Upper tropospheric clouds play important roles in transporting water vapor
into the stratosphere and regulating thermodynamical processes inside the
troposphere. The amount of cloud ice in the upper troposphere is
a key component of hydrological cycle in the Earth atmosphere.
The amount of ice mass at these levels depends on strength,
frequency and size of deep convection, but remains poorly quantified, especially
on a global basis. Formation and removal processes of cloud ice are
complex, involving aerosol-cloud interactions and other cloud
microphysics. Treatment of these clouds in global circulation
models (GCMs) remains relatively primitive and oversimplied in many
cases. Deep convection and cloud microphysics parameterizations are used
as tunable variables in a sense to balance the top and bottom of atmospheric
radiation [Slingo and Slingo, 1988; Randall et al., 1989;
DelGenio et al., 1996; Ramanathan et al., 2001; Graf, 2004].
Observations of cloud ice amount and variability in the upper troposphere
provide guidance in improving these models towards more realistic physics.
It is challenging to measure cloud ice with remote sensing means from space. Most passive nadir-viewing techniques do not have sufficient vertical resolution to resolve cloud profile at tropopause heights. Limitations include lacks of sensitivity of all ice particle sizes, inability of penetrating thick clouds, contamination of surface radiation, and field-of-view smearing, etc. Satellite infrared (IR) and visible techniques are sensitive to high cirrus clouds but limited to small (<30 microns or ice water content < 0.01g/m3) ice particles on the uppermost cloud layers. These observations cannot provide ice mass measurement for thick-and-dense clouds like those from deep convection. On the other hand, low-frequency microwave techniques are able to penetrate most clouds and interacts only with very large ice particles (ice water contents of ~0.1g/m3) and rain droplets, but field-of-view smearing, surface radiation and mixed-phase clouds often complicate observations. Scattering-based algorithms have been developed and used to deduce cloud ice water path (IWP) from various nadir-viewing sensors with the best success over ocean surfaces [e.g., Liu and Curry, 2000; Zhao and Weng, 2002]. However, ice concentrations near the tropopause are often too low for the nadir-viewing microwave techniques to detect.
Narrow beamwidth limb techniques at high microwave frequencies (200-600GHz), such as MLS flawn UARS and Aura, offer additional opportunity to measure cloud ice in the upper troposphere Wu et al. [2006]. Unlike nadir-viewing sensors, limb techniques do not rely on cloud top temperature for cloud detection. Nevertheless, the high-frequency microwave observations can penetrate to most clouds in the upper troposphere, and are affected little by cloud inhomogeneity, low clouds and surface radiation. Initial results show that Aura MLS IWC has a similar morphology to those produced by global circulation models, where some model-model comparisons exhibit larger differences than MLS-model differences Li et al. [2005]. The observational constraint from MLS cloud ice measurements has led to significant improvements in performation of some models (e.g., ECMWF) through revised parameterization schemes.
Because of large spatial and temporal variabilities, clouds are often undersampled by ground-based and spaceborne instruments. The sparse and incoherent observations make it difficult to reliably characterize clouds and their micro/macro-physical properties. Most of the observations are limited by constraints of instrument sensitivity and regional atmospheric conditions, and can hardly compare to observations in other regions and by other techniques. As a result, the parameterization schemes used in numerical models often perform biasly towards one or other kind of observations. This situation must be changed!
As part of NASA's A-Train synergic observing system,
the Aura MLS (launched in 2004) is flying in formation with Aqua (launched in 2002), and
CloudSat 94-GHz Cloud Profiling Radar (CPR) (launched in 2006) [Stephens
et al. 2004] with coincident measurements < 15 and < 7 min,
respectively. Aqua AIRS (Atmospheric Infrared Sounder) [Aumann
et al., 2003] and MODIS (Moderate Resolution Imaging
Spectroradiometer) [Platnick et al., 2003]
provide high-resolution horizontal coverage at visible
and infrared frequencies whereas the CloudSat CPR makes high
resolution measurements of cloud ice profiles. During the first two-year mission,
CloudSat/CALIPSO footprints were not collocated with Aura MLS. Since May 8,
2008, a new maneuver from Aura aligns up the MLS tangent point within the same
(+/-10 km) sampling plane of CloudSat/CALIPSO curtains. The MLS submm-wave
receivers will provide the key sensitivity to link the radar and lidar
cloud measurements in the upper troposphere. Together, these A-Train observations yield unprecedented coverages on cloud
properties and variabilities, which vary by more than four orders of
magnitude and are coupled with processes caught between
chaos and order.
Importance of Particle Size Distribution
The Aura MLS monthly data (V2.2) in NetCDF or IDL-save formats can be downloaded from ftp://mls.jpl.nasa.gov/pub/outgoing/dwu/cloud/emls/monthly/
