Mirosław Darecki

Diffuse Attenuation Coefficient of Downwelling Irradiance: An Evaluation of Remote Sensing Methods

coastal waters, with Kd(490) ranging from � 0.04 to 4.0 m � 1 . The derived values are compared with the data calculated from in situ measurements of the vertical profiles of downwelling irradiance. The comparisons show that the two standard methods produced satisfactory estimates of Kd(l) in oceanic waters where attenuation is relatively low but resulted in significant errors in coastal waters. The newly developed semianalytical method appears to have no such limitation as it performed well for both oceanic and coastal waters. For all data in this study the average of absolute percentage difference between the in situ measured and the semianalytically derived Kd is � 14% for l = 490 nm and � 11% for l = 443 nm.

Optical characteristics of two contrasting Case 2 waters and their influence on remote sensing algorithms

This paper describes the results of measurements of inherent and apparent optical properties of two contrasting Case 2 waters (Southern Baltic and off the west coast of Ireland). The experiments were carried out over two seasons, both before the peak of the phytoplankton growth period, and during the bloom events. The first study was made when the concentration of chlorophyll (phytoplankton pigment) was similar in both waters and in the range of 0.6–3.23 mg m � 3 . The second study was made when concentrations of chlorophyll were very high, reaching 14 mg m � 3 in the west coast of Ireland and 70 mg m � 3 in Southern Baltic. Optical measurements of surface reflectance were made with a profiling spectroradiometer. The spectral shapes of the particulate absorption spectra and the diffuse attenuation coefficient in both regions and seasons are compared. The contribution of detritus and coloured dissolved organic matter absorption to the apparent optical properties of water have impact on the accuracy of remote sensing retrieval algorithm for chlorophyll a. Differences have been found between algorithms based on conventional spectral bands (e.g. SeaWiFS) and proposed new spectral channels. The most accurate ratio for chlorophyll retrieval (490/550) was for the western Irish shelf (R 2 =76%, and the standard error of the estimate ranged from 30% to 37%), with poor results for this ratio in the Baltic. The Rrs (550/590) ratio gave better results for the Baltic (R 2 =75%, and the standard error of the estimate ranged from 20% to 55%), with poor results for the Irish shelf. The results show that a unique combination of spectral bands needs to be applied in Baltic waters to achieve an acceptable accuracy of the in-water remote sensing algorithm whereas the more commonly accepted band ratios were acceptable for the Irish Shelf. This suggests that for accurate determination of chlorophyll from satellite-borne sensors in different coastal waters a wider choice of spectral bands is needed. r 2003 Elsevier Science Ltd. All rights reserved.

Empirical relationships between coloured dissolved organic matter (CDOM) absorption and apparent optical properties in Baltic Sea waters

Coloured dissolved organic matter (CDOM), also know as yellow substances, dissolved in Baltic Sea waters is a dominant constituent, absorbing light in the blue and green parts of the spectrum. The relative share of CDOM absorption in the total absorption of blue light can reach 80%. This causes a high correlation between the CDOM absorption coefficient a y(400) and irradiance diffuse attenuation coefficient K d(412). The seasonal variations of CDOM and particulate absorption have a combined effect on the magnitude and shape of the irradiance diffuse attenuation coefficient spectra, which is represented by a shift of the maximum transmission wavelength towards the red part of the spectrum. Such an effect has been observed in the Gulf of Gdansk during recent experiments. CDOM absorption also influences the shape of remote sensing reflectance spectra. Appropriate spectral bands for the construction of a local, empirical, remote sensing algorithm for quantification of CDOM absorption are proposed, based on the seasonal variability of remote sensing reflectance spectra. The empirical relationships between CDOM absorption and downwelling irradiance diffuse attenuation coefficient and spectral reflectance are presented and their potential for quantifying CDOM absorption is explored. The CDOM absorption coefficient is highly correlated with downwelling irradiance diffuse attenuation coefficient in the blue. The estimation of CDOM absorption using the downwelling irradiance measurements can be achieved with low errors: −4.6% systematic and 25.3% random. The estimation of CDOM absorption from empirical relationships between absorption and remote sensing reflectance suffers lower accuracy: 4% for systematic and 32% for random errors of estimates.

SeaWiFS ocean colour chlorophyll algorithms for the southern Baltic Sea

Simultaneous measurements of the remote sensing reflectance and chlorophyll‐a concentration at nearly 700 stations in the Baltic Sea were collected in the years 1993–2001. The collected dataset was used for in situ validation of the standard Sea‐viewing Wide Field of view Sensor (SeaWiFS) algorithms. The results confirm a large discrepancy between the observed chlorophyll‐a concentrations and the concentrations calculated from the current SeaWiFS operational algorithm equations. New forms are proposed as chlorophyll algorithms for the SeaWiFS sensors or sensors with similar spectral bands. These forms include new coefficients and lower order of equations. As a two‐band algorithm, a simple polynominal log‐log equation is proposed: log(Chla) = −0.141–2.865R, where R is a band ratio equal to that of SeaWiFS OC2 algorithm. The four‐band algorithm takes the form: log(Chla) = −0.028–3.9015R, where R is a band ratio equal to that of SeaWiFS OC4 algorithm. These new forms allow improving accuracy for chlorophyll retrieval in the Baltic. Further, a considerable improvement in the accuracy of the chlorophyll concentration estimated from remote sensing algorithms required the use of a new set of the spectral bands in the equations. An algorithm with such a new set of two spectral bands has the form: log(Chla) = 0.738–4.681R, where R = log[R rs(550)/R rs(590)]. The accuracy of this algorithm is much better than the previously mentioned algorithms and the authors propose application of the new spectral bands in future satellite sensors.

Modeling of wave-induced irradiance fluctuations at near-surface depths in the ocean: a comparison with measurements

We develop a computationally fast radiative transfer model for simulating the fluctuations of the underwater downwelling irradiance E(d) at near-surface depths, which occur due to focusing of sunlight by wind-driven surface waves. The model is based on the hybrid matrix operator-Monte Carlo method, which was specifically designed for simulating radiative transfer in a coupled atmosphere-surface-ocean system involving a dynamic ocean surface. In the current version of the model, we use a simplified description of surface waves, which accounts for surface slope statistics, but not surface wave elevation, as a direct source of underwater light fluctuations. We compare the model results with measurements made in the Santa Barbara Channel. The model-simulated and measured time series of E(d)(t) show remarkable similarity. Major features of the probability distribution of instantaneous irradiance, the frequency content of irradiance fluctuations, and the statistical properties of light flashes produced by wave focusing are also generally consistent between the model simulations and measurements for a few near-surface depths and light wavelengths examined. Despite the simplification in the representation of surface waves, this model provides a reasonable first-order approximation to modeling the wave focusing effects at near-surface depths, which require high temporal and spatial resolution (of the order of 1 ms and 1 mm, respectively) to be adequately resolved.

Immersion factors for the RAMSES series of hyper-spectral underwater radiometers

The immersion factors If of sample RAMSES radiance and irradiance hyper-spectral radiometers were experimentally determined in the 400?700?nm spectral range using state of the art protocols, with the aim of supporting the use of these sensors in accurate radiometric measurements for ocean colour applications. Values of If obtained from the characterization of RAMSES-ACC-VIS irradiance sensors showed a dispersion of 2% in the blue increasing to 4% in the red spectral region. Results obtained from the characterization of RAMSES-MRC-VIS radiance sensors having a 20? in-air full-angle field of view highlighted an average spectral bias of ?2.4% in the theoretical If values determined just using the refractive indices of the intervening medium and of the optical window. This average spectral bias exhibited values of ?0.8% for the RAMSES-ARC-VIS radiance sensors having a 7? in-air full-angle field of view. Using these results, reference If values have been proposed for the RAMSES series of hyper-spectral radiometers.

Visual prey availability and distribution of foraging little auks (Alle alle) in the shelf waters of West Spitsbergen

As diving seabirds use vision underwater, it is presumed they should preferentially select sites where their preferred food items are not only abundant but also clearly visible. To test this, we studied the optical properties of the seawater in the West Spitsbergen Shelf, in combination with zooplankton abundance in the feeding grounds of the planktivorous little auks from the nearby colonies in Hornsund. We estimated the relative attractiveness of the foraging sites using a novel parameter—visual prey availability (VPAv), which relates density and proportion of the preferred food item (Calanus glacialis) of the little auk, in total zooplankton, to the optical properties of the seawater. We found a significant positive correlation between the density of foraging little auks and VPAv values. Birds chose areas where C. glacialis was both abundant and clearly visible, because of the clarity of the water and low proportion of other zooplankton species. The birds avoided foraging over the warmer Atlantic-type waters, characterised by a high abundance of zooplankton taxa mostly ignored by birds and where VPAv values were low. VPAv values could potentially also be applied to other visual planktivores for which prey preference and visual acuity are known.

Empirical Model for Phycocyanin Concentration Estimation as an Indicator of Cyanobacterial Bloom in the Optically Complex Coastal Waters of the Baltic Sea

Commonly used parameters to assess cyanobacteria blooms are chlorophyll a concentration and cyanobacterial cell counts. Chlorophyll a is contained in all phytoplankton groups and therefore it is not a good estimator when only detection of cyanobacteria is desired. Moreover, laboratory determination of cyanobacterial cell counts is difficult and it requires a well-trained specialist. Instead of that, cyanobacterial blooms can be assessed using phycocyanin, a marker pigment for cyanobacteria, which shows a strong correlation with the biomass of cyanobacteria. The objective of this research is to develop a simple, remote sensing reflectance-based spectral band ratio model for the estimation of phycocyanin concentration, optimized for the waters of the Baltic Sea. The study was performed using hyperspectral remote sensing reflectance data and reference pigment concentration obtained in the optically complex coastal waters of the Baltic Sea, where cyanobacteria bloom occur regularly every summer, often causing severe damages. The presented two-band model shows good estimation results, with root-mean-square error (RMSE) 0.26 and determination coefficient (R2) 0.73. Moreover, no correlation with chlorophyll a concentration is observed, which makes it accurate in predicting cyanobacterial abundance in the presence of other chlorophyll-containing phytoplankton groups as well as for the waters with high colored dissolved organic matter (CDOM) concentration. The developed model was also adapted to spectral bands of the recently launched Sentinel-3 Ocean and Land Color Imager (OLCI) radiometer, and the estimation accuracy was comparable (RMSE = 0.28 and R2 = 0.69). The presented model allows frequent, large-scale monitoring of cyanobacteria biomass and it can be an effective tool for the monitoring and management of coastal regions.

An attempt to use measurements of sun-induced chlorophyll fluorescence to estimate chlorophyll a concentration in the Baltic Sea

The possibility of using natural fluorescence -- sun-induced chlorophyll a fluorescence (SICF) in remote sensing and/or contact investigations of chlorophyll a concentration (Ca) has often been discussed in recent years. The relationship between chlorophyll a concentration (Ca) and chlorophyll a fluorescence at 683 nm seems to evident and suggests a method of obtaining the Ca concentration on the basis of SICF intensity. Unfortunately this relationship is not as direct as we would like because SICF depends on numerous physiological properties of phytoplankton like fluorescence quantum yield and light absorption capacity which depend on an environmental factors. The influence of these parameters on SICF is shown in this paper and the relationships are presented. Using these we made attempt to achieve the direct functional dependence between the natural fluorescence (obtained from the upward reflectance coefficient spectrum measured at 665, 683 and 710 nm) and the chlorophyll a concentration. The analysis carried out in this study shows that the possibility determining Ca concentrations is limited to great optical depths in the sea ((tau) greater than 3). However using SICF for estimation Ca in the surface layer is probably impossible.

Statistical relationships between photosynthesis and abiotic conditions in the ocean: the IOPAS initial model for remote sensing applications

A model of marine primary production with a set of statistical relationships linking physiological parameter of the photoplankton with abiotic factors of the sea has been developed. The study is based on empirical data analyzed from about 3,500 stations of various ocean regions. The data comes from Polish and Russian expeditions as well as from literature.