerosol impact on the
              south-Asian monsoon

                   Dilip Ganguly
            NOAA-GFDL/Princeton University

            V. Ramaswamy and Paul Ginoux
• Diverse sources of aerosols: Natural and Anthropogenic

• Aerosols affect climate, mainly in two ways
 Direct: Scattering and absorbing the Solar & terrestrial radiation

 Indirect: Altering Microphysical        Aerosol ‘Hot spots’ in the globe
 properties and lifetime of clouds
                                           MODIS AOD at 500 nm for Sep 2000
                                                           Kaufman et al., 2002, Nature
• It still remains a challenge to
make realistic estimates of the
impact of aerosols on climate

• Distribution of aerosols around
the globe is heterogeneous and
they have a short residence time
 Two different pathways have been identified by which
aerosols could effect the South Asian Monsoon system
 The Solar dimming effect:

 [Ramanathan et al. 2005, Chung and Ramanathan, 2005.]
 Based on INDOEX measurements

 ABC (Atmospheric Brown Cloud) Forcing


Aerosol imposed cooling of land and northern Indian ocean cools
the surface, stabilizes the atmosphere, reduces evaporation,
causes a spin down of the Hadley circulation and weakens the
 Elevated heat pump effect:
                         Lau et al., [2006], Lau and Kim, [2006]
Anomalous heating due to BC in the foothills of Himalayas and
warming of upper troposphere over Tibetan plateau caused by
transport of dust, heats the atmosphere making it unstable with
increased convection and results in advancement and
intensification of Indian Monsoon

                                               distribution of
                                                aerosols (BC
                                               and dust) and
                                                solar heating
                                               across Tibetan
 Has the aerosol-monsoon-climate puzzle been solved ?
 Ramanathan et al. (2005), Lau et al. (2006), Randles and
 Ramaswamy (2008) and Meehl et al. (2008) have identified some of
 the possible mechanisms by which aerosols can effect the
 monsoon over South Asia

      But, there are caveats in each of these studies

• Indirect effects of aerosols have been ignored

• Aerosol distributions of Lau et al. (2006) and Meehl et al. (2008)
needs to be evaluated with observations over this region

• Magnitude and the distribution of ‘ABC forcing’ used by
Ramanathan et al. (2005) does not show good correlation with
the recent data available over land areas within this region
 A renewed attempt is being made at NOAA-GFDL to investigate the
impact of aerosols on the climate over South Asia.

 A realistic assessment of the impacts of aerosols on climate
demands GCMs to simulate accurate distribution of aerosols in the

 It is therefore important to evaluate model results globally and
understand the origin of any discrepancy in aerosol properties
simulated by the GCMs and observations.

 Evaluation of models are generally performed by comparing Aerosol
Optical Depth (AOD) with satellite data. However, this may not
sufficient as AODs are diagnosed from simulated mass concentration
by making further assumptions.
     Climatological AOD over the Indian subcontinent for January
                    AM2                      MODIS

Monthly     mean
AOD at 500 nm
for the year 2002
observations at
Kanpur and AM2
There is some problem with the GCM which can be due to many reasons
                            Among them…

      Error in the meteorological values simulated by the GCM

   One solution for this problem could be Nudging the winds and
              specific humidity with NCEP Reanalysis

                                               AM2 is underestimating
                                                  RH in the lower
Assimilation of NCEP-AM2 data with sounding data from Asia

         Before Assimilation                       After Assimilation

  But all these efforts did not result in any significant improvement in the
     simulated AOD to match with satellite data or other observations
 Comparison of AOD simulated by AM2 with observations


                               AERONET AOD
• There are differences between AM2 results and observations

• The problem could be due to various reasons
• How to get things right in the GCM ?
•Sulfate is overestimated and organics is underestimated by AM2 over
the North American region.                     Ginoux et al., JGR[2006]
How do we evaluate any model (GCM) over
regions where measurements of aerosol
composition and concentration are not available ?

                    Measurements of aerosol concentration over
We do not have :
                    large parts of the world

                   Long term and systematic measurements of
We do have :       optical and physical parameters of aerosols
                   across the globe
1.     Aerosol Robotic
     Network (AERONET)
Direct Product: AOD at 7 wavelengths
Inversion Products:
            Single scattering albedo
            Size distribution

2.   Micro-Pulse Lidar
     Network (MPLNET)

Extinction profile of aerosol at 523 nm

3.    Cloud-Aerosol Lidar
        with Orthogonal
     Polarization (CALIOP)
CALIOP onboard CALIPSO satellite

Extinction profile of aerosol at 523 nm
Develop a technique to derive the concentration of aerosols from optical
measurements in places where direct measurements of aerosol composition
are not available and use this information to constrain the GCM.

Finding an optimum combination of aerosol concentration which gives the
closest values of AOD, single scattering albedo and size distribution of
aerosols as AERONET when distributed according to the profile from
                            Ganguly, Ginoux and Ramaswamy et al. (2009), JGR
Methodology                 Ganguly, Ginoux and Ramaswamy et al. (2009), GRL

  Optical properties of individual components

                      mass growth           mass specific
    Dry mass            factor             extinction cross
  Construction of concentration profile

  Assumption made
 We assume log-normal distribution and external mixture of
following aerosol components: Sulfate, BC, OC, Dust (two modes), Sea Salt
                                       (two modes)
 Compute parameters like AOD, single scattering albedo (w),
volume fraction of coarse (Cvc) and fine (Cvf) mode aerosols
  Minimization process

                                                           Uncertainty / Error
                                                           associated with
                                                           each parameter

    where          and         are the computed and observed
    aerosol parameters and M (total number of parameters)=13 with
    contributions of 7, 4, 1, and 1 corresponding to AOD, w, Cvf and Cvc

    Finally, the aerosol model with the lowest        is selected.
Application of our technique over different parts of the world
   Kanpur, India        GSFC, United States   Etosha Pan, Namibia
Comparison of surface concentration of aerosol components
from the results of minimization with those simulated by the
  GFDL AM2 model and measurements available over USA

                        GSFC, 2002
Comparison of results over Southern Africa

             Mongu, Zambia
Using this technique, we are able to evaluate
the model results in a more consistent way

How do we use these results to constrain the

              Still remains a challenge
              but we made an attempt
 Impact of aerosols on present and future climate
    using a General Circulation Model (GCM)

   Correct distribution of aerosol concentration
    and optical depths simulated by the GCM

Production   Transport      Optical properties   Removal
                           hygroscopic growth

             Assuming these correct
                        Adjust these
  Constraining the model based on the results of our retrieval over North

              Before            AERONET AOD

Changes made in AM2 based on our results
• Decreased sulfate from fossil fuel by 50%
• Increased BC from biofuel by a factor of two
• Increased the biogenic emissions of OC over boreal forests
• Included OC emission from sea spray
• Increased the wet deposition of aerosols by increasing the scavenging ratio
We did some similar changes in the aerosol inventory over
the South Asian region based on the results of our retrieval

We are currently running an improved version of the coupled
atmosphere-mixed layer ocean model SM2.1 of NOAA-GFDL to
investigate the climate response to the direct as well as
indirect effects of anthropogenic verses natural aerosols from
South Aisa on monsoon dynamics.
Thank You for your attention

To top