The world's first space-borne ocean color observation geostationary satellite was launched on June 27, 2010. Systems
and Korea Ocean Satellite Center was established for receiving, processing and distributing images captured
Geostationary Ocean Color Imager (GOCI) since 2005. Trials test of the systems had been conducted continuously for
stabilized operation since 2009. Systems in KOSC were set up to operate from receiving image to distributing data
nonstop. Because this means that stabilized operation of each system and relation of them is important, it is crucial to
figure out problem when anomaly occurred and analyze effect on each system. Also it is very significant to figure out
additional unexpected problem during in-orbit test period, analyze it and then propose solutions to it, because operation
of geostationary satellite for ocean is the first in the world.
In conclusion, we artificially make emergencies and propose solutions responding to them before lunching satellite. Also
we analyze anomalies which are occurred during in-orbit test period, then seek solutions responding them for setting up
stabilized operation. The results drawing from the paper will good source to KOSC which operate system of GOCI and
agencies concerned for 7 years from now.
The data processing software system of Geostationary Ocean Color Imager (GOCI) is composed of the image preprocessing
system (IMPS) and the GOCI data processing system (GDPS). IMPS generate GOCI level 1B from raw
satellite data and GDPS is the post-processing system to generate GOCI level 2.
IMPS have a radiometric correction module as IRCM and a geometric correction module named as INRSM. The former
is focused on equipment's mechanical noise reduction and radiometric accuracy and the latter image navigation and
image registration accuracy by landmark matching method and image mosaic method.
GDPS have the atmospheric correction algorithms, as the spectral shape matching method (SSMM) and the sun glint
correction algorithm (SGCA), and BRDF algorithm to solve bi-directional problem. Several Case-II water analytical
algorithms, like chlorophyll concentration, suspended sediment and dissolved organic matter, are contained in GDPS.
Also, GDPS will generate the value added product like water quality, fishery ground information, water current vector,
During in-orbit test period planned six months after successful launch of satellite, IMPS and GDPS will be verified with
respect to those requirements and algorithms and functionality and accuracy by pre-defined test procedure like test,
inspection, demonstration. And then those configuration parameters will be modified and the algorithm descriptions will
be updated. In this paper, we will present the preliminary analyzed results of data processing system test and update
planning during in-orbit test.
Since Geostationary Ocean Color Imager (GOCI) data are not available yet, we used daily MODIS Chlorophyll a (Chla)
data to illustrate how GOCI data can be used for physical and biological interaction research. For physical features, we
used the daily New Generation Sea Surface Temperature (NGSST) for the Yellow Sea, the South Sea, and the East Sea
from January 2005 to December 2009. Since the cloud contamination in ocean color observations are always
programmatic, analyzing physical and biological interactions have been limited. In order to examine whether we can use
NGSST for Chla using a linear regression, we investigated their relations to obtain cloud free Chla. The results show
that the ES and the SS have relatively small root mean error (RMSE) than that in the YS. In addition to time series of
two different observations, we applied empirical Mode Decomposition (EMD) to extract different spatial features from
both Chla and SST imagery. We selected off the west coast of the ES for a jet like feature on August 13, 2007. The
Chla meandering features were different from previously reported upwelling features in the area. The features seem like
to be modulated by waves, which were appeared in SST decomposition modes, i.e., Intrinsic Mode Decomposition
Although the methods were applied to MODIS observations, which are coarser spatial and temporal resolutions than
those of GOCI, these methods will provide better results with GOCI observations because of better resolutions.
The Geostationary Ocean Color Imager (GOCI) on board the Communication Ocean Meteorological Satellite (COMS)
requires accurate atmospheric correction for the purpose of qualified ocean remote sensing. Since its eight bands are
affected by atmospheric constituents such as gases, molecules and atmospheric aerosols, understanding of aerosolradiation
interactions is needed. Aerosol optical properties based on
sun-photometer measurements are used to analysis
aerosol optical thickness (AOT) under various aerosol type and loadings. It is found that the choice of aerosol type
makes little different in AOT retrieval for AOT<0.2. These results will be useful for aerosol retrieval of COMS/GOCI
Precise geographic mapping of remote sensing (RS) data is prerequisite for time series analysis of multiple RS data.
Conventional georegistration of scanned image along satellite track, of which spatial resolutions on earth surface are not
uniform, requires heavy computational efforts. A rapid and precise mapping method, referred to as the Piecewise Affine
Transform (PAT) method, is developed. In PAT method, each polygon of which vertices are defined by ground control
points (GCP) are mapped to target grids, and affine transforms are used in determining the pixel location of scanned
image in mapping data inside of each polygon. The PAT method does not involve any 'proximity searching', and
mapping by PAT method is very rapid. For typical scanned data of a few thousands lines, mapping look up table (LUT)
can be made within a second in personal computer. Using LUT data base, mapping can be easily made by the nearest
neighbour or bilinear interpolation method. The PAT algorithm is tested by mapping MODIS scanned image, and found
precise mapping was achieved.
The first geostationary ocean color sensor, Geostationary Ocean Color Imager (GOCI), on board the Korean
Communication Ocean and Meteorological Satellite (COMS), was successfully launched on June 26 of 2010. GOCI
includes 8 spectral bands in visible and near-infrared wavelengths with a coverage area of 2,500×2,500 km2 centered at
36°N and 130°E over the Korean seas. GOCI will provide an important capability to monitor ocean phenomenon with
one hour temporal and 500 m spatial resolutions for a better understanding of biogeochemical processes in the Korean
seas. However, there are uncertainties in estimating bio-optical properties since water properties in large areas of
Koreans are optically characterized as Case-2 waters due to strong tidal mixing and large amount of river discharges.
The newly-developed semi-analytical algorithm of diffusion attenuation coefficient at the wavelength of 490 nm,
Kd(490), for the turbid coastal waters was assessed using in situ radiometric and Kd(490) measurement obtained from
clear and turbid waters over the global ocean. Results of the Kd(490) data using the new model is well correlated with the
in situ Kd(490) measurements. Synoptic maps of Kd(490) data from the Moderate Resolution Imaging Spectroradiometer
(MODIS) onboard the Aqua satellite using the new model were derived in the Yellow Sea and East China Sea. The
MODIS-derived Kd(490) data show significant increased values along the turbid coastal waters including the Bohai Sea
and the Yangtze River Estuary. In general, the highest Kd(490) appeared in winter and the lowest Kd(490) are presented
in summer over the all area. Interannual variability of Kd(490) in timing and magnitude is apparent, but there is no
consistent trend of interannual variability across all areas.