Title: MERIS land surface Albedo from data fusion with MODIS BRDFs, its validation using MISR, POLDER and MODIS (gap-filled albedo) and Data Dissemination using DDS and OGC
1MERIS land surface Albedo from data fusion with
MODIS BRDFs, its validation using MISR, POLDER
and MODIS (gap-filled albedo) and Data
Dissemination using DDS and OGC
Jan-Peter Muller (UCL) Carsten Brockmann, Marco
Zühlke, Norman Fomferra (BC) Jürgen Fischer, Réné
Preusker, Thomas Schröder (FUB) Peter Regner
(ESA/ESRIN) Professor of Image Understanding and
Remote Sensing MISR MODIS Science Team Member
(NASA EOS Project) HRSC Science Team Member (ESA
Mars Express 2003) Chair, CEOS-WGCV Terrain
mapping sub-group
2Overview
- Context
- Objectives
- BRDF/Albedo retrieval approach
- BRDF/Albedo algorithm details
- Initial Results
- Validation approach
- Preliminary Validation results
- Future Prospects
3Context
- All governments with space agencies agreed in
Brussels in February 2005 on a common strategy
for Earth Observation called - GEOSS (Global EO System of Systems) which has 9
societal benefit areas including climate
modelling, biodiversity, ecology and hazard
monitoring - ESA and the European Union have now established
the funding for their GMES (Global Monitoring of
Environment and Security) programme which
embodies these GEOSS principles - CEOS (Committee on Earth Observing Systems) is
now the provider of the space segment including
setting ISO-level standards for cal/val - Main push is to improve interoperability between
products from the same agency and products
between different space agencies - Development of MERIS spectral and broadband
albedo is the first example of trying to fuse at
the processing chain/algorithm level between
products from different space agencies - Albedo required for climate GCM model
verification by Hadley Centre
4Objectives
- Derivation of a one-year (2003) land surface
albedo from MERIS for - 13 of the 15 MERIS wavelengths (excluding 2
inside O2 absorption bands) - 2 broadband albedos (0.4-0.7µm, 0.4-3µm)
- 16-day and MONTHLY time step for 2003
- Input Level 2 RayleighO3 corrected
- 0.05º and 10km sinusoidal spatial resolutions
- Publication of MERIS albedo browse images (as Web
Map Services layers) within the CEOS-WGISS EO
Data Portal (http//iceds.ge.ucl.ac.uk) - Publication of the associated albedo files
downloadable through a cascaded Web Coverage
Server - Main driver is to improve the retrieval of
atmospheric parameters from MERIS. Hence,
spectral albedos at the MERIS wavelengths are
required - Secondary driver is the production of spectral
and broadband albedos for use by the European
climate and weather forecasting bureaus - Processsing software incorporated into the
platform-independent (Java-based) BEAM software
so that anyone can produce their own albedo
products for any other time periods - Validation by inter-comparison with other EO
sensors only envisaged at present
5BRDF/albedo approach(1)
- Inputs are orthorectified, cloud-cleared,
atmospherically-corrected Spectral/Surface
Directional Reflectances (SDRs) from level 2 data
at 1.2km spatial resolution and a typical
sampling of every 2-3 days - BRDF retrieval is NOT directly performed on these
SDRs as sampling of the bi-directional plane is
insufficient for most land surfaces given the
narrower swath (1130km) and lower temporal
sampling (every 2-3 days at the equator) of MERIS
cf. instruments such as MODIS (2550 km and daily
sampling) - Instead the BRDF shape and BRDF models are taken
for the 4 common spectral bands from the MOD43C2
(0.05º) product (see below for an
intercomparison). N.B. Bands also common with
MISR/POLDER
6BRDF/albedo approach(2)
- Using magnitude inversion, MERIS BRDFs are
calculated for each set of SDRs which are
co-located with the MODIS 0.05º pixel where MODIS
returns a value - Linear spectral interpolation is performed for
the isotropic component of the BRDF for the
remaining 9 MERIS spectral bands. (In future, it
is planned to use spectral databases such as
ASTER or SDRs from CHRIS/PROBA or CAR data to
refine this approach) - Currently spectral interpolation for the 2 sets
of broadband albedos (0.4-0.7µm, 0.4-3µm) is
performed using the MISR-equivalent bands. Work
is in progress to refine this approach - QC information is provided for 4 common spectral
albedos and Nadir BRDF Adjusted Reflectances
(NBAR) through statistical summaries of
intercomparison with MOD43C1 (albedo)/MOD43C3
(NBAR)
7BRDF retrieval vegetation
Kernel-Driven Semiempirical BRDF Model
BRDF Model Linear combination of two BRDF shapes
and a constant BRDF shapes described by kernels,
which are Trigonometric functions of incidence
and view angles Derived from physical models for
surface scattering (Ross-Thick Li-Sparse Model
Reciprocal (RTLSMR) for leaf cloud and
shadows) Analytical Form
- where
- is a constant for isotropic scattering
- are trigonometric functions
providing shapes for geometric-optical and
volume-scattering BRDFs and - are constants that weight the
two BRDFs
MOD43C2 Product supplies values of f for each
0.05º pixel and separate C code to calculate k
8BRDF retrieval vegetation
Magnitude inversion
- We determine a on a per-band basis by
- a least squares minimisation of the difference
between directional reflectances (SDRs)
predicted by the MOD43C2 BRDF parameters and
those actually measured by the MERIS sensor - The predicted measurements are found by running
the RTLSMR model in the forward mode using
the MOD43C2 BRDF parameters under the same
view and - illumination angles as the MERIS measurements
available - Performed on 4 common spectral bands between
MODIS (469,555,645,859nm) and MERIS
(490,560,665,865nm)
9Albedo retrieval vegetation
Black-sky, White-sky and solar zenith dependence
Direct Hemispherical Reflectance, is given
by
Black-sky (NO diffuse) albedo, is given by
Diffuse bi-Hemispherical reflectance, is
given by
White-sky (diffuse ONLY) albedo, is given by
10Albedo retrieval vegetation
Black-sky, Blue-sky and solar zenith dependence
The solar angle dependence can be approximated
by,
Under actual atmospheric conditions given the
aerosol optical depth, the blue-sky albedo
is given by
Where is the fraction of diffuse
skylight
11Albedo retrieval vegetation
Narrow-to-broadband conversion
Gao et al. (2003) derived a first approximation
to broadband albedo conversion factors based on
those from MISR which are taken from his paper
with VIS (0.4-0.7µm), NIR (0.7-3µm) and Shortwave
(0.4-3µm)
12Albedo retrieval scheme
MOD43C2 BRDF (0.05º) QA1 flags
BIN MERIS SDRs (0.05º x 0.05º) over 16-day MOD43C2
MOD43C3 NBAR (0.05º)
QA2 Nsamps, ave stddev, min, max
MAGNITUDE INVERSION with MOD43C2
CALCULATE ltMERISgt NBAR OVER MODIS 16 DAY PERIOD
INTERCOMPARE WITH MOD43C3
MERIS 0.05º 16- DAY NBAR
CALCULATE MERIS NBAR 0.05º DAILY
INTEGRATE MERIS ALBEDO FOR 16-DAY PERIOD
DIFF STATS
INTERCOMPARE WITH MOD43C1
MERIS 0.05º 16- DAY ALBEDOS
MONTHLY/ SEASONAL AVERAGE RE-PROJECT TO 10Km/0.05º
QA3 Nsamps, std.dev.
MOD43C1 ALBEDO (0.05º)
N.B. Status Sample products produced for
Europe. Global production completion due by end
January 2006.
INTERPOLATE ALBEDO VALUES TO 9 OTHER BANDS
INTEGRATE TO VIS, NIR, SW Broadbands
MERIS 10KM/005º 13- SPECTRAL
4 BROADBAND MONTHLY SEASONAL ALBEDOS
13First MERIS albedo product DoY 257 (16-day time
period 14/9/03-29/9/03) all bands
14First MERIS albedo product DoY 257 (16-day time
period 14/9/03-29/9/03) Band 5 (green)
15Validation approach(1)
- Difference statistics between MERIS-Albedo and
MODIS gap-filled albedo product (Moody et al.,
2005) for common bands being analysed for the
same 16-day time periods - Inter-comparisons are also being performed with
- MISR 0.5º true monthly level-3 product (2003)
- POLDER2 0.05º resampled from 6km sinusoidal
gridded 30-day products reported on the 15th of
each month (Apr03-to-Oct03) - MODIS gap-filled albedo product sampled for
weighted average of constituent 16-day time
periods within the months of Jan, Feb, Sep, Oct,
Nov-03 - Initial inter-comparisons follow with POLDER2,
MODIS gap-filled and MISR - Detailed inter-comparison also shown for MODIS
gap-filled and MISR
16Validation issue finding temporal coincidences
for 16-day products to match them up against
monthly climate modelling requirements
17MERIS (16-day,DoY257-272) cf. POLDER2 (30-day,
DoY244-273) at 0.05º resolution
N.B. Poor atmospheric correction of POLDER-2
18MERIS (16-day,DoY257-272) cf. POLDER2 (30-day,
DoY244-273) at 0.05º resolution with coastlines
MERIS 865nm
POLDER2 865nm
N.B. Very poor geocoding of POLDER-2. Decided NOT
to perform any further inter-comparisons with
MERIS and MISR until this problem is fixed
19MERIS cf. MODIS gap-filled albedo for common
bands (16-day, DoY257-272) at 0.05º resolution
MERIS 665, 560, 490
MODIS 665, 560, 470
N.B. Noticeable differences in colour and bright
albedo patterns
20MERIS vs MODIS gap-filled albedo for common bands
(16-day, DoY257-272) at 0.05º resolution
490 vs 470
665 vs 665
865 vs 869
N.B. 2D correlation improves with increasing
wavelength
21MERIS (weighted average DOY 241(13), 257(16),
273(1)) cf. MISR (30-day, DoY244-273) at 0.5º
resolution
665,560,443nm
672,558,443nm
865,665,560nm
867,672,558nm
N.B. MISR higher Albedo cf. MERIS
MERIS
MISR
22MERIS weighted average DOY 241(13/30),
257(16/30), 273(1/30) vs MISR (30-day,
DoY244-273) at 0.5º resolution
560 vs 558
443 vs 443
665 vs 672
865 vs 857
867,672,558nm
N.B. MISR albedo values higher than MERIS but
overall good correlation. Plan to compare
instantaneous MISR albedo at 1.1km with MERIS
16-day. This requires modification of BEAM
ingest for MISR Level 2AS data. This is planned
later in 2006.
23MODIS gap-filled product weighted average DOY
241(13/30), 257(16/30), 273(1/30) MINUS MISR
(30-day, DoY244-273) at 0.5º resolution
470nm
555nm
665nm
859nm
(MODIS-MISR)/MISR normalised difference albedo .
MISR always HIGHER than MODIS 2 5 10 20
50 100
24Conclusions
- First demonstration of data fusion of MERIS and
MODIS - Substantial interest in user community for
monthly (and seasonal albedo products. Little
interest in 16-day products - Simple weighted average of number of days within
a 16-day cycle appears to provide reasonable
values of monthly albedos - Significant differences between MISR and
gap-filled MODIS albedos with MISR consistently
higher than either MODIS or MERIS albedos - Some differences can be explained due to the
derivation of snow-free MODIS gap-filled product - Good agreement (as expected) between MERIS and
MODIS gap-filled products for common spectral
bands
25Planned Prospects
- Improvement in POLDER georeferencing so POLDER
can be used to compare against MISR and MERIS - Intercomparisons of monthly MISR vs MODIS
gap-filled albedo for 5 years of data - Intercomparisons of MISR L2AS with MODIS
gap-filled albedo, POLDER and MERIS - Improvement of spectral interpolation using
CHRIS/PROBA and GSFC-CAR measurements including
development of CHRIS/PROBA processing chain
within BEAM based on MERIS - Production of further years of MERIS spectral
albedos (2002, 2004, 2005, 2006) at current
resolutions - Development of modified processing chain for
production of MERIS 300m spectral albedos for
2005 using MOD43B1 (500m, Collection 5) BRDFs
including dealing with snow (explicitly) - Publication of MERIS spectral albedo browse
products as WMS layers within ICEDS and for use
by other WMS browsers as cascaded datasets - Publication of underlying MERIS spectral albedos
as WCS layers at BC including direct linkage to
BEAM and subsetting via WMS
26Data delivery current options
1- Physical media
2- Internet
ftp or http
Near Real Time (NRT) Rolling Archive or On
request
Archived data gradual availability
27ENVISAT data access to Near Real Time data
Internet Internet Telecom satellite broadcast
7-days Rolling Archive (ftp or http) 7-daysWeb File Server (http) Data Dissemination System (DDS) broadcast
Complete product Geographical selection Antenna needed
advantages Off-the-shelf Complete product Worldwide access Off-the-shelf User select part of a product and decrease the volume for download Worldwide access Independent of Internet, i.e. of network performances issues
inconvenience Rely on Internet(user network performances) Rely on Internet(user network performances) Specific equipment needed European reception coverage only (under extension to Africa)
cost for users No cost (apart from Internet connection) No cost (apart from Internet connection) 350 1.2m antenna(DDS-Europe)
28Data Dissemination System (DDS)
DDS Europe
DDS Africa
C-band
Typical 1.2 m DDS receiving antenna
About 2.5 m receiving antenna
Dissemination of global MERIS RR Level 1 Level
2 products
29ICEDS portal (http//iceds.ge.ucl.ac.uk)