Report on Overseas Visit
Faull AGU Meeting, San Francisco
December 15 - 19 1996
R M Mitchell
CSIRO Division of Atmospheric Research
1. SATELLITE RADIOMETRIC CALIBRATION
I attended this meeting at the
invitation of Dr C.R.N. Rao of NOAA/NESDIS to contribute a paper
to the special session entitled `Calibration and Characterization
of Meteorological Satellite Sensors'. The session began with an
overview of current calibration practice for the AVHRR given by
Dr Rao. Current in-flight calibration updates are based on
the instrument response over a target in the South-eastern Libyan
desert, which experience has shown to be radiometrically stable.
The site has not been independently characterized. From November
1996, calibration updates are being released electronically on
a monthly basis, a timely development considering the plan to
establish an Australian calibration data base supported by the
EOC (see below).
My presentation, entitled `Identification and Characterisation
of a Bright Calibration Site in the Strzelecki Desert, South Australia',
began with a brief overview of Australian calibration and validation
work, followed by a detailed discussion of the bright target identified
at Tinga Tingana in the Strzelecki Desert. This site was selected
with the aim of fulfilling a variety of needs, including cross-calibration
of different satellite sensors, validation of satellite aerosol
retrievals and verification of atmospheric correction methods.
Because of this, much emphasis was placed on site characterization,
both in terms of surface angular reflectance pattern and atmospheric
properties such as water vapour and aerosol loading. Preliminary
characterization carried out during the September 1996 field trip
was reported. In the context of the session, this approach contrasted
favourably with the status of the African desert sites, for which
little in situ characterization is available.
Another presentation of particular interest was given by Dr Chris
Mutlow of RAL regarding visible and near-infrared calibration
of the ACER/2 sensor. This instrument is equipped with an on-board
calibration system which has proved effective in tracking changes
in radiometric response. However, surface targets are still important
for interchannel calibration and for consistency checks. It was
claimed that absolute radiometric accuracy approaching 1% can
be achieved, a significant improvement over vicarious calibration
methods (5%). The ATSR/2 calibration will be used to further improve
AVHRR calibration by analysis of observation over specified targets.
However, for this cross-calibration to be successful, accurate
site characterization is essential; hence the interest in our
work in this area.
Additional relevant work was presented by Dr W. Wang of the National
Satellite Meteorological Center in Beijing, currently visiting
scientist at NOAA/NESDIS. Dr Wang and colleagues are establishing
a calibration site in the Gobi Desert. The site is of intermediate
reflectance ( 20%), in contrast to the brighter sites such as
the Libyan desert ( 40%) or Tinga Tingana ( 35%). A range of target
reflectances is important in characterizing the response function
of the newer AVHRR/3 instrument, the first of which is due for
launch on NOAA K in August 1997. By contrast to the AVHRR/2, the
visible and near-infrared channels of the new instrument have
bi-linear response functions, necessitating a range of target
reflectances to pin down the slopes in the high and low reflectance
regimes.
Informal discussion
As mentioned above, NOAA/NESDIS is now issuing monthly
updates of the calibration status of the AVHRR solar reflectance
channels. At present this applies only to the responsivity, or
the factor g in the expression
I = (C-C_0)/g
where I is the radiance, C is the count, and C_0 is the space
count. In a parallel development, the EOC is setting up a calibration
web page which will include in-flight monitoring of the space
count. Dr Rao expressed interest in this development with a view
to adopting the space count analysis in the NOAA scheme, subject
to assessment. Dr Rao also encouraged me to coordinate a submission
from AVHRR users in CSIRO relating to implementation options for
the new AVHRR/3 instrument.
2. OTHER SESSIONS ATTENDED
2.1 The First IGAC Aerosol Characterization Experiment (ACE-1)
This session ran parallel to the satellite calibration session,
so I attended only parts of it. The extent of characterization
of composition, size distribution and spatial distribution of
aerosols over the remote Southern Ocean was impressive. Several
speakers commented on the importance of sea salt aerosol relative
to non-sea-salt sulphate in the pristine conditions, in contrast
to more polluted areas where sulphates dominate the radiatively
active size range between 0.3 and 0.7 um.
2.2 Aerosol, clouds and climate
This session contained several papers dealing with aerosol
measurements overland, and hence was directly relevant to our
proposed work toward aerosol measurement over the Australian continent.
Dr M. Andreae presented results from an observational campaign
in the Negev Desert, a site chosen because it lies at the maximum
of the radiative forcing pattern due to sulphate aerosols advected
southward from Europe. The aerosol distribution is dominated by
fine mode sulphate aerosols except during the heat of the day
when local sand and dust is lofted by convection. Surface scattering
coefficients were found to be in the range 50-150Mm-1, which may
be contrasted with the range 5-50Mm-1measured at Leigh Creek,
SA, during our field trip in September 1996. To a first approximation
this large difference is simply due to the absence of a significant
sulphate component in the outback Australian aerosol. Dr. Andraea
also attempted to relate surface scattering to column optical
depth by combining nephelometer and sun photometer measurements.
Scale heights required to reconcile the two data sets were in
the range 1-3km, corresponding to the height of the mixed layer.
This relation is of relevance to our work in attempting to characterize
Australian continental aerosols with minimal instrumentation.
The same relation was investigated by Dr M. Bergin in the context
of the ARM site in the US Southern Great Plains. Dr Bergin found
that optical depth estimates based on nephelometer measurements
of scattering coefficient at the surface multiplied by the depth
of the mixed layer consistently underestimated the column optical
depth measured by a sun photometer, to the extent of 30-50%. It
was claimed that this is most likely due to instrumental effects
such as dehydration or loss of large particles in the nephelometer.
In principle a discrepancy of this sort would be expected if the
aerosol was significantly absorbing, since the sun photometer
measures column extinction (scattering plus absorption) while
the nephelometer measures only scattering. However, Dr Bergin
discounted this by quoting single scattering albedos of 0.95.
The issue of aerosol absorption is of relevance to the Tinga Tingana
site since there is prima facie evidence for the seasonal appearance
of absorbing aerosol over north-east South Australia from SBUV
satellite measurements.
2.3 Detection of climate change and attribution to causes
Dr Judith Lean gave an interesting presentation in this session
showing that small changes in solar irradiance can explain most
of the global warming over the period 1600-1850, and about half
that observed since 1850. This finding reduces the required anthropogenic
forcing substantially. Later sessions dealt with the question
of predicting these solar changes. However it appears that the
physics of the sun's interior is insufficiently well understood
to enable predictions at the required precision. Finally, Dr Joyce
Penner described recent work aimed at further understanding the
radiative forcing due to aerosol by correlating regional temperature
distributions produced by a model with observation. Dr Penner
reported that, contrary to expectation, the addition of a biomass
aerosol to the model does not improve regional correlation over
that obtained with a sulphate aerosol alone.
SUMMARY
The meeting provided an excellent opportunity to maintain
contact with international developments in the area of satellite
sensor calibration, through presentation of our recent work in
establishing a calibration site, learning of related work overseas,
and through informal discussions. In addition, the aerosol sessions
were an unexpected bonus in the extent to which the work reported
was directly relevant to our planned observational program of
Australian continental aerosols.
Thanks are due to the EOC for providing generous financial support.