Kein Folientitel

1
WP 8 Impact on Satellite Retrievals
Partners (according to Contract)
University of lAquila (DFUA 12) Vincenzo Rizi
Ecole Polytechnique (EPFL 13) Bertrand Calpini
Observatory of Neuchatel (ON 14) Valentin Mitev
Met. Institute Munich (MIM 17) Matthias Wiegner
2
Contract Objectives and input to workpackage
1. Quantify the sensitivity of radiances at the
top of the atmosphere (toa) to variations of the
aerosol distribution as typically observed at
lidar sites. This will be used to provide an
estimate of the impact of aerosols on retrievals
of a variety of quantities using optical methods
from spaceborne platforms. 2. Provide aerosol
distributions for calibration purposes upon
request from relevant groups operating satellite
sensors.
3
Contract Description of work

• This is done by running adequate radiative
  transfer models
• using the aerosol data from the lidar
  measurements as input.
• As a result, inversion algorithms for surface
  properties can be
• critically reviewed and improved.
• 2. Additional lidar measurements of aerosol
  vertical distribution
• will be performed at special occasions when
  satellite overpasses
• close to an EARLINET station occurs and the sky
  is not overcast.
• For this part to stay within the limits of the
  available resources
• it will be done only on request and in close
  co-operation with
• the satellite community.

4
Why is this WP relevant (1. reason)?
multispectral and multiangular radiances from
satellites used to retrieve aerosol optical depth
critical over land surfaces (unknown ground
albedo, orography)
complex retrieval algorithms require validation
Validation of atmospheric parameters
5
Why is this WP relevant (2. reason)?
multispectral and multiangular radiances from
satellites used to retrieve surface properties
(e.g. precision farming)
determine atmospheric influence (unknown aerosol
effect)
corrections of atmospheric masking require input
and validation
Validation of surface parameters
6
Why is this WP relevant (3. reason)?
Radiative transfer calculations required to
understand satellite data and to develop
parameterizations
aerosol distribution is one controlling parameter
realistic modelling requires typical input
datasets
Improvment of aerosol understanding
7
Strategy of WP 8
Demonstrate the benefit of lidar data on the
basis of examples (i.e. motivate users)
Offer our datasets to the satellite
community (i.e. activate users)
Identify cases where lidar data are useful (i.e.
attract users)
Note WPs ordered by time allocated WP8 is
15. out of 20.
8
First contribution
Goal Support satellite retrieval validation by
supplying lidar data (possible candidates
Envisat community)
Actions proposed Supply survey of available
aerosol extinction profiles (site, time,
wavelength)
Data dissemination On request (Webpage)
9
Second contribution
Goal Full characterization of surface and
atmosphere of exactly the same scene (for
calibration of satellite sensor and algorithms)
Requirements co-incidence and co-location and
very small satellite pixel required.
Time and Place May, June, July and August 2002
in Gilching
Acquisition mode CHRIS 18 km swath, 25 m
resolution, 19 spectral bands, along track (5
angles)
In co-operation with
10
Third contribution
Goal Support model development by supplying
lidar data
Actions Performing model calculations with
realistic aerosol extinction profiles to
investigate the influence of aerosols on
toa-radiances
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
Next steps of WP 8
Start May 2000 End December 2002
Deliverables
April 2002 Report on aerosol impact on satellite
retrievals
Other Deadlines
May 2002 Contribution to Annual Report
15
Stop here
16
Goals of the Work-Package
The goal of Workpackage No. 8 includes the
modeling of the aerosol influence on radiances
measured by satellites and the provision of
additional lidar measurements on request. What
does this mean? Measurements Lidar data are
available since May 2000. Dedicated measurements
simultaneous to satellite overpasses make sense
if pixel are small and cloud free conditions can
be guaranteed. On the other hand, the existing
data base can be used for validation of
satellite measurements and their
products. Model calculation Models for
atmospheric corrections (e.g., to retrieve
surface properties) and models to derive aerosol
properties can be supported by supplying lidar
data. Remark The development of such models
itself is beyond the scope of EARLINET. Both
classes are linked and cannot be considered
separately.
17
Small but relevant Workpackage
Satellite provide global coverage but vertical
resolution is poor. Aerosols are hard to be
detected by passive radiometers, thus Validation
and additional data are useful. Aerosols
influence retrievals of geophysical and
atmospheric parameters
Global climate models will benefit of high
resolved aerosol data as Well as (conventional)
dedicated aerosol missions
18
EARLINET and Satellites
General remarks Meteorological satellites
suitable for aerosol remote sensing require
good spatial and spectral resolution. For that
reason, SeaWIFs is presently the most promising
candidate. Geostationary satellites have poor
radiometric accuracy and spectral resolution,
GOME et al. have very poor spatial resolution,
Landsat et al. have very poor temporal sampling,
and sensors with very high spatial resolution
are not yet in orbit (MERIS 250 m, Chris et
al. 25 m). Thus, we follow two options Option
1 plan dedicated experiments on the compare
same atmospheric volume-concept risk overcast
conditions and Option 2 select data sets
already available on the validate aerosol
parameters-concept
19
Option 2 Aerosol Validation
Output Several calibration points for the map of
aerosol optical depth derived from (e.g.) v.
Hoyningen-Huenes SeaWIFs retrieval. Information
of special aerosol stratifications that might
help to explain possible deviations. E.g., check,
whether algorithm works in the presence of
Saharan dust layers. Provision of information of
the aerosol type (if possible, e.g., from
trajectories, lidar data themselves, auxiliary
data) to support satellite retrieval algorithm
(input for them).
Possible co-operation with University of Bremen
(v. Hoyningen-Huene)
20
Option 2 Aerosol Validation (contd.)
Background Information A SeaWIFs algorithm to
derive aerosol optical depth exists and has been
(successfully) applied. SeaWIFS has a spatial
resolution of about 1 km and a coverage of 1800
km (swath width) similar to AVHRR Algorithm
works best in the spectral range between 412 -
510 nm (surface is dark) lidar data of 532 nm
can be extrapolated. Times to be compared should
be in spring and early summer (green vegetation
no problems with water stress) MERIS will have a
better resolution but will be available not
before spring 2002. Sciamachy has a very poor
spatial resolution.
21
Option 2 Aerosol Validation (contd.)
To be Discussed Data from stations not directly
involved in this work package would be required.
Is that possible? Who will calculate the
optical depth from the extinction profiles (owner
or M.W.)? Is an extra qualitity check
required/desired by the owner of the
data? Selection of episodes from the diurnal
cycle subset (best time of the day is 11-13
hours)? How many episodes should be selected
(one, two, more?) Should we include algorithms
to derive aerosol optical depth over land from
other institutes (answer from Berlin is pending)?
22
Goals of the Work-Package
The goal of Workpackage No. 8 includes the
modeling of the aerosol influence on radiances
measured by satellites and the provision of
additional lidar measurements on request. What
does this mean? Measurements Lidar data are
available since May 2000. Dedicated measurements
simultaneous to satellite overpasses make sense
if pixel are small and cloud free conditions can
be guaranteed. On the other hand, the existing
data base can be used for validation of
satellite measurements and their
products. Model calculation Models for
atmospheric corrections (e.g., to retrieve
surface properties) and models to derive aerosol
properties can be supported by supplying lidar
data. Remark The development of such models
itself is beyond the scope of EARLINET. Both
classes are linked and cannot be considered
separately.