History and Plan of GLI
ocean group
History from ADEOS-2 launch to the first data
release
During calibration and valibration phase from December 2002
to December 2003, GLI ocean team has concentrated our effort on
standard-algorithm improvement, vicarious calibration, and product validation
(see Table-1). We also set about data application, new-algorithm development,
and standard-algorithm improvement toward the next update after data release in
December 2003.
Table-1 history of ocean group activity
(2003/01 - 2003/12)
Vicarious calibration coefficients for the ocean
color were derived by global scheme (using global GLI radiance, SeaWiFS global
ocean color products, and atmospheric correction look-up tables) and match-up
scheme (using MOBY, AERONET, and a radiative transfer model). The derived
coefficients show that we need to apply the coefficients for ocean color
processing and they are temporally stable during the mission period except for
channels 1-3.
Using the vicarious coefficients, GLI ocean color
products achieved sufficient quality considering we took only one-year after the
launch. Comparisons with 20-120 match-ups show that error ratio of normalized
water-leaving radiance (nLw) is 20~50% for 380-565nm, and 80~280% for 625-680nm.
GLI chlorophyll-a concentration (CHLA) is sometimes scattered around coastal
areas (about 120%), but generally agrees well with ground observation data
(correlation coefficients is 0.87 on log-scale). Bulk sea surface temperature
(SST) is validated by comparing with buoy observations distributed by Global
Telecommunication System (GTS) and ship observations (RMSE is 0.67K for more
than 4000 clear match-ups).
Plan for the next
step
Version 2 algorithm update is planned in June 2004
and the products will be released in September 2004. The new algorithms will
improve temporal change of vicarious coefficients in channel 1-3, absorptive
aerosol correction, retrievals of in-water parameters in coastal area,
cloud-area detection, and stripe noise correction. GLI has six 250-m channels
from blue to shortwave infrared which are designed for the land monitoring (wide
spectral response and low SNR). We tried to make 250-m ocean color products
using the 250-m channels, and successfully produced nLw and CHLA images which
can detect fine spatial structures in the coastal area. We are planning our
action items as Table-2.
Table-2
Work plan of GLI ocean group
| Item |
Status |
1. Work for GLI Version
2
a) Improve Atmos.Corr.
- absorptive
aerosol correction (PI&GAIT)
Investigate
characteristics of the absorption (RESTEC)
- sunglint correction
(PI)
b) Tune parameters
- cloud detection for SST
(GAIT)
- cloud detection for ocean-color atmospheric
correction (Fukushima)
- coccolith, redtide, case-2 flags
(PI)
c) Collect more match-up data from collaborators
(PI&GAIT)
d) Re-process match-up data using new
L1B (operator&RESTEC)
e) Re-analyze match-up data
(PI&GAIT)
f) Consider daily change of cloud in the PAR
algorithm (PI&GAIT) |
on going
no start
on
going
on going
on going
no
start
on
going
no start
on
going
no start
|
2. Development of GLI research
products
a) Make 250m ocean products
(GAIT)
- Calculate GLI250m LUT (RESTEC)
b)
Fluororesence (PI&GAIT)
c) make flow of PAR Level 3
(GAIT)
d) Apply GLI algorithm to other high resolution satellite
data (AVNIR, MERIS..) |
on going
complete 2004/03
stop
no
start
no start |
3. Open GLI sample products via HP
(HP: GAIT, CDROM: EORC)
a) GLI around Japan (1-km)
process & open by GLI latest version algorithm
(code by
GAIT, operation and HP by RESTEC (test 06/01-, operation
07/01-))
b) 1-km Ocean products in other areas (Southeast Asia,
California, Europe,
Northeast America, Equat.Pac, and Oceania.. ) (GAIT)
c)
0.125-deg global radiance data (code & open by GAIT, operation by
RESTEC)
d) 0.125-deg global Rayleigh corrected reflectance
dataset (if requested) (GAIT)
e) 0.125-deg global ocean
products (GAIT) |
on
going
no start
complete
no start
complete |
4. EORC MODIS ocean
processing
a) MODIS NRT processing and open by GLI
latest algorithms around Japan
(code & val by GAIT,
operation & HP by RESTEC)
- calculate OTSK1A LUT at
MODIS channels (RESTEC, 2004/03)
- operation & open by
middle of Apr. 2004
- re-process from Jan 2003 to now using
GLI system (RESTEC, from 2004/07/01 using Ver.2 code)
b) Ocean
product of MODIS 500m around Japan
- develop operation code
(GAIT, 05/mid)
- operation & open (RESTEC, start from
2004/05/mid+2months)
c) 0.125-deg global radiance by MODIS
5km data
(code by GAIT, operation by RESTEC, open to
PI)
- collect MODIS 5km data
- use for vical &
algorithm development
- Start operation by middle of Apr.
2004
d) Analyze MODIS MUD (by GAIT&PI)
- use
for future algorithm development
- made from NRT data around
Japan |
on going
complete
no start
no
start
no start
no start
on going
no start
no start
no start
no
start |
5. Long term ocean color
dataset
a) Process OCTS data by GLI latest
algorithms (GAIT)
- make OCTS LUT of GLI format (RESTEC by
2004/09/30)
b) build up a closer
connection with ADEOS/Ocean homepage (RESTEC) |
no start
no
start
no start |
6. Contribute to SST project by
Tohoku Univ.
a) Provide MODIS NRT data
(operation by RESTEC) |
on
going |
last updated
on 12 April 2004 Murakami
