Reconstruction of satellite chlorophyll images under heavy cloud coverage using a neural classification method

Abstract We present a method able to fill large data gaps in satellite chlorophyll (CHL) images, based on the principle of Self-Organizing Maps (SOM) classification methods. The method makes use of complementary oceanic remote sensing observations: sea surface temperature (SST) and sea surface height (SSH). It relies on the assumption that a state of the ocean can be locally defined by its values of SST, SSH and CHL, and that a codebook of possible (SST, SSH, CHL) situations, if large enough, allows the reconstruction of incomplete situations of CHL. To account for the spatio-temporal context, each situation is characterized by the values of the three variables CHL, SST and SSH over a 3 × 3 spatial window surrounding the pixel, and at three successive times. In a first step, the feasibility and robustness of the method was assessed on a synthetic data set derived from a high-resolution coupled bio-physical ocean model. The SOM reconstruction method was particularly powerful in the case of intense cloud cover and to reconstruct CHL at scales larger than ~ 10 km. Subsequently, the method was adapted to real satellite data, which required introducing iterative learning and dealing with multi-modal determination of the referent vectors. The proposed improvements of the method proved satisfactory in retrieving CHL missing values under heavy cloud coverage and in improving the quality of weekly CHL products.

[1]  M. Lévy,et al.  Does the low frequency variability of mesoscale dynamics explain a part of the phytoplankton and zooplankton spectral variability? , 2004, Proceedings of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences.

[2]  Robert Frouin,et al.  Coverage opportunities for global ocean color in a multimission era , 1998, IEEE Trans. Geosci. Remote. Sens..

[3]  J. Beckers,et al.  Multivariate reconstruction of missing data in sea surface temperature, chlorophyll, and wind satellite fields , 2007 .

[4]  Robert H. Woodward,et al.  Improvements in coverage frequency of ocean color: combining data from SeaWiFS and MODIS , 1998, IEEE Trans. Geosci. Remote. Sens..

[5]  Teuvo Kohonen,et al.  Self-Organizing Maps , 2010 .

[6]  Ruoying He,et al.  Sea Surface Temperature Patterns on the West Florida Shelf Using Growing Hierarchical Self-Organizing Maps , 2006 .

[7]  Gurvan Madec,et al.  Large-scale impacts of submesoscale dynamics on phytoplankton: Local and remote effects , 2012 .

[8]  Sylvie Thiria,et al.  Multifractal analysis of oceanic chlorophyll maps remotely sensed from space , 2011 .

[9]  Jean-Michel André,et al.  Spatial/temporal variability of algal biomass and potential productivity in the Mauritanian upwelling zone, as estimated from CZCS data , 1987 .

[10]  Giulietta S. Fargion,et al.  Merger of ocean color data from multiple satellite missions within the SIMBIOS project , 2003, SPIE Asia-Pacific Remote Sensing.

[11]  M. Lévy.  The Modulation of Biological Production by Oceanic Mesoscale Turbulence , 2008 .

[12]  Antonio Turiel,et al.  Inferring missing data in satellite chlorophyll maps using turbulent cascading , 2008 .

[13]  E. Lorenz.  Atmospheric Predictability as Revealed by Naturally Occurring Analogues , 1969 .

[14]  P. Schlittenhardt,et al.  Upwelling and boundary circulation off Northwest Africa as depicted by infrared and visible satellite observations , 1991 .

[15]  S. Maritorena,et al.  Merged satellite ocean color data products using a bio-optical model: Characteristics, benefits and issues , 2010 .

[16]  Robert H. Weisberg,et al.  Patterns of ocean current variability on the West Florida Shelf using the self-organizing map , 2005 .

[17]  Bertrand Chapron,et al.  Three‐dimensional reconstruction of oceanic mesoscale currents from surface information , 2008 .

[18]  Richard T. Barber,et al.  Primary production off northwest Africa: the relationship to wind and nutrient conditions , 1977 .

[19]  A. H. Murphy,et al.  Use of an Analogue Procedure to Formulate Objective Probabilistic Temperature Forecasts in The Netherlands , 1983 .

[20]  C. Walton,et al.  Nonlinear Multichannel Algorithms for Estimating Sea Surface Temperature with AVHRR Satellite Data , 1988 .

[21]  Janet W. Campbell,et al.  The lognormal distribution as a model for bio‐optical variability in the sea , 1995 .

[22]  Edward R. Abraham,et al.  The generation of plankton patchiness by turbulent stirring , 1998, Nature.

[23]  Solomon Kullback,et al.  Information Theory and Statistics , 1960 .

[24]  M. Lévy,et al.  Grid degradation of submesoscale resolving ocean models: Benefits for offline passive tracer transport , 2012 .

[25]  F. d’Ovidio,et al.  Mixing structures in the Mediterranean Sea from finite‐size Lyapunov exponents , 2004, nlin/0404041.

[26]  Adrian P. Martin,et al.  Bringing physics to life at the submesoscale , 2012 .

[27]  Giulietta S. Fargion,et al.  Application of machine-learning techniques toward the creation of a consistent and calibrated global chlorophyll concentration baseline dataset using remotely sensed ocean color data , 2003, IEEE Trans. Geosci. Remote. Sens..

[28]  Tereza Cavazos.  Using Self-Organizing Maps to Investigate Extreme Climate Events: An Application to Wintertime Precipitation in the Balkans , 2000 .

[29]  W. Hazeleger,et al.  A model study on internally generated variability in subtropical mode water formation , 2000 .

[30]  P. Klein,et al.  Three-dimensional stirring of thermohaline fronts , 1998 .

[31]  Gurvan Madec,et al.  Impacts of sub-mesoscale physics on phytoplankton production and, subduction , 2001 .

[32]  J. Beckers,et al.  EOF Calculations and Data Filling from Incomplete Oceanographic Datasets , 2003 .

[33]  Francesco d'Ovidio,et al.  Stirring of the northeast Atlantic spring bloom: A Lagrangian analysis based on multisatellite data , 2007 .

[34]  Sylvie Thiria,et al.  Self-Organizing Maps and Unsupervised Classification , 2005 .

[35]  M. Conkright,et al.  Comparison of global chlorophyll climatologies: In situ, CZCS, Blended in situ -CZCS and SeaWiFS , 2003 .

[36]  C. McClain,et al.  Modis Validation, Data Merger and Other Activities Accomplished by the Simbios Project: 2002-2003 , 2013 .

[37]  P. Delecluse,et al.  OPA 8.1 Ocean General Circulation Model reference manual , 1998 .

[38]  S. De Monte,et al.  Fluid dynamical niches of phytoplankton types , 2010, Proceedings of the National Academy of Sciences.

[39]  B. Hewitson,et al.  Self-organizing maps: applications to synoptic climatology , 2002 .

[40]  Anthony J. Richardson,et al.  Using self-organizing maps to identify patterns in satellite imagery , 2003 .

[41]  Patrice Klein,et al.  Impact of sub-mesoscale physics on production and subduction of phytoplankton in an oligotrophic regime , 2001 .

[42]  Robert Frouin,et al.  The Proposal for the NASA Sensor Intercalibration and Merger for Biological and Interdisciplinary Oceanic Studies(SIMBIOS) Program, 1995 , 2002 .

[43]  Richard B. Alley,et al.  North Atlantic climate variability from a self-organizing map perspective , 2007 .

[44]  Gurvan Madec,et al.  Modifications of gyre circulation by sub-mesoscale physics , 2010 .

[45]  R. A. Leibler,et al.  On Information and Sufficiency , 1951 .

[46]  Thomas M. Smith,et al.  Reconstruction of Historical Sea Surface Temperatures Using Empirical Orthogonal Functions , 1996 .

[47]  H. Storch,et al.  The Analog Method as a Simple Statistical Downscaling Technique: Comparison with More Complicated Methods , 1999 .

[48]  Fouad Badran,et al.  Retrieval of aerosol type and optical thickness over the Mediterranean from SeaWiFS images using an automatic neural classification method , 2006 .

[49]  Richard W. Reynolds,et al.  A Real-Time Global Sea Surface Temperature Analysis , 1988 .

[50]  Warren S. Wooster,et al.  SEASONAL UPWELLING CYCLE ALONG THE EASTERN BOUNDARY OF THE NORTH ATLANTIC. , 1976 .

[51]  Jean-Marie Beckers,et al.  Cloud filling of ocean colour and sea surface temperature remote sensing products over the Southern North Sea by the Data Interpolating Empirical Orthogonal Functions methodology , 2011 .

[52]  G. Lapeyre,et al.  Does the tracer gradient vector align with the strain eigenvectors in 2D turbulence , 1999 .

[53]  A. Oschlies,et al.  Basin-scale pCO2 maps estimated from ARGO float data: A model study , 2009 .