Workpackage 3Role of aerosols vs. dynamics for different cloud systemsLeader: Philip Stier (University of Oxford, United Kingdom) |
WP3 will determine the key processes controlling cloud systems in contrasting environments and the relative role of natural vs. anthropogenic aerosol (precursor) emissions in each of them. It makes use of data and process understanding gained in WP1 and WP2, conducts case and process studies and evaluates and improves ESMs used for climate projections, aerosol radiative forcing and feedback studies in WP4. These objectives are split into 4 tasks:
This task will develop a joint framework to disentangle the effects of aerosol vs. dynamics for process and case studies in tasks 3.2 and 3.3. For each of the key regimes investigated in task 3.2, this includes: identification of background vs. anthropogenic conditions based on measurements from WP1; compilation of aerosol and aerosol precursor emissions, including a volatility basis set based organic aerosol precursor emission inventory; consistent implementation of parameterisations from WP2; provision of joint large scale forcing fields suitable for high resolution models and ESMs; coordination of suitable evaluation datasets of aerosol and cloud properties and their diagnostics in the models used in Task 3.2.
Arctic: High-resolution aerosol-cloud coupled model simulations of the
Arctic region will be performed to understand how aerosol and cloud
properties will change in a future climate with reduced sea ice coverage.
Simulations will be performed with the UM-UKCA, WRF and
UCLA-LES-SALSA. These simulations will identify processes
controlling the behaviour of aerosols and clouds in the present-day
summertime Arctic boundary layer, assess how
aerosols and clouds respond to reductions in sea ice and the climate
response of boreal forests and identify the roles of
biogenic sources, advected anthropogenic aerosol and local shipping
emissions for Arctic clouds and climate. This work will
be supported by the UK ACCACIA campaign with involvement of ULEEDS and
UMAN, and Met Office Large Eddy Model (LEM), providing access to in-situ
ground-based and aircraft data as well as to forcing datasets for the
high-resolution simulations.
Amazon: In the tropical environment we will
exploit the highly variable seasonal aerosol regimes to understand aerosol
invigoration of convection (e.g. Andreae et al., 2004; Rosenfeld et al.,
2008). We will use a range of models with different resolutions and
complexity of aerosol and cloud processes: The convective plume model
ATHAM,
the UCLA-LES, UCLA-LES-SALSA, WRF-Chem and
WRF-Chem-HAM with the same HAM aerosol microphysics as in MPI-ESM,
which will be coupled with the spectral bin microphysics scheme provided by
HUJI and schemes for co-condensation of SOAs and ice-nucleation. These case
studies will be used as a test-bed for a process-based evaluation of the
ESMs used for feedback studies in WP4. Process analysis (PA) modules in
ATHAM and WRF-Chem (Pöhlker et al., 2012) will be used to examine effects
of dynamical vs. aerosol processes. Data from WP1 will be used to
initialise the models WRF-Chem runs investigating the effect of different
aerosol sources on convection. The UCLA-LES will be run for very long
simulation periods using identical ECMWF forcing data with and without
time-varying aerosol (prescribed from the BACCHUS measurements) to ask
whether including aerosol variability in the LES improves the agreement
with observations, such as satellite derived droplet numbers, cloud top
pressures or the relationship between AOD and cloud properties.
Barbados: The frequent occurrence of shallow cumulus cloud potentially
susceptible to aerosol perturbations make this an ideal site to study
aerosol cloud interactions in a highly climate-relevant environment. The
proposed simulations will be complemented by long-term observations from
the Barbados observatory. We will set-up UCLA-LES and
UCLA-LES-SALSA over Barbados forced by ECMWF boundary conditions,
consistent with the global MPI-ESM simulations in Task 3.4 to investigate
whether including aerosol variability in the LES improves the agreement
with observations, such as satellite derived CDNC, cloud top pressures or
the relationship between AOD and cloud properties. Additionally we will
setup the MPI-ESM with the convective cloud field model using identical
ECMWF forcing data in single column mode over Barbados, evaluating the
simulated cloud spectrum with long-term observations from the Barbados
observatory and investigating potential aerosol effects on the cloud
spectrum.
This tasks combines traditional bottom-up CDNC/ICNC closure studies with a
novel concept of top-down closures studies:
Bottom-up: We will use in-situ
data provided by WP1 to conduct parcel model closure studies of cloud
microphysical parameters at the sites addressed in Task 3.2. We will
combine standard approaches with the detailed ACPIM cloud parcel model
and optimal parameter estimation methods based on Markov Chain Monte
Carlo (MCMC) parcel modelling (Partridge et al., 2012). In addition, we
will run MPI-ESM with the convective cloud field model (CCFM) that
simulates an ensemble of convective parcels that will be coupled with the
spectral bin microphysics scheme provided by HUJI and schemes for
co-condensation of SOAs and ice-nucleation, in single column mode nudged by
observed meteorology. This approach will constrain, evaluate and improve
ESM parameterisations.
Top down: Based on the vertical profile of
the dependence of the effective radius on cloud top temperature from
NPP/VIIRS imager (375 m) data retrieved in WP1 and temperature we will
retrieve the number of activated aerosol particles at cloud base, Na, in
non-precipitating convective clouds (HUJI, Freud et al., 2011). We will
combine Na with assumptions on cloud base updraft and knowledge of aerosol
hygroscopicity to infer CCN concentrations from satellites (Rosenfeld et
al., 2012). Similar relationships will be obtained for the development of
the mixed phase and glaciation temperature, Tg, of convective clouds
(following Rosenfeld et al., 2011, but with the greater accuracy of the
NPP/VIIRS). The combination of bottom-up and top-down approaches will
provide unique constraints on the satellite inferred cloud microphysical
properties and CCN concentrations as validation of products used in Task
3.4.
This task has two main objectives: to evaluate the ESMs used in WP4 through data retrieved and process studies conducted in Task 3.3 and to improve the ESMs to investigate aerosol cloud interactions on the global scale. We will set up MPI-ESM-HAM, MPI-ESM-CCFM, MPI-ESM-SALSA, HadGEM-UKCA and NorESM for a hindcast covering the BACCHUS observational period with the meteorology nudged to the same ECMWF data used as forcing data in the process studies in Task 3.2. The models will be evaluated with observations compiled in WP2 and Task 3.3. The role of natural vs. anthropogenic emissions will be quantified through simulations with specific emissions. On the observational side, FMI will compile satellite observations of aerosol and cloud properties over the BACCHUS focus regions over land and ocean surfaces, using improved algorithms for AATSR/SLSTR and MODIS from task 1.5 to provide information on the spatial variation of aerosol and cloud properties complementing local in-situ observations. This data will be used to detect aerosol signatures in cloud observations and for identification of the effect of different aerosol types (pollution vs. clean background, absorbing vs. non-absorbing aerosol) and studies in the twilight zone near cloud edges. The combination of global ESM results, global satellite datasets and compilations of global in-situ datasets will provide unique constraints on aerosol cloud interactions in the ESMs used in WP4.