The Ocean and Land Color Instrument (OLCI) onboard Sentinel 3A satellite was launched in February 2016. Level 2 (L2) products have been available for the public since July 2017. OLCI provides the possibility to monitor aquatic environments on 300 m spatial resolution on 9 spectral bands, which allows to retrieve detailed information about the water quality of various type of waters. It has only been a short time since L2 data became accessible, therefore validation of these products from different aquatic environments are required. In this work we study the possibility to use S3 OLCI L2 products to monitor an optically highly complex shallow lake. We test S3 OLCI-derived Chlorophyll-a (Chl-a), Colored Dissolved Organic Matter (CDOM) and Total Suspended Matter (TSM) for complex waters against in situ measurements over Lake Balaton in 2017. In addition, we tested the machine learning Gaussian process regression model, trained locally as a potential candidate to retrieve water quality parameters. We applied the automatic model selection algorithm to select the combination and number of spectral bands for the given water quality parameter to train the Gaussian Process Regression model. Lake Balaton represents different types of aquatic environments (eutrophic, mesotrophic and oligotrophic), hence being able to establish a model to monitor water quality by using S3 OLCI products might allow the generalization of the methodology.

Ocean Color remote sensing has a great importance in monitoring of aquatic environments. The number of optical imaging sensors onboard satellites has been increasing in the past decades, allowing to retrieve information about various water quality parameters of the world's oceans and inland waters. This is done by using various regression algorithms to retrieve water quality parameters from remotely sensed multi-spectral data for the given sensor and environment. There is a great number of such algorithms for estimating water quality parameters with different performances. Hence, choosing the most suitable model for a given purpose can be challenging. This is especially the fact for optically complex aquatic environments. In this paper, we present a concept to an Automatic Model Selection Algorithm (AMSA) aiming at determining the best model for a given matchup dataset. AMSA automatically chooses between regression models to estimate the parameter in interest. AMSA also determines the number and combination of features to use in order to obtain the best model. We show how AMSA can be built for a certain application. The example AMSA we present here is designed to estimate oceanic Chlorophyll-a for global and optically complex waters by using four Machine Learning (ML) feature ranking methods and three ML regression models. We use a synthetic and two real matchup datasets to find the best models. Finally, we use two images from optically complex waters to illustrate the predictive power of the best models. Our results indicate that AMSA has a great potential to be used for operational purposes. It can be a useful objective tool for finding the most suitable model for a given sensor, water quality parameter and environment.

Gaussian Process Regression (GPR) have experienced tremendous success in biophysical parameter retrieval in the last years. The GPR provides a full posterior predictive distribution so one can derive mean and variance predictive estimates, i.e. point-wise predictions and associated confidence intervals. GPR typically uses translation invariant covariances that make the prediction function very flexible and nonlinear. This, however, makes the relative relevance of the input features hardly accessible, unlike in linear prediction models. In this paper, we introduce the Sensitivity Analysis (SA) of the GPR predictive mean and variance functions to derive feature rankings and spectral spacings, respectively. The methodology can be used to uncover knowledge in any kernel-based regression method, it is fast to compute, and it is expressed in closed-form. The methodology is evaluated on GPR for global ocean chlorophyll prediction, revealing the most important spectral bands and their spectral spacings. We illustrate the (successful) methodology in several datasets and sensors.

The monitoring of Chlorophyll-a (Chl-a) concentration in high northern latitude waters has been receiving increased focus due to the rapid environmental changes in the sub-Arctic, Arctic. Spaceborne optical instruments allow the continuous monitoring of the occurrence, distribution, and amount of Chl-a. In recent years, the Ocean and Land Color Instruments (OLCI) onboard the Sentinel 3 (S3) A and B satellites were launched, which provide data about various aquatic environments on advantageous spatial, spectral, and temporal resolutions with high SNR. Although S3 OLCI could be favorable to monitor high northern latitude waters, there have been several challenges related to Chl-a concentration retrieval in these waters due to their unique optical properties coupled with challenging environments including high sun zenith angle, presence of sea ice, and frequent cloud covers. In this work, we aim to overcome these difficulties by developing a machine-learning (ML) approach designed to estimate Chl-a concentration from S3 OLCI data in high northern latitude optically complex waters. The ML model is optimized and requires only three S3 OLCI bands, reflecting the physical characteristic of Chl-a as input in the regression process to estimate Chl-a concentration with improved accuracy in terms of the bias (five times improvements.) The ML model was optimized on data from Arctic, coastal, and open waters, and showed promising performance. Finally, we present the performance of the optimized ML approach by computing Chl-a maps and corresponding certainty maps in highly complex sub-Arctic and Arctic waters. We show how these certainty maps can be used as a support to understand possible radiometric calibration issues in the retrieval of Level 2 reflectance over these waters. This can be a useful tool in identifying erroneous Level 2 Remote sensing reflectance due to possible failure of the atmospheric correction algorithm.

This paper evaluates two alternative regression techniques for oceanic Chlorophyll-a (Chl-a) content estimation. One of the investigated methodologies is the recently introduced Gaussian Process Regression (GPR) model. We explore two feature ranking methods derived for the GPR model, namely Sensitivity Analysis (SA) and Automatic Relevance Determination (ARD). We also investigate a second regression method, the Partial Least Squares Regression (PLSR) for oceanic Chl-a content estimation. Feature relevance in the PLSR model can be accessed through the Variable Importance in Projection (VIP) feature ranking algorithm. The paper thus analyses three feature ranking models, SA, ARD and VIP, which are all derived from different fundamental principles, and uses the ranked features as inputs to the GPR and PLSR to assess regression strengths. We compare the regression performances using some common performance measures, and show how the feature ranking methods can be used to find the lowest number of features to estimate oceanic Chl-a content by using the GPR and PLSR models, while still producing comparable performance to the state-ofart algorithms. We evaluate the models on a global MERIS matchup dataset. Our results show that the GPR model has the best regression performance for most of the input feature sets we used, and our conclusion is this model can favorably be used for Chl-a content retrieval, already with two features, ranked by the either the SA or ARD methods.