Remote sensing of ocean color from space aims at retrieving from a noisy top-of-atmosphere radiance the values taken by some relevant quantities like the chlorophyll-a concentration or the marine reflectance. From a mathematical perspective, it is an ill-posed inverse problem with a highly nonlinear operator. Few techniques are available in the case of a nonlinear inverse problem; even its theoretical study is far from easy, yet some techniques may be used in a practical setting when the noise distribution is known. However in the case of ocean color remote sensing, the noise encompasses several types of error owing to the forward operator approximation (radiative transfer model) as well as to calibration and pure measurement noise. Hence the noise distribution is unknown. In this work, a stochastic technique is proposed to first infer a noise distribution, which is next used to retrieve the marine reflectance in a least-square prediction setting by a regression model. The methodology is illustrated on actual data originating from the SeaWiFS sensor.

Ocean environment attracts more and more attention, whereas monitoring ocean environment using ocean color imagery (e.g., MODIS) has become one of the important fields in modern oceanography. After analyzing a few individual modules of the radiative transfer process, an end-to-end numerical model for ocean remote sensing is developed. This model combines MODTRAN and HYDROLIGHT, both are state-of-the-art radiative transfer models for atmosphere and water, respectively. Also, a simple but realistic cloud model is added to it. This combined model calculates the downward radiance on water surface using MODTRAN and the independent cloud model, which replaces the empirical and semi-empirical models used in RADTRAN (Gregg and Carder 1991). Especially, with information of cloud's location and brightness from by the cloud model, the combined model provides more accurate radiance on water surface. Further, the effects of cloud position (from 30 degree to 180 degree for the cloud central azimuth angle) and coverage (from about 10% to 80%) on retrieved chlorophyll concentration by standard MODIS algorithm is analyzed. It is found that the nearer the cloud to the Sun the smaller the effect on the derived chlorophyll.

A methodology is presented to estimate aerosol altitude from reflectance ratio measurements in the O2 absorption A-band. Previous studies have shown the impact of the vertical distribution of scatterers on the reflectance ratio. The reflectance ratio is defined as the ratio of the reflectance in a first spectral band, strongly attenuated by O2 absorption, to the reflectance in a second spectral band, minimally attenuated. First, a sensitivity study is performed to quantify the expected accuracy for various aerosol loadings and models. An accurate, high spectral resolution, radiative transfer model that fully accounts for interactions between scattering and absorption is used in the simulations. Due to their adequate spectral characteristics, POLDER and MERIS instruments are considered for simulations. For a moderately loaded atmosphere (i.e., aerosol optical thickness of 0.3 at 760 nm), the expected error on aerosol altitude is about 0.3 km for MERIS and 0.7 km for POLDER. More accurate estimates are obtained with MERIS, since the spectral reflectance ratio is more sensitive. Second, the methodology is applied to MERIS and POLDER imagery. Estimates of aerosol altitude are compared with lidar profiles of backscattering coefficient acquired during the AOPEX-2004 experiment. Retrievals are consistent with measurements and theory. These comparisons demonstrate the potential of the differential absorption methodology for obtaining information on aerosol vertical distribution.

Ocean color products (aerosol, normalized water-leaving radiance, and chlorophyll-a concentration) were produced using 250-m resolution data (center wavelengths at 462, 543, 661, 824, 1645, and 2194nm) of Global Imager (GLI) on ADEOS-II by GLI standard ocean atmospheric correction algorithms (i.e., basically same as the 1-km algorithm). The 250-m ocean color products can show finer spatial structures than standard 1km products, but we found some problems; (a) there is large noise because GLI 250-m channels were designed originally for bright area (land vegetation), (b) we have to use channels which are not optimized for the atmospheric correction (wide bandwidth, large noise, and including slight water-vapor absorption), and (c) we need more consideration about cloud shadow, sea-surface reflectance (sunglint and white cap), and shallow bottom. Sea surface reflectance has fine spatial structures in the actual ocean due to fine structure of surface winds influenced by the coastal topography, and we often cannot distinguish the surface reflection from aerosol scattering above the sea surface. JAXA is planning a new mission, "Second generation GLI (SGLI) on GCOM satellites", which has visible and near-infrared channels with 250-m spatial resolution improving noise level and bandwidth applicable to the coastal ocean-color observations.

A Chinese Moderate Resolution Imaging Spectrometer (CMODIS) on "Shenzhou-3" spaceship was launched on March 25, 2002. CMODIS has 34 channels, with 30 visible and near-infrared channels and 4 infrared channels. The 30 channels are 20nm width with wavelength ranging from 403nm to 1023nm. The radiance calibration of CMODIS was finished in the laboratory measurements before it was launched and the laboratory calibration coefficients were used to calibrate the CMODIS raw data. Since none of on-board radiance absolute calibration devices including internal lamps system and calibration system which is based on solar reflectance and lunar irradiance were installed with the sensor, how about the accuracy of CMODIS-measured radiance is a key question for the remote sensing data processing and ocean applications. A new model was developed as a program to evaluate the accuracy of calibrated radiance measured by CMODIS at the top of the atmosphere (TOA). The program can compute the Rayleigh scattering radiance and aerosol scattering radiance together with the radiance component from the water-leaving radiance to deduce the total radiance at TOA under some similar observation conditions of CMODIS. Both the multiple-scattering effects and atmosphere absorbing effects are taken into account on the radiative transfer model to improve the accuracy of atmospheric scattering radiances. The model was used to deduce the spectral radiances at TOA and compared with the radiances measured by Sea-viewing Wide Field-of-view Sensor (SeaWiFS) to check the performance of the model, showing that the spectral radiances from the model with small differences from those of SeaWiFS. The spectral radiances of the model can be taken as reference values to evaluate the accuracy of CMODIS calibrated radiance. The relative differences of the two radiances are large from 16% to 300%, especially for CMODIS at the near-infrared channels with more than one time larger than those of the model. It is shown that the calibration coefficients from the laboratory measurements are not reliable and the radiance of CMODIS needs to be recalibrated before the data are used for oceanography applications. The results show that the model is effective in evaluating the CMODIS sensor and easily to be modified to evaluate other kinds of ocean color satellite sensors.

This study shows match-up analysis of chlorophyll-a concentration in coastal area in Upper Gulf of Thailand. An applicability of atmospheric correction are investigated in turbid area. When a suspended matter concentration is over 7 g/m3 in a mouth of Bangpakong river, atmospheric correction was failed, then chlorophyll-a concentration could not be estimated. Three algorithms which are MODIS (Moderate Resolution Imaging Spectroradiometer) standard, neural network for GLI (Global Imager) and regional empirical algorithm are compared using match-up data set. The regional algorithm has better correlation than other algorithms and its RMSE was minimum in three algorithms. MODIS standard algorithm has good performance in higher than 1mg/m3, however, CHL was overestimated in lower concentration.

Estimated total sediment mass in the fluid mud system is about 4.25×10^8 tons per year as river inputs into Yangtze River estuary. The spatial-temporal distribution of suspended sediment concentration (SSC) is concerned by estuarine and coastal researchers. In this work, in situ measurements were carried out in Yangtze River estuary in Sep.2004, Jul. and Aug.2005 and Feb.2006. An experiment in barrel on land as a complementary measurement was conducted in July 2006. All of the spectral signatures observed show three spectrum-peak (at wavelengths 590nm, 690nm and 810nm around) characters more sensitive to SSC variation at bands 500-900nm. Remote sensing reflectance (Rrs) at 590nm around is bigger than that at 690nm around when SSC is about <0.1g/l, but this situation would be opposite when the SSC surpasses about 0.15g/l. The Rrs at 810nm around is lower than that respectively at 590nm and 690nm around, except that SSC increased to one close to or even > 1.0g/l. Here a parameter SRP (Sediment Response Parameter) described a changeable rate of ratios among three spectral peaks was defined, which presented a logarithmic decreasing tendency with the SSC increasing. The preliminary result shows that SRP can reflect SSC variation in the estuary.

With the increasing recognition of the need for using the NIR bands for chlorophyll retrieval in coastal waters it is necessary to account not only for the spectral modulation of the total elastic backscatter by the chlorophyll absorption spectra, as it is normally done, but to also take into account the spectral signature of the backscatter itself, whether from mineral or organic particulates, including algae, and to assess how these factors effect retrieval algorithms. Based on our recent field measurements in coastal waters, we have undertaken a study to examine the spectral behavior of the backscatter to total scattering ratio as a function of suspended solids and chlorophyll loadings. The total scattering spectra is obtained using the WET Labs AC-S instrument which provides hyperspectral measurements of absorption and attenuation, in conjunction with the bb9 instrument which provides direct measurement of backscatter, as well fluorescence measurement of chlorophyll concentration [Chl]. The relevant WET Labs absorption and attenuation data were then used as input into Hydrolight radiative transfer simulations to obtain the backscattering ratio spectral distributions. Preliminary NIR algorithms, which were evolved for high [Chl] coastal waters and which focus on the contribution of spectral changes due to chlorophyll backscattering in the NIR, are presented. It is expected that these algorithms will ultimately prove to be less dependent on regional tuning.

An empirical method is presented here to estimates bulk particulate refractive index using the measured inherent and apparent optical properties from the various waters types of the Arabian Sea. We have used the empirical model, where the bulk refractive index is modeled in terms of the backscattering ratio and the hyperbolic slope of the particle size1. Empirical models are obtained to determine the slope of the particle size and backscattering ratio as a function of the remote sensing reflectance. We have used these algorithms with the IRS-P4 OCM satellite data and derived the refractive index image. The values of indexes are found to be lower for the open ocean and relatively higher for the coastal waters. Distinct features are observed even in the in the waters with similar chlorophyll concentrations. This algorithm provides scope for the classification of water types and detection of blooms.

An artificial neural network (ANN) based procedure is developed to estimate concentrations of Chlorophyll-a, Suspended Particulate Matter (SPM) and absorption due to chromophoric dissolved organic matter (CDOM) in the seawater from satellite detected normalized water-leaving radiance (nLw) of the IRS-P4 Ocean Colour Monitor (OCM) satellite data. An ocean colour reflectance model was used to generate surface spectral reflectance corresponding to first five bands of IRS-P4 OCM sensor, using three optically active oceanic water constituents as inputs. ANN model making use of reflectance in five visible bands was trained, tested and validated to invert the spectral reflectance for the simultaneous estimation of three optically active constituents. The retrieved chlorophyll-a concentrations from ANN based algorithm were compared with the corresponding chlorophyll-a distribution obtained by global empirical algorithms e.g. Ocean Chlorophyll-4 (OC4) algorithm. ANN derived chlorophyll-a estimates were found to have reduced the over estimation in coastal waters often observed with the use of band ratio algorithms.

Costal Zone Color Scanner (CZCS) was the first of its kind sensor flown in 1978 for ocean studies. It gave quantitative estimate of phytoplankton. After that, many missions were launched by various agencies like IRS-P3-MOS, Sea-WiFS, OCTS, MODIS, MERIS etc. The data from these sensors was utilized to study, develop retrieval algorithms, validate the models and define the new sensor parameters for better extraction and estimation of chlorophyll and other ocean biological and physical parameters on a routine basis. ISRO's Ocean Color Monitor (OCM-1) was launched in 1999 onboard Oceansat-1 (IRS-P4), polar orbiting satellite. This paper highlights some of the important features, specifications and performance of OCM sensor. It also briefly gives some of the enhancements planned for OCM-II. OCM provides multi-spectral imagery with narrow spectral width having large field of view of ±43° and ground resolution of 360m. The sensor is designed to cater to land and ocean applications globally and accordingly it has high radiometric sensitivity with large dynamic range. The radiometric performance is realized using low noise circuit designs and practices. The optics and detectors are mounted on a highly stable structure, which facilitates achieving band-to-band registration, minimizes veiling glare and temperature excursions and avoids sun glints. To provide continuity in services, OCM-II is being developed for Oceansat-II, slated launch second half of 2007. OCM is the only sensor which meets both ocean and land applications simultaneously.

The Monitoring Aerosols in the Ultraviolet Experiment (MAUVE) and the Short-Wave Infrared Polarimeter Experiment (SWIPE) instruments have been designed to collect, from a typical sun-synchronous polar orbit at 800 km altitude, global observations of the spectral, polarized, and directional radiance reflected by the earth-atmosphere system for a wide range of applications. Based on the heritage of the POLDER radiometer, the MAUVE/SWIPE instrument concept combines the merits of TOMS for observing in the ultra-violet, MISR for wide field-of-view range, MODIS, for multi-spectral aspects in the visible and near infrared, and the POLDER instrument for polarization. The instruments are camera systems with 2-dimensional detector arrays, allowing a 120-degree field-of-view with adequate ground resolution (i.e., 0.4 or 0.8 km at nadir) from satellite altitude. Multi-angle viewing is achieved by the along-track migration at spacecraft velocity of the 2-dimensional field-of-view. Between the cameras' optical assembly and detector array are two filter wheels, one carrying spectral filters, the other polarizing filters, allowing measurements of the first three Stokes parameters, I. Q, and V, of the incident radiation in 16 spectral bands optimally placed in the interval 350-2200 nm. The spectral range is 350-1050 nm for the MAUVE instrument and 1050-2200 nm for the SWIPE instrument. The radiometric requirements are defined to fully exploit the multi-angular, multi-spectral, and multi-polarized capability of the instruments. These include a wide dynamic range, a signal-to-noise ratio above 500 in all channels at maximum radiance level, i.e., when viewing a surface target of albedo equal to 1, and a noise-equivalent-differential reflectance better than 0.0005 at low signal level for a sun at zenith. To achieve daily global coverage, a pair of MAUVE and SWIPE instruments would be carried by each of two mini-satellites placed on interlaced orbits. The equator crossing time of the two satellites would be adjusted to allow simultaneous observations of the overlapping zone viewed from the two parallel orbits of the twin satellites. Using twin satellites instead of a single satellite would allow measurements in a more complete range of scattering angles. A MAUVE/SWIPE satellite mission would improve significantly the accuracy of ocean color observations from space, and will extend the retrieval of ocean optical properties to the ultra-violet, where they become very sensitive to detritus material and dissolved organic matter. It would also provide a complete description of the scattering and absorption properties of aerosol particles, as well as their size distribution and vertical distribution. Over land, the retrieved bidirectional reflectance function would allow a better classification of terrestrial vegetation and discrimination of surface types. The twin satellite concept, by providing stereoscopic capability, would offer the possibility to analyze the three-dimensional structure and radiative properties of cloud fields.

The Changjing River triangle area, where includes Jiangsu and Zhejinag province and Shainghai and so is called as Changjing delta, is a key area of Chinese economic development, but the economic sustainable development of Changjing delta in last ten years is restricted by coastal water quality deterioration, such as nitrogen and phosphorus content increasing, eutrophication, red tide and man activity pollution. The routine marine water quality assessment by boat, buoy and coastal observation station sampling is difficult to monitor its special and timely variation. In this paper, first, the situation of water quality of Changjing delta is introduced. Second, the satellite remote sensing algorithm of retrieve the parameters of water quality, such as total nitrogen (TN), total phosphorus (TP) and transparency, are discussed in detail. Finally, the rule of water quality classification briefly is mentioned and the water quality classification images are presented in the paper. The preliminarily result shows that the ocean color satellite data has its latent capability of quasi-realtime coastal water quality monitoring.

The understanding of the dynamics of suspended matter is crucial for coastal engineering projects and for food web in the coastal environment. The turbidity determines also optical properties of the waters, and is crucial for the coral reef system. Gulf of Kachchh is one of the macrotidal sites of India with occurrence of numerous harbors and an ecologically fragile coral reef system. The macrotidal gulf has high magnitude currents which amplify inland with tide into the gulf. A sequential analysis of imageries of the Ocean Color Monitor of IRS P-4, during different tidal phases reveals that there is a large variation in the TSM contents of the gulf associated with tidal cycle. The suspended matter dispersal under different tidal phases in the gulf is evaluated. The source to sink pathways of the suspended matter into different regions of the gulf, specifically in the vicinity of the National Marine Park, are determined which is important for restoring the stressed coral reefs.

Coral reefs are complex marine ecosystems and environment-sensitive that are constructed and maintained by biological communities that thrive in the warm tropical ocean. Remote sensing provides a more efficient way to observe and monitor shallow coral reefs than alternative field survey. Dong-Sha Atoll is located at the northern continental margin of the South China Sea. The area of the atoll is more than 300 km2 with an average depth of 10 m, and has being abused by destructive fishing during recent decades. In this paper, three satellite broadband multi-spectral imageries (Quickbird2, ETM+, and SPOT5) are used to survey this area. In the preprocessing, low-resolution images are geometrically corrected to high-resolution imagery. For information extraction, unsupervised classification functions are adopted for ETM+ and SPOT5 data, and supervised classification method for Quickbird2. The classification results are (or not) merged into coral reef, and operated by vectorization, simplification, and topological analysis. There are 1331 coral reefs larger than 100 m2 with the detection limitation of 3×3 pixels at multi-band data of Quickbird2. The results extracted from SPOT5 data are better than that of ETM data at silty lagoon due to its higher resolution, and less at atoll reef bodies in the absence of blue waveband.

The history of satellite radar altimetry stems from the need to capture a global view of the surface topography of the oceans. As altimeters are specifically designed for global observations, they encounter major problems in coastal regions, such as relatively poor sampling and inaccuracy of the corrections, so measurements are generally discarded. Nevertheless, a global archive of 15 years of raw data from a series of missions is presently available. The huge amount of unused data in coastal regions can be re-analyzed, improved and more intelligently exploited, possibly promoting coastal altimetry to the rank of operational service. This paper outlines the obstacles limiting the use of the data, discusses some areas of improvement, shows the lessons learned from a case-study in the Mediterranean Sea, and shows that the improved coastal altimetry concept can be extended to other regions, e.g. along the coasts of India. This paper also explores the implications of adopting the emerging vision of the Internet infrastructure in the coastal altimetry context: a collection of unstructured information becomes a network of linked data and software, necessary to perform the specialized on-the-fly processing of the raw data to provide ready-to-use geophysical parameters such as sea level and significant wave height.

The coastal currents in the Northern Indian Ocean are believed to play a prominent role in the heat and salt exchanges between Arabian Sea and Bay of Bengal. Hence it appears timely to carefully monitor their structure and variability at all timescales from intraseasonal to interannual. The objective of this study is to determine to what extent the sea surface height (SSH) variability associated to these coastal processes can be observed with satellite altimetry, in the Northern Indian ocean sector. Indeed, radar altimetry, which allows to measure SSH at centimetric accuracy, has been shown to be a powerful tool to obtain a wealth of information about open-ocean dynamics. Unfortunately, today, the use of standard satellite altimetric products in coastal zones remains challenging. This study explores a newly released coastal altimetric dataset, obtained from a complete reprocessing of the Topex/Poseidon data. We first present an objective method to derive geostrophic current from the raw SSH. Then we present the validation of the altimetric SSH against in situ observations. Finally we briefly analyze the observed variability of the East India Coastal Current at various timescales.

Possibility of predicting surface boundary layer winds over coastal land and ocean has been explored in this paper. Prediction has been effected using a modern nonlinear data-fitting algorithm known as genetic algorithm (GA) based on the Darwinian evolutionary theory. Time series of tower-mounted anemometer measured wind speed has been used for carrying out forecast over land while time series of satellite scatterometer derived winds has been used for forecast over coastal ocean. The prediction over land can feed into weather advisories required for rocket launching stations while prediction over coastal ocean can be of use in offshore industries.

Ship features in high-resolution spaceborne Synthetic Aperture Radar (SAR) imagery has crucial significance for ship classification from satellite. In this paper, we discuss the features of merchant Ships including oil tanker, container ship and bulk carrier in SAR imagery, which is comprised of geometrical feature, scattering feature, tonnage information with Radar Cross Section (RCS) and wake. The study show that the ship lengths measured from SAR imagery has a good correlation with the real lengths, but the correlation of ship beam is worse. Ship scattering feature has positive correlation with the ship structure, which maybe is a feature to distinguish container ship from other vessels. A new equation about ship length and its displacement in tons is presented in this paper. The relation suggested by Skolnik M.I between ship tonnage and RCS is tested but not validated. We also validate the means of extracting ship speed by ship turbulence wake in SAR imagery.

NOS/NOAA routinely generates HAB bulletins for the Southeast US coastal waters over the Gulf of Mexico and Florida basin using ocean color products from SeaWiFS and lately MODIS sensor on Aqua. From the consideration of life and safety, availability of products that can provide continuity in case the current sensors retire or fail before VIIRS becomes operational is very important. NOAA CoastWatch program is exploring the suitability of products from non-US satellite platforms as a possible option. We need to investigate inter-sensor product variation. We present results from such an evaluation study done with data from MERIS for the US coastal waters. A time series of the variability of color products available from the three sensors is presented. Comparisons are analyzed for different geographical areas and different optical ranges with a view to investigate systematic effects. We plan to extend this analysis to other sensors in future.

In the present study, Integrated Coastal Zone Management Plan (ICZMP) has been developed for Mangalore Coast in Karnataka, along the West Coast of India, by analyzing the remotely sensed data and conventional data. The various data products used in the present study includes, IRS-1C LISS-III+PAN and IRS-P6 LISS IV remotely sensed data, Naval Hydrographic Chart and Survey of India (SOI) toposheets. Different thematic maps prepared in the present study includes, land use/ land cover map, bathymetry map, shoreline configuration map, transportation and drainage network maps, GPS survey map, CRZ map, contour map, DEM, inundation map and coastal erosion vulnerability map. The results of the present study are encouraging. Some of the specific conclusions of the study are; eight coastal vulnerability sites have been identified, significant increase in the built-up area and decrease in the agricultural land, no large scale erosion or deposition in the vicinity of coastal structures such as seawalls, breakwaters and entrance channel of New Mangalore Port Trust and the beaches along the Mangalore Coast are maintaining dynamic equilibrium. To get the online information about all these, Coastal Zone Information System (CZIS) has been developed through V. B. 6. 0. using results of various data analyses.

The remote sensing method has many advantages such as large coverage, multi-temporal dynamic monitoring and historical monitoring backward. In this paper, remote sensing images from Landsat-5, Landsat-7 and CBERS-1 satellites are acquired for extracting shoreline, inning and tidal flat information. The characteristics of inning and inter-tidal flat are analyzed. Then, the spatio-temporal evolvement of inter-tidal flat under the influence of inning projects is presented. The work shows that the remote sensing technology is a valuable method for multi-temporal dynamic monitoring and historical backward monitoring for estuarine tidal flat, and that inter-tidal flat evolvement is influenced by inning projects mainly.

Coastal changes are dynamic in nature and periodic study of coastal areas is essential for the sensitive management and planning of the coastal environment. This paper attempts to study these changes occurring over the years due to natural and anthropogenic factors. Documentation and analysis of such changes is essential to develop sustainable environmental management plans for coastal zone. The study involves understanding the coastal environment, its geomorphology, changes in the land use patterns and the socio-economic factors influencing them. Accurate, comprehensive and real-time scientific data is essential to monitor such changes over decades, which is best made available through remote sensing. Hence, it has been used as an important tool for analysis. The area has been studied with the help of Survey of India maps and satellite imagery available for the period from 1973 to 2005. After an initial understanding of the entire coastline of Goa, a specific study area along the CRZ of Mandovi River was selected for detailed analysis.

Trichodesmium spp. is widely spread in the Arabian Sea. It form dense patches. During 2000-2005 (5 years period) extensive sampling was done in the Arabian Sea covering large area and different months starting from November to May. Three prominent sites are identified as Trichodesmium bloom sites in the Arabian Sea: 1) Lakshadweep waters 2) Off Goa and 3) Off Gujarat area. Bloom of around 100 km2 area with 4 to 400737 filamentsL-1 concentrations are recorded. Two species of Trichodesmium are encountered based upon seasonality and environmental conditions. OCM derived chlorophyll a during bloom was as high as 0.5 to 1 mgm-3, at time increased upto 5 mgm-3 and depends upon number of filaments in water. The Trichodesmium features were identified at 869, 670, 555nm in OCM data. Trichodesmium was detected as stripes and eddies in OCM images. The bloom patches appear darker which is taken as measurement of spread of the bloom in water. Total 133 stations are covered during 5 years period out of which 63 stations showed presence of Trichodesmium with discolouration of water. In offshore water Trichodesmium was detected as early as November and continued upto April month whereas in the coastal water Trichodesmium prevailed from February to May. The seasonality of these blooms was observed with respect to inshore/offshore and two species of Trichodesmium is discussed with the help of OCM data processed for chlorophyll a during Trichodesmium growth period, at 3 identified sites in the Arabian Sea.

Chlorophyll-a maps derived from IRS-P4 Ocean Color Monitor (OCM) was used to study the distribution pattern of phytoplankton biomass in the eastern Arabian sea off Karnataka-Goa coast, southwest coast of India. The data was compared with in-situ measurements of chlorophyll-a concentration estimated for 100 stations covering an area of more than 4000 km^2 in the above region. The presence of dense algal blooms spread over an area of almost 100 km^2 representing Trichodesmium sp. 18km off Kumta-Gokarna in the eastern Arabian Sea depicts high value (30 to 40 mg/m^3) of chlorophyll concentration. Similarly, around the Nethrani Island, off Bhatkal, the surface concentration ranged from 5 to 10 mg/m^3. The secchi depth varied from 4 to 8 m near the island. The sea surface area enveloping this high bloom (around the island) depicts a normal distribution of chlorophyll-a ranging from of 0.1 to 5 mg/m^3. It is suggested here, that the low salinity value (35 to 35.2 %) around the Nethrani Island enhances the algal bloom due to enrichment of nutrients in the shallow marine environment through probable inputs of nutrient charged fresh water from the island aquifers. Near river mouths, the values are marginally high in the range of 3 to 5 mg/m^3, probably enforced by riverine nutrient inputs well depicted by the Tadri River. The satellite (IRS-P4 OCM) derived images of chlorophyll during summer also shows high values as a band parallel to the coast. During the occurrence of algal blooms this band, parallel to the coast, widens offshore and this phenomenon of widening is typically absent during non-bloom summer scenarios, as identified for summer 2005.

Arabian Sea is highly influenced by monsoon systems like SW monsoon (June-September) and NE monsoon (December-February). This affects distribution pattern of phytoplankton, availability of nutrients and changing temperature specially during winter cooling period (February-March). These and other conditions like quality and quantity of light influence phytoplankton (type and concentration) in the Arabian Sea. In our study we have observed monsoon related peaks in phytoplankton and chlorophyll a in the Arabian Sea. These chlorophyll a (Chl a) concentrations detected by OCM sensor onboard IRS-P4 satellite is helping us to work out seasonality of phytoplankton in the Arabian Sea, which is of the great importance in the field of biology and biogeochemistry of this region. However, during these study subsurface chlorophyll a maxima (SCM) was observed as characteristics feature of chlorophyll a distribution in the Arabian Sea. The subsurface chlorophyll a maxima varies from 30m to 55m in the Arabian Sea during various seasons. During November at St. 1 surface chlorophyll a was 1.503 mgm^-3 and subsurface chlorophyll maxima was 12.692 mgm^-3. Similarly, at St. 13 surface chlorophyll a was 0.584 mgm^-3 and surface chlorophyll maxima was 8.517 mgm^-3. During upwelling, nutrients remained unused at sub surface due to shortage of light which may lead to subsurface blooms, detection of which is critical for precise estimation of chlorophyll a from ocean colour sensor. During our 5 year study (covering all the seasons) in northeastern (NE) Arabian Sea, we have observed more than 50 % stations were with subsurface chlorophyll a maxima where chlorophyll a was approximately 10 times higher compared to surface value. The high chlorophyll some time detected by OCM is mainly because of detection of subsurface chlorophyll maxima by the satellite but may not actually sampled during ship studies. The satellite penetration depth (ze), subsurface chlorophyll maxima depth, OCM derived chlorophyll a for different seasons in NE Arabian Sea is presented and discussed in this paper.

The Arabian Sea, situated in the western part of the northern Indian Ocean is a tropical basin. It is bounded on the east by the Indian peninsula, on the north by Baluchistan and Sindh provinces of Pakistan and on the west by the landmass of Arabia and Africa. The environmental factors that influence this tropical basin are the seasonally changing winds from the northeast during winter (November-February) and southwest during summer (June to September). Accordingly, the waters of the basin will experience seasonal variations. The study aims at understanding the seasonal and inter-annual variation of the Arabian Sea using satellite-derived data. The spatial domain selected for the present study is 40 degrees E and 78 degrees E longitude and equator to 30 degrees N. The remote sensing data with respect to sea surface temperature (SST), sea surface wind, sea surface height (SSH), and chlorophyll pigment concentration during January 2002 to December 2005 were used to understand the spatio-temporal variability of the Arabian Sea. The monthly mean SST data was obtained from Modis aqua, winds from Quikscat and chlorophyll pigment concentration from SeaWiFS. The SSH anomaly data was obtained from the merged product - Topex/Poseidon ERS 1/2 satellite which is 7-day snapshot. The spatial resolution of these data is 0.3 degrees latitude x 0.3 degrees longitude. Geographical information system (GIS) was used for processing and analysing the above parameters to determine the variability and detection of oceanic processes that are responsible for such variability.The study showed a very strong inverse correlation between SST and chlorophyll concentrations. Arabian Sea undergoes cooling during summer due to upwelling and advection, and in winter due to surface cooling under reduced solar heating. Upwelling along the coasts of Somalia, Arabia, and the west coast of India brings cold and nutrient rich sub-surface waters to the surface, which supports the observed high chlorophyll concentrations. During winter the convective mixing in the northern Arabian Sea supports high chlorophyll pigment concentrations. Due to strong solar heating, SST was warmest in spring (April), which supported least chlorophyll concentration.llite

The seasonal cycles of surface chlorophyll (SCHL) in the Indian Ocean (IO) are regionally described by means of 6 parameters: the timing of the bloom onset and of the bloom peak, and the integrated SCHL value in between these two extrema for both winter and summer blooms. This description, based on a climatology constructed from 7 years of SeaWiFS data, provides a regional image of the influence of the two monsoons on phytoplankton blooms. Over a large part of the basin, the seasonal cycle is characterized by two distinct growth periods, one in summer during the South West Monsoon (SWM), the other in winter during the North East Monsoon (NEM). However, in some specific areas such as the southwestern coast of India, there is no maxima during the NEM. The bloom areas during the SWM and the NEM show totally different regional patterns. Important lags in the timing of the blooms are identified, and are also associated with distinct regional patterns. The next step in the understanding of the SCHL seasonal cycles in the IO will be to relate the regional patterns of the timing and amplitude of the blooms with those of the physics of the IO.

Strong winds associated with hurricanes generate upwelling of cold water and transfer nutrients to the sea surface, supporting the development of significant phytoplankton blooms. Here we study the effect of the passage of hurricanes on the fields of sea surface temperature and chlorophyll-a in the pelagic ocean. A case-study is given for Hurricane Ignacio, the first storm of the 2003 season in the Eastern Tropical Pacific Ocean, on the south-eastern waters of Baja California Peninsula. The net reduction of in situ water temperature was -10° C and the phytoplankton pigment increase was 10 fold. Detailed features of the distribution of both characteristics of the sea surface are evident in the synoptic satellite imagery (MODIS/Aqua), with extreme thermal changes of -6 °C and increases up to 25 fold in chlorophyll-a. The satellite-derived averaged changes computed in the area of impact (~45,000 km^2) show temperature reduction of -1.3 °C and 1.5 fold increase of phytoplankton biomass. The physical and biological features studied for Hurricane Ignacio are crucial for understanding the ecosystem function around the southern Baja California peninsula, a region with strong dynamics in the carbon cycle. The study demonstrates how hurricanes induce phytoplankton blooms, a critical resource in the food chain, in particular for the pelagic fisheries. Systematic use of satellite remote-sensing may be advantageous to quantify at short, middle and long term, the impact of hurricanes on ocean biology at spatial and temporal scales of local and regional interest.

Observational data from the Arabian Sea Monsoon Experiment (ARMEX-Phase IIA) in the southeastern Arabian Sea (SEAS) showed intense warming with the SST up to 31.5°C during April-May 2005. Analysis of 5-day repeat cycles of temperature and salinity profiles from an ARGO float (ID No. 2900345) in a 3°x1° box closer to ARMEX-II buoy (8.3°N, 72.68°E) in the SEAS during January-September 2005 revealed evolution of warm pool (SST>28°C) in spring 2005. The Argo data derived D20 (depth of 20°C isotherm) showed the influence of remote forcing during January-May, and local wind forcing during southwest monsoon. Low salinity waters (<34.0) occupied the top 30 m during January-February followed by temperature inversions (up to 0.5°C) in the 30-60 m depth range. From the peak spring warming, the SST dropped gradually by 3.5°C by end-July with the advent of southwest monsoon followed by a decrease in net heat gain upto 100 W/m^2. The merged weekly products of sea surface height anomalies and the NLOM simulated surface currents showed complex surface circulation consisting of seasonal Lakshadweep High/Low in winter/summer. The examined oceanic and atmospheric variables showed an intraseasonal variability with 41 to 63 day period, coinciding with the Madden-Julian Oscillation.

The merged ERS-1/2, TOPEX/Poseidon and Jason-1 altimeter weekly sea level anomalies (SLAs) for the period 1997-2005 were analyzed to study the variability of sea level and computed geostrophic currents in relation to the equatorial jets in the Indian Ocean. Year-to-year-variability in SLA was large (small) at 77°E (93°E) with a pronounced (weaker) semi-annual variation. The computed geostrophic currents, using the second order differential equation, in the equatorial region (2°S-2°N) revealed the existence of spring and fall equatorial jets in each year, but with considerable variation. The 9 year mean (excluding 1997-98) SLA and the derived currents revealed that spring (fall) jet was weak (strong) with higher (lesser) magnitudes of SLA in the eastern basin. The computed surface currents agreed well with the Acoustic Doppler Current Profiler (ADCP) measured currents and OGCM for Earth Simulator (OFES) simulated currents. The interannual variability of the equatorial jets revealed that the fall jet extended for longer duration than the spring jet. The equatorial Kelvin wave with a wavelength of ~1100 km was identified. Westward propagating Rossby wave was identified at 5° N more clearly. The zonal currents at 5° N (south of Sri Lanka) revealed intense anticyclonic (cyclonic) eddy activity between 75° and 80° E during summer (winter) monsoon.

Light absorption, chlorophyll-specific absorption and package effect of phytoplankton in Zhujiang River mouth was studied on the two cruises of "Shiyan 2". The vertical distribution of phytoplankton in waters can be simulated by Gaussian equation; light absorption of phytoplankton was made on 25 mm GF/F filter, maximum light absorption of phytoplankton was at the depth 5-10m underwater. The value of chlorophyll-specific absorption was negatively correlated with chlorophyll a concentration. Package effect index also had a negative relationship with chlorophyll concentration. These results indicated that the characteristics of light absorption by phytoplankton varied in different waters were properly related to the phytoplankton concentration.

In this paper, the light attenuation, fluorescence and absorbance of DOM in Lake TaiHu was mainly studied from July 2001 to December 2002. Results showed that DOM mainly contributed to the attenuation at the ultraviolet, and the best correlation coefficient was at 355 nm. The absorption of DOM could be well described by the exponential decay model. The slope value S, retrieved from absorbance of 350 nm and 440 nm, was 0.0046 nm^-1(SD=0.0015,n=284), far less than the published 0.014 nm^-1. The absorbance at 350 nm (a(350)) well correlated with DOM concentration (R^2=0.673, n=284), and also linearly correlated with fluorescence at 430 nm (R^2=0.497, n=214). The relationship between DOM concentration and fluorescence intensity was also regressed, and showed that it was polynomial.

The current exact Rayleigh scattering calculation of ocean color remote sensing uses the look-up-table (LUT), which is created for a special remote sensor and cannot be applied to other sensors. For practical application, a general Rayleigh scattering LUT which can be applied to all ocean color remote sensors is produced. The vector radiative transfer equation (VRTE) in a plane-parallel atmosphere with the adding-doubling method is deduced and solved in detail. Comparing with the exact Rayleigh scattering radiance derived from MODIS exact Rayleigh scattering LUT, it is proved that the relative error of Rayleigh scattering calculation with the adding-doubling method is less than 0.25%, which meets the required accuracy of atmospheric correction of ocean color remote sensing. Therefore, the adding-doubling method can be used to generate the exact Rayleigh scattering LUT for ocean color remote sensors. Finally, with the adding-doubling method, the general exact Rayleigh scattering LUT is generated. On the basis of this general LUT, the calculated Rayleigh scattering radiance is tested by comparing with LUTs of MODIS, SeaWiFS and the other ocean color sensor, showing that relative errors are all less than 0.5%, and this general LUT can be applied to all ocean color remote sensors.

It is necessary to perform atmospheric correction for extracting quantitative information from Moderate Resolution Imaging Spectroradimeter (MODIS). In this paper, a feasible method which atmospheric correction of MODIS image was performed by using meteorological data was discussed. The reflectance of two bands was obtained from MODIS and the approach based on Dark Object Subtraction (DOS) method was used to calculate the path radiance with the reasonable dark-object reflectance choice. The atmospheric transmittance can be derived from the atmospheric parameters at overpass time of satellite by using interpolation among local meteorological records. At last the reflectance of objects was obtained. This approach was evaluated by comparing the normalized difference vegetation index (NDVI) in close date before and after atmospheric correction of MODIS, and by using single day MOD09. Results showed that NDVI by using this approach is close to truly data. It will be feasible to use meteorological visibility and vapor pressure data to perform atmospheric correction of MODIS data.

Spaceborne radar scatterometers operating in microwave frequency bands have several science and operational applications in Oceanography, meteorology, agricultural and geophysical sciences. The basic parameter measured by a scatterometer is the Backscattering coefficient (&psigma;^0) for a certain frequency, polarization and observational geometry. Before addressing a specific application, it is needed that the &psigma;^0 signatures be analyzed over natural, undisturbed and uniform/quasi-uniform target areas. As a prelude to ISRO's forthcoming OceanSat-II mission, carrying a Ku-band scatterometer, QuikSCAT scatterometer measured &psigma;^0 data are analyzed over its dynamical range using the global natural targets.

This study aims at discovering changes of ecosystem service value based on spatial-temporal variety of land use features on the coastal area in Lianyungang City from 1978 to 2000. Three LANDSAT images (1978 MSS, 1987 TM and 2000 ETM) were used to estimate changes in the size of six land-use categories (forest, cropland, urban, wetland, lakes/rivers and coastal water). Coefficients which integrated from Costanza et al.'s (1997) ecosystem services valuation model and Xie Gaodi et al.'s (2003) "ecosystem services value unit area of Chinese terrestrial ecosystems" were used to estimate changes in ecosystem services delivered by each land categories, and a sensitivity analysis was conducted to determine the affecting factors of ecosystem service value. The total annual ecosystem service values in coastal area of Lianyungang City decreased from RMB 8.23×10^9 in 1978 to RMB 7.53×10^9 in 1987, however it increased from RMB 7.53×10^9 in 1987 to RMB 10.73×10^9 in 2000. The decline of ecosystem service value from 1978 to 1987 is largely attributable to 53.49% loss of forest at the same period, while the increase of service value is contributed by 147.70% increase of wetland from 1987 to 2000. Change of wetland and forest has the major contribution to the change of total ecosystem service value of Liangyungang city. Waste treatment, water supply and climate regulation are the dominant ecosystem services of this area. Forest and wetland should be strictly conserved in further urban development.

Using Landsat-5 TM image on November 24, 2004 and May 13, 2005; and quasi-synchronous surveyed data, the dynamic change of suspended silt and chlorophyll a in the Yellow River Estuary were inversed and the coastal line of this region was interpreted. The mean correlation coefficient between measured data and spectral signature value of every image is 0.85 which proved that the inversion is satisfying. The results show that there were different seasonal patterns between suspended silt and chlorophyll a. The concentration of suspended silt in the surface seawater in the estuary averaged 307.8mg/L in winter and 352.7mg/L in spring. While the Chlorophyll a concentration in the surface seawater averaged 0.71μg/L in winter and 0.59μg/L in spring. The surveyed data show one distinct boundary in concentration of suspended silt along the sampled stations. The reversed data from band1, band2, band3, band4 and band6 also show the same boundary at the same location. The results of interpreting coastal line show that there is a rapid alteration in the Yellow River Estuary coastal line. Although in low or median water level period, with advance extent of the coastal line, there were prominent on the south shore beside the Laizhou Bay than the other places. Integrated with the practical measured data, by means of RS method, the study can get the same conclusion as the other methods did on judging boundary sign between River and Sea. So it also can be regarded as an efficient method to confirm boundary of River and Sea, and be the most available way to study dynamic state of coastal line of river mouth.

Sequence of the images from IRS P4 / OCM satellite and extensive shipboard sampling programme are used to understand the seasonal variation of phytoplankton abundance and types in the Northeastern (NE) Arabian Sea and Lakshadweep Sea. An appreciable degree of spatial and temporal variability is observed in chlorophyll a distribution from November to April months, as well as coastal and offshore stations, indicating marked seasonality in phytoplankton distribution in NE Arabian Sea. During November month (fall intermonsoon) average chlorophyll a (Chl a) by fluorometer was (0.799 mgm^-3) and by OCM it was 0.584 mgm^-3. The higher chlorophyll a observed was due to Trichodesmium (cyanobacteria) blooms. During December the average chlorophyll a was 0.34 mgm^-3 also due to Trichodesmium filaments in water column. During January onwards winter cooling led to increase in nutrients which enhanced chlorophyll a value to 0.64 mgm^-3 due to growth of flagellates (as seen by high chlorophyll b besides chlorophyll a) in water column. February, March and April supported moderately high chlorophyll value (0. 3 to 0.5 mgm^-3) due to growth of prasinophytes (as seen by pigment prasinoxanthin) and blooms of the Noctiluca miliaris. Time series monitoring of Noctiluca bloom was also conducted using OCM based chlorophyll images in NE Arabian Sea. During February chlorophyll a retrieved by OCM was 0.3 to 0.9 mgm^-3. Pigment analysis of water samples indicated the equal important of accessory pigment like zeaxanthin, prasinoxanthin, beta-carotene. The relevance of these pigments estimated by HPLC like zeaxanthin (cyanobacteria), fucoxanthin (diatoms), peridinin (dinoflagellates) is presented and discussed. Similarly, exercise is conducted in Lakshadweep waters where Trichodesmium related peak in chlorophyll a was observed during March onwards in OCM data. The average chlorophyll a in NE Arabian Sea at surface during November was (0.726 mgm^-3), December (0.34 mgm^-3), January (0.723 mgm^-3), February (0.344 mgm^-3), March (0.963 mgm^-3) and April 0.665 mgm^-3. Similar trend was observed in primary productivity estimates. The attempt is made to work out seasonality in the productivity of the Arabian Sea using OCM derived chlorophyll and relation of enhancement in productivity due to development of winter blooms in the Arabian Sea. The environmental conditions (temperature, wind, nutrients and mixed layer depth) affecting these blooms responsible for year to year variation in bloom biomass and productivity is also presented in detail.

The HYDROLIGHT ocean radiative transfer code is a well established standard for simulating the water leaving radiance from the visible to the near infrared. However, its efficient use is a task not easily accomplished for running thousands of different cases and processing the generated results. HydroPPT is a tool integrated with HYDROLIGHT to generate input data and Fortran files and process the digital output files in batch mode. This paper describes workflows and examples for simulating a dataset. This paper also provides some tips to port HYDROLIGHT to some other compilers (e.g., Intel visual Fortran compiler) and platforms (e.g., SGI IRIX, HP TRU64, SUN Solaris, etc) to improve the computing efficiency and effectively use the computing resources. Based on our tests, Intel visual Fortran compiler provides improved performance for HYDROLIGHT code than that by the LAHEY Fortran compiler on the Intel processor based platforms.

In this study, measurements and models are used to test the closure between remote-sensing reflectance and IOPs. Measurements include those by AC9 (Wetlabs, Inc.) and HS6 (HOBI labs, Inc.), while models include both empirical models (e.g., Voss' beam attenuation coefficient model) and radiative transfer model (e.g., Hydrolight). It is found that, generally, AC9 works better than HS6 in providing scattering and backscattering coefficients. HS6 need more accurate calibration; absorption coefficients by AC9 are consistent with those by Spectrix or Spectrometer. Good linear relationship is found between AC9 measured beam attenuation coefficients (c) and the Voss model; while those measurements by HS6 needs some adjustments before feeding to HYDROLIGHT.

Red tides occurred frequently in the East China Sea in recent years. A red tide dominated by Prorocendrum donghaiense and Karenia mikimotoi happened in the Zhejiang Coastal Waters in late-May 2005. EOS MODIS water color data were utilized in extracting the red tide information. Sea surface temperature data derived from NOAA AVHRR and EOS MODIS were also analyzed to understand the possible formation mechanism of this red tide. The relationships between the red tide and related oceanographic features were discussed based on image data. The results indicate that the red tide was associated with several oceanographic processes, such as coastal front. This study showed that the combination of remote sensing data of water color and sea surface temperature can be useful in studying and understanding the oceanography of red tides.

Geostationary Ocean Color Imager (GOCI) onboard its Communication Ocean and Meteorological Satellite (COMS) is scheduled for launch in 2008. GOCI includes the eight visible-to-near-infrared (NIR) bands, 0.5km pixel resolution, and a coverage region of 2500 x 2500km centered at 36N and 130E. GOCI has had the scope of its objectives broadened to understand the role of the oceans and ocean productivity in the climate system, biogeochemical variables, geological and biological response to physical dynamics and to detect and monitor toxic algal blooms of notable extension through observations of ocean color. To achieve these mission objectives, it is necessary to develop an atmospheric correction technique which is capable of delivering geophysical products, particularly for highly turbid coastal regions that are often dominated by strongly absorbing aerosols from the adjacent continental/desert areas. In this paper, we present a more realistic and cost-effective atmospheric correction method which takes into account the contribution of NIR radiances and include specialized models for strongly absorbing aerosols. This method was tested extensively on SeaWiFS ocean color imagery acquired over the Northwest Pacific waters. While the standard SeaWiFS atmospheric correction algorithm showed a pronounced overcorrection in the violet/blue or a complete failure in the presence of strongly absorbing aerosols (Asian dust or Yellow dust) over these regions, the new method was able to retrieve the water-leaving radiance and chlorophyll concentrations that were consistent with the in-situ observations. Such comparison demonstrated the efficiency of the new method in terms of removing the effects of highly absorbing aerosols and improving the accuracy of water-leaving radiance and chlorophyll retrievals with SeaWiFS imagery.

Secchi depth provides the oceanographer with the first hand information about transparency and penetration of light in the water. Here we present results of the Secchi depth and the optical properties measured in the Arabian Sea. Our analyses show spatial and temporal variability of Secchi depth and their dependence on the optical properties beam attenuation and diffuse attenuation the biological parameter of Chlorophyll. The in-situ measured inherent and apparent optical properties have been used to understand the underwater light properties and their relations to the Secchi depth in various water types. The Secchi depth model is validated using the measured optical properties. We also present an empirical method to determine Secchi depth from the satellite ocean color sensor, and the application of the same to the IRS-P4 OCM is found to provide comparable results to the measured values.

The remote sensing constitutes a vast field of study whose repercussions are many and varied on environmental management. The phenomenon of dust clouds is a major climatic event in Africa. But the observation means of this phenomenon are still too much limited. The development of an approach consisting in the detection of dust clouds from satellite images can be a solution. In this work, we present a new approach for dust clouds detection in the infrared images coming from the METEOSAT satellite. It is then proved necessary of finding automatic or semi-automatic analysis methods to assist their detection and interpretation. Thus we are interested in image fusion methods for detection structures in the images. In this paper, we present some statistical methods which enable to extract texture features from the images. Then, we describe the method used for selection the best attributes for the images segmentation into three classes: "water clouds", "ocean" and "continent". We then use a method which enable us to segment the class "continent" of the image for dust clouds detection. Finally, we compare our results with another one which shows the zone of presence or absence of dust clouds. This comparison shows that we are in concord because visually, we have a good analogy of shape on the dust clouds zone as well as on the part without dust clouds.

A methodology is proposed to correct satellite ocean-color imagery for the perturbing effects of the atmosphere and surface progressively, starting from the near infrared and advancing to the visible. First, a set of spectral bands is selected in the near infrared, for which the water body can be considered black, except in one of the spectral bands. The top-of-atmosphere reflectance in the selected bands, after correction for molecular scattering and sun glint contributions, is linearly combined to retrieve the ocean signal in the spectral band where the water body is not black. The coefficients of the linear combination minimize the perturbing effects, which are due to scattering and absorption by aerosols, and reflection by the surface. These effects are decomposed into principal components in the modeling. Second, other sets of spectral bands are selected, that progressively include shorter wavelengths. At each step, only the marine signal in one spectral band is unknown and therefore estimated. The methodology is developed for the Sea-viewing Wide Field-of-view Sensor (SeaWiFS), but is generally applicable to ocean-color sensors that measure in the visible and near infrared. Without measurements above one micrometer, however, the atmospheric correction is only accurate over Case-1 waters. Theoretical performance is evaluated from radiation-transfer simulations for a wide range of geophysical and angular conditions, including absorbing aerosols. Only Case-1 waters are considered in the simulations. The perturbing influence of the atmosphere and surface is minimized adequately for each set of wavelengths, except when the aerosol loading is large. The residual effects in the linear combination exhibit a bias of magnitude increasing with aerosol optical thickness. The bias can be reduced globally, by taking into account all the eigenvectors of the decomposition in principal components, not only the most significant ones. Errors in the estimated marine signal increase with decreasing wavelength (the residual effects at longer wavelengths propagate) and with increasing aerosol optical thickness. They become unacceptable when the aerosol optical thickness at 550 nm is above 0.3. Performance can be improved by optimizing the sets of selected wavelengths, or by using aerosol optical thickness estimated from the satellite data.