Aucun titre de diapositive

1
CALIBRATION OF SOLAR SPECTRUM SATELLITE
RADIOMETERS OVER DOME CONCORDIA,
ANTARCTICA Delphine Six1 - Michel Fily1 -
Patrice Henry2 Steve Warren3 Richard Brandt 3
Philippe Goloub 4 1 Laboratoire de
Glaciologie et Géophysique de lEnvironnement
CNRS 54 rue Molière - 38 402 Saint Martin
dHères Cedex (six_at_lgge.obs.ujf-grenoble.fr) 2
Centre National dEtudes Spatiales 31 041
Toulouse Cedex 3 University of Washington
Department of Atmospheric Sciences Seattle,
USA 4 Laboratoire dOptique Atmosphérique,
Université des Sciences et techniques de Lille,
59 655 Villeneuve dAscq
PROJECT SUMMARY
Remote sensing is an essential tool for studying
our environment at many different scales. The
calibration of remote satellite sensors is
necessary for quantitative measurements of the
earth surface or atmospheric geophysical
parameters. It is particularly necessary to
assess long term changes from satellite
sensors. Different calibration techniques are
used before launch, internal sources, and
natural targets. Dome Concordia, located on the
high plateau of the Antarctic ice sheet, would be
another type of natural target, like deserts for
example. The Dome Concordia area has many
characteristics which are of interest for
calibration. The snow surface is spatially
homogeneous at the 100 km scale, surface
roughness is weak with a very small slope, and
temporally due to low accumulation and slow wind.
The site is located at 3200 m with a clear
atmosphere with almost no aerosol and very little
water vapor. At its latitude (75S), the site is
frequently viewed by many satellites at different
angles and, in this regard, is better than South
Pole. The presence of the Dome Concordia station
will permit ground measurements for many years.
Several studies have investigated the use of the
Antarctic and Greenland ice sheets to determine
temporal changes of satellite sensors in the
solar spectrum but, so far, no ground truth
experiment has been carried out. This is the main
objective of our project. This project has two
components. First, we examine the variability of
the ground snow surface from data of the SPOT4
satellite (Vegetation 1 sensor) to test the
quality of this area as a calibration site. Then,
we carried out field measurements to characterize
the surface and the atmosphere over Dome C
station.
PART I SPOT4/VEGETATION DATA PROCESSING
Objective To test the snow ground surface
stability to assess Antarctica as a new future
calibration site and to determine the drift of
the sensors
Snow reflectance data analysis Spatial
variations - Main results ? Reflectance
different from one zone to the other ?
Roughness present at the bottom of the area ?
Reflectance standard deviation generally, never
more than 3 Mensual variations - Main results
Due to the evolution of the irregularities over
the season (wind effect, T effect) ? Maximum
of variability in January and February. November
and December are more stable Annu
al variations - Main results ? No more
than 2 of evolution (on top of the area, for
November and December data)
The sensor Satellite SPOT 4, launched in
1998 Sensor VEGETATION 1, ground pixel
resolution of 1km 4 channels Available data
4 years (1998-2002) Area under study 716
716 km centred over Dome C Station (75S,
123E)
Preliminary investigations ? Cloud masking
Using variance calculation on each pixel and
the information contained in the B4 channel (1.6
microns). Example of cloud
masking (images in B0, B4 and mask of 24th
December 1999) ? Atmospheric corrections Using
the 6S model (Second Simulation of the Satellite
Signal in the Solar Spectrum, Vermote et al. -
1997) adapted to an Antarctic atmosphere by
Alvain (2001) Introduction of daily ozone data
from TOMS satellite in each picture. ? Snow
surface reflectance data base Characteristics of
the data base grid boxes of 120120 km, each
month, each year, only on B0 channel solar zenith
angle step 2 in the backward (azimuth angles
between 50-70) and forward hemisphere (azimuth
between 230 250) 36 sections (Dôme C
is located box n15) obtained from the initial
picture, and representation of the
back and forward hemisphere
Standard deviation of ground reflectance in
function of viewing zenith angle (in the back and
forward hemispheres) for Nov., Dec., Jan. and
Feb. Example for solar zenith angle from 60 to
66 Azimuth angles 50-70 and 230-250
Evolution of the reflectance in function of time.
The reference year is year 1 (1998), for
November, December, January and February, in the
forward hemisphere.
PART II DOME C FIELD CAMPAIGN
Objectives to measure the snow surface and
atmospheric characteristics at Dome C station
Work done or tested in the field Atmospheric
characteristics Surface
characteristics Daily
ozone measurements (MICROTOP)

Logistical aspects 2 scientists in the field
From 2nd January to 5th February 2003 Location
1.3 km far from summer camp. undisturbed surface
(no tracks, no hills) Map of Dôme C
station (summer and winter camp, and the field
measurement site)
Aerosols composition (size, distribution.)
CIMEL sunphotomter MICROTOP back up ? Very
low values of Aerosol Optical Thickness (values
lower than 0.1)
Construction of a 33 meters tower (January 2003)
And bidirectional reflectance tests from the top
of the 33 meters tower (R.Brandt)
Daily ozone value from 3rd January to 4th
February 2003
CONCLUSION AND PERSPECTIVES
Data processing Vegetation images show that
Dome C area is will adapted for a new calibration
site as the spatial variability of the
reflectance is less than 3. The drift of the
sensor was then tested and shows that the drift
is in the limit of the conditions proposed by the
CNES, never more than 2 on 4 years. Field
campaign CIMEL instrument gave 7 days of
complete Aerosol Optical Thickness data. These
data will be analyzed and compared with MICROTOP
back up. The ozone value will also be compared
with TOMS satellite data. Results of
bidirectional reflectance of snow were
successfully tested this year before continuous
measurements next year. During the next season
2003-2004, we will plan CIMEL measurements and
bi-directional measurements 24 hours per day.
Acknowledgments Centre National dEtudes
Spatiales (CNES), University of Washington
(Seattle), Laboratoire dOptique Atmosphérique
(Lille), Institut Polaire Français (IPEV),
Institut Polaire Italien (ENEA) Ref Vermote
E. et al, 1997, Second Simulation of the
Satellite Signal in the Solar Spectrum, 6S user
guide, version 2 Alvain S., 2001, Développement
dune méthode de correction atmosphérique pour
létalonnage du capteur SPOT4/Végétation sur
lAntarctique, Rapport de Diplôme Etudes
Approfondies, Université de Grenoble.