Two regional bio-optical algorithms are combined to retrieve the Chlorophyll-a($\text {Chl-}\textit {a}$) concentration in the Black Sea. The first is a band-ratio algorithm that computes$\text {Chl-}\textit {a}$as a function of the slope of Remote Sensing Reflectance ($\mathit {R_{\text {RS}}}$) values at two wavelengths using a polynomial regression that captures the overall data trend, enhancing extrapolation results. The second algorithm is a Multilayer Perceptron neural net based on$\mathit {R_{\text {RS}}}$values at three individual wavelengths that features interpolation capabilities helpful to fit data non-linearities. A new merging scheme is then designed to benefit from the complementarity of the two approaches. Remote sensing data employed to demonstrate the merging of regional results for the Black Sea are those acquired by the Ocean and Land Color Instrument on board Sentinel-3A to acknowledge the need for data products of higher accuracy within the long-term Copernicus program.

Summary This paper presents a high‐performance computing (HPC) framework for Monte Carlo (MC) simulations in the ocean color (OC) application domain. The objective is to optimize a parallel MC radiative transfer code named MOX, developed by the authors to create a virtual marine environment for investigating the quality of OC data products derived fromin situmeasurements of in‐water radiometric quantities. A consolidated set of solutions for performance modeling, prediction, and optimization is implemented to enhance the efficiency of MC OC simulations on HPC run‐time infrastructures. HPC, machine learning, and adaptive computing techniques are applied taking into account a clear separation and systematic treatment of accuracy and precision requirements for large‐scale MC OC simulations. The added value of the work is the integration of computational methods and tools for MC OC simulations in the form of an HPC‐oriented problem‐solving environment specifically tailored to investigate data acquisition and reduction methods for OC field measurements. Study results highlight the benefit of close collaboration between HPC and application domain researchers to improve the efficiency and flexibility of computer simulations in the marine optics application domain. © 2016 The Authors.Concurrency and Computation: Practice and ExperiencePublished by John Wiley & Sons, Ltd.

This study investigates standard and regional algal pigment index! (API!) estimates in the Atlantic off the southwestern Iberian Peninsula. Standard API! data are those delivered by the Medium Resolution Imaging Spectrometer (MERIS) orbiting sensor. Equivalent quantities are computed by applying a regional inversion scheme using as input both MERIS and in situ remote sensing reflectances (Rrs). Reference data for the development of the regional algorithm and for the analysis of tested products include field measurements of total concentration of chlorophyll a (TChla) and coincident Rrsvalues collected at different distances from the coast. Validation results, based on matchup analysis, identifies a systematic overestimation of standard API1 versus the reference TChla values. The additional comparison of product maps in selected regions of interest confirms this tendency and demonstrates the feasibility and relevance of using regional algorithms for investigating spaceborne products. Analogous applications are hence devised for the early-stage evaluation of the forthcoming Sentinel-3/OLCI data products.

This study investigates how the number of data collected for algorithm development in typical case-1 waters and optically complex cases can affect the remote sensing of ocean color (OC) products. Marine conditions dominated by the colored dissolved organic matter (CDOM) and nonalgal particles are considered. The applied OC inversion schemes are based on multilayer perceptron (MLP) neural nets for data classification and regression. Simulated data for MLP training are generated with a forward OC model. Results show that a disproportion of samples representing different marine optical cases influences the MLP learning and hence also the data product retrieval, mostly in mixed case-1 and CDOM-dominated environments. A postclassification correction is then employed for performance improvements. Methodological developments are presented, acknowledging the coastal water monitoring prioritized by the Copernicus Earth Observation program.

This letter discusses the normalized water-leaving reflectance obtained from two atmospheric corrections built in the MEGS 8.0 processor applied to the Medium Resolution Imaging Spectrometer (MERIS) marine data. In situ reference data for intercomparisons with satellite-derived reflectance products are from the Ocean Color component of the Aerosol Robotic Network (AERONET-OC). Case of study is the northern Adriatic Sea AERONET-OC site characterized by both Case-1 and Case-2 waters. The accuracy of MERIS pigment indices is also discussed in relation to uncertainties and biases affecting atmospherically corrected data.

Advances in satellite ocean color technologies and methodologies are expected to lead to the generation of coastal water bio-optical products with accuracies close to those targeted for oceanic regions. In view of contributing to such a progress, multilayer perceptron neural networks complying with standard Medium Resolution Imaging Spectrometer (MERIS) pigment indices were developed relying on regional highly accurate in situ data. This work illustrates and discusses the application to sample MERIS imagery of those neural networks trained to produce pigment indices in seas characterized by increased levels of bio-optical complexity: the Baltic, the Northern Adriatic, and the Western Black Seas.

Statistical performance tuning of a parallel Monte Carlo (MC) radiative transfer code for ocean color (OC) applications is presented. A low observed-to-peak performance ratio due to highly sparse computations is compensated by online and offline tuning techniques based on a statistical indicator of products accuracy. Run-time adaptive control employs the accuracy indicator to set up two complementary tuning criteria: one general to MC computations and the other specific to OC applications. The same accuracy indicator is also used for pre-execution tuning of a threshold parameter. Numerical simulations of real case scenarios showed that the proposed methods consistently led to faster runs, while satisfying application accuracy requirements. Specifically, speed-ups range from 2.17 to 7.44 times when compared with the un-optimized version of the MC code. The applied techniques are orthogonal to parallelization, so that the reported performance gains are further amplified by parallel speed-ups.

This letter, addressed to the analysis of remote sensing (RS) images of the sea-surface temperature (SST) off the Portugal coast, presents a novel approach to automatically detect and characterizemesoscale eddy-related structures. The complexity of this task is due to the dynamics of the investigated region, where upwelling currents and bathymetry effects produce countless and highly heterogeneous SST patterns, features of interest may have smooth boundaries, and edges associated to strong temperature gradients may not correspond to any eddy. All these limit the effectiveness of an image processing based on edge features (which can be successfully applied to automatically detect eddies in other oceanographic areas, for instance, close to the Gulf Stream). The proposed scheme exploits the iso-SST patterns associated to the eddy-related structure to code with a rule-based definition the process that allows for their visual identification. In practice, this enables revealing various morphological parameters of the eddy-related structure (i.e., the location, scale, symmetry, and rotation) and supports the exploitation of SST data allowing for annotating the RS image and benchmarking the subjectivity of the visual survey.

The accurate merging of primary radiometric ocean color products such as the normalized water-leaving radiance requires combining data from various space missions, which may be affected by different uncertainties as resulting from absolute calibration and minimization of the atmospheric effects. A statisticalcorrectionschemebased on a multilinear regression algorithm is used here to remove systematic differences betweeninsituand remote-sensing measurements. The application of thecorrectionschemeto Sea-viewing Wide Field-of-View Sensor (SeaWiFS) and Moderate Resolution Imaging Spectroradiometer (MODIS) primary radiometric products improves the convergence between remote-sensing andinsitumeasurements, with the largest effects at 412 and 443 nm. Specifically, the scatter and bias of MODIS derived with respect toinsituLwnat 412 nm have shown values of 12% and 3% forcorrectedwith respect to values of 34% and -28% foruncorrecteddata, respectively. Similarly, the scatter and bias for SeaWiFS-derivedLwnat 412 nm have shown values of 14% and 4% forcorrectedwith respect to 32% and -20% foruncorrecteddata. Results at 667 nm for MODIS and at 670 nm for SeaWiFS, although displaying a reduction in the scatter of data, have shown a significant residual bias of about 11% and 17% with respect toinsituvalues. Finally, it was shown the need for restricting the application of thecorrectionschemeto data with atmospheric and marine optical features represented within the reference data set used to define the correction coefficients.

Water-type (WT) definition as a function of bio-optical properties has relevance to the exploitation of satellite ocean color data because it may provide a mean to optimize the atmospheric correction process or select the most appropriate bio-optical algorithm for the determination of optically significant constituents in seawater. Within such a framework, this paper defines a WT index assuming the existence of two extreme conditions. On one side, there are waters with optical properties totally depending on the phytoplankton component. On the other side, waters characterized by fully unrelated bio-optical quantities. The index of intermediate WTs is then defined on a statistical basis, and the resulting scheme is applied to a set of experimental data collected in an optically complex coastal region. Finally, in view of supporting remote sensing applications, a neural network algorithm is implemented to identify WTs using the normalized water leaving radiance.

In situ autonomous systems are commonly used for the collection of measurements for the vicarious calibration of satellite data and the successive validation of derived products. However, the use of autonomous systems creates the need of assessing the quality of the large volume of collected data. Within the framework of ocean color activities, this work investigates the consistency of normalized water leaving radiances spectra produced from measurements taken with an above-water autonomous system installed on an oceanographic tower. The study has shown the need of addressing the problem under two different levels of inference. The first level, so-called self-consistency, has demonstrated the capability of identifying spectra with a low statististical representativeness within the dataset itself. The second level, so-called relative-consistency, has provided the possibility of evaluating whether a spectrum is relatively consistent to a reference set of quality-assured data.

MOTIVATION: Clustering techniques such as k-means and hierarchical clustering are commonly used to analyze DNA microarray derived gene expression data. However, the interactions between processes underlying the cell activity suggest that the complexity of the microarray data structure may not be fully represented with discrete clustering methods. RESULTS: A newly developed software tool called MILVA (microarray latent visualization and analysis) is presented here to investigate microarray data without separating gene expression profiles into discrete classes. The underpinning of the MILVA software is the two-dimensional topographic representation of multidimensional microarray data. On this basis, the interactive MILVA functions allow a continuous exploration of microarray data driven by the direct supervision of the biologist in detecting activity patterns of co-regulated genes. AVAILABILITY: The MILVA software is freely available. The software and the related documentation can be downloaded from http://www.ncrg.aston.ac.uk/Projects/milva. User 'surrey' as username and '3245' as password to login. The software is currently available for Windows platform only.

MOTIVATION: Despite theoretical arguments that so-called 'loop designs' for two-channel DNA microarray experiments are more efficient, biologists continue to use 'reference designs'. We describe two sets of microarray experiments with RNA from two different biological systems (TPA-stimulated mammalian cells and Streptomyces coelicolor). In each case, both a loop and a reference design were used with the same RNA preparations with the aim of studying their relative efficiency. RESULTS: The results of these experiments show that (1) the loop design attains a much higher precision than the reference design, (2) multiplicative spot effects are a large source of variability, and if they are not accounted for in the mathematical model, for example, by taking log-ratios or including spot effects, then the model will perform poorly. The first result is reinforced by a simulation study. Practical recommendations are given on how simple loop designs can be extended to more realistic experimental designs and how standard statistical methods allow the experimentalist to use and interpret the results from loop designs in practice. AVAILABILITY: The data and R code are available at http://exgen.ma.umist.ac.uk CONTACT: [email protected].

Multilayer perceptron (MLP) neural network algorithms were developed to retrieve the absorption coefficient spectra of the colored dissolved organic matter (CDOM) and nonpigmented particulate matter (NPPM) from the remote sensing reflectance R/sub rs/ of optically complex waters. The two MLP algorithms, consisting of one hidden layer with ten neurons and requiring R/sub rs/ at 412, 490, and 665 nm as inputs, were trained with a comprehensive experimental dataset of the Northern Adriatic Sea coastal waters. The products of the proposed regional MLP algorithms showed higher accuracies than regional band-ratio algorithms, and exhibited average uncertainties of 20% and 25% in the determination of CDOM and NPPM absorption coefficients at 412 nm, respectively.

An operational system for autonomous above-water radiance measurements, called the SeaWiFS Photometer Revision for Incident Surface Measurements (SeaPRISM), was deployed at the Acqua Alta Oceanographic Tower in the northern Adriatic Sea and used for the validation of remote sensing radiometric products in coastal waters. The SeaPRISM data were compared with simultaneous data collected from an independent in-water system for a wide variety of sun elevations along with different atmospheric, seawater, and sea state conditions. The average absolute differences between the above- and in-water determinations of water-leaving radiances (computed linearly) were less than 4.5% in the 412-555-nm spectral interval. A similar comparison for normalized water-leaving radiances showed average absolute differences less than 5.1%. The comparison between normalized water-leaving radiances computed from remote sensing and SeaPRISM matchup data, showed absolute spectral average (linear) differences of 17.0%, 22.1%, and 20.8% for SeaWiFS, MODIS, and MERIS, respectively. The results, in keeping with those produced by independent in-water systems, suggest the feasibility of operational coastal networks of autonomous above-water radiometers deployed on fixed platforms (towers, lighthouses, navigation aids, etc.) to support ocean color validation activities.

Novelty detection is used to identify the range of applicability of empirical ocean color algorithms. This method is based on the assumption that the level of accuracy of the algorithm output depends on the representativeness of inputs in the training dataset. The effectiveness of the novelty detection method is assessed using two datasets: one representative of the northern Adriatic Sea coastal waters and the other representative of open sea waters. The two datasets are independently used to develop neural network algorithms for the retrieval of chlorophyll-a concentration (Chl-a). The range of applicability of the individual algorithms is presented using remote sensing data derived from the Sea-viewing Wide-Field-of-view Sensor (SeaWiFS) for three selected regions: the central Mediterranean Sea, the North Sea, and the Baltic Sea. An extension of the novelty detection technique is also proposed to blend the individual algorithms and to avoid discontinuities in the resulting Chl-a maps.

The determination of phytoplankton in seawater, quantified as chlorophyll-a concentration (Chl-a) or absorption of pigmented matter (a/sub ph/), is a major objective of optical remote sensing. The accuracy of multilayer perceptron (MLP) neural network algorithms in determining Chl-a and a/sub ph/ at 443 nm as a function of the multispectral remote sensing reflectance (R/sub rs/) was investigated for optically complex waters. The implementation of the MLP algorithms was carried out relying on an experimental dataset collected in a coastal region of the northern Adriatic Sea. The performance of the algorithms was assessed on both separate and combined Case 1 and Case 2 water types. The proposed MLP algorithms showed a better accuracy both with respect to other algorithms developed on the basis of the same dataset as well as with respect to independent algorithms operationally used for the processing of Sea-viewing Wide Field-of-view Sensor (SeaWiFS) data. The study also showed a high accuracy in determining a/sub ph/(443) and, thus, further confirmed the possibility of computing the inherent optical properties of seawater significant components from the R/sub rs/ spectra.