Diurnal variability of turbidity and light attenuation in the southern North Sea from the SEVIRI geostationary sensor G. Neukermans a, b, c, d, ⁎, K.G. Ruddick a , N. Greenwood e a Management Unit of the North Sea Mathematical Models (MUMM), Royal Belgian Institute for Natural Sciences (RBINS), Gulledelle 100, B-1200 Brussels, Belgium b Université Lille Nord de France, F-59000 Lille, France c Université du Littoral Côte d'Opale (ULCO), Laboratoire d'Océanologie et Géosciences (LOG), F-62930 Wimereux, France d Centre National de la Recherche Scientifique (CNRS), UMR 8187, F-62930 Wimereux, France e Centre for Environment, Fisheries and Aquaculture Science (Cefas), Pakefield Road, Lowestoft, Suffolk, NR33 0HT, UK abstract article info Article history: Received 23 November 2011 Received in revised form 10 May 2012 Accepted 3 June 2012 Available online 10 July 2012 Keywords: PAR attenuation Atmospheric correction Tidal variability Satellite data validation This study follows up on the successful feasibility study of Neukermans et al. (2009) for mapping suspended matter in turbid waters from the SEVIRI sensor on board the METEOSAT geostationary weather satellite plat- form. Previous methodology is extended to the mapping of turbidity, T, and vertical attenuation of photosyn- thetically active radiation (PAR), K PAR . The spatial resolution of the SEVIRI products is improved from 3 km × 6.5 km to 1 km × 2 km using the broad high resolution visual band. The previous atmospheric correc- tion is further improved and the uncertainties on marine reflectance due to digitization are considered. Based on a two year archive of SEVIRI imagery, available every 15 min, the diurnal variability of T and K PAR is inves- tigated during cloud free periods and validated using half-hourly T and K PAR data obtained from a system of moored buoys (SmartBuoys) in the southern North Sea. Based on numerous match-ups, 80% of SEVIRI de- rived T and K PAR are within 53% and 39% of SmartBuoy T and K PAR , respectively. Results further show that on cloud free days, the SEVIRI T and K PAR signals are in phase with the SmartBuoy data, with an average dif- ference in the timing of the maximum T and K PAR of 11 min and 23 min, respectively. It is concluded that di- urnal variability of T and K PAR can now be mapped by remote sensing offering new opportunities for improving ecosystem models and monitoring of turbidity. Limitations of the current SEVIRI sensor and per- spectives for design of future geostationary sensors and synergy with polar orbiting satellites are discussed. © 2012 Elsevier Inc. All rights reserved. 1. Introduction Polar-orbiting multispectral ocean colour sensors such as the Sea- viewing Wide Field-of-view Sensor (SeaWiFS), the Moderate Resolu- tion Imaging Spectroradiometer (MODIS), and Medium Resolution Im- aging Spectrometer (MERIS) provide 2-day coverage of the global ocean and coastal zones since their respective launches in 1997 and 2002. These sensors have become well-established sources (McClain, 2009) of concentration of chlorophyll a, [Chl a] (see Table 1 for notation), and suspended particulate matter, [SPM], and there has been consider- able progress towards many new products including particulate and dissolved organic and inorganic carbon (Stramski et al., 1999; Vantrepotte et al., 2011), particle size distribution (Loisel et al., 2006), phytoplankton species composition (Alvain et al., 2008), vertical light attenuation (Stumpf et al., 1999), turbidity (Nechad et al., 2009; Stumpf et al., 1999; Woodruff et al., 1999) etc. During the last decades the spectral and spatial resolution of space-borne ocean colour sensors has improved, from multispectral to hyperspectral (e.g., Hyperspectral Imager for the Coastal Ocean, launched in September 2009), and from 1 km nadir pixel resolution down to less than 100 m in coastal areas. The quality and quantity of atmospheric corrections and bio-optical al- gorithms has also significantly progressed. Even though further progress can still be expected for polar-orbiting sensors in terms of sensor design and processing algorithms, their sam- pling frequency, typically once per day, is insufficient for many studies and applications. Many physical and biogeochemical processes in coastal regions show variability at time scales shorter than the daily sampling fre- quency of polar-orbiting sensors. For example, in situ measurements have shown that [SPM] can vary by a factor two or more during the day due to horizontal advection and/or vertical resuspension forced by tides or wind events (Eisma & Irion, 1988; Thompson et al., 2011). Hence, long term data series from polar-orbiting sensors are affected by aliasing that can only be treated indirectly (e.g., Stumpf et al., 1993). Furthermore, cloudi- ness and/or sun glint reduce data availability from typically once per day (e.g. mid-latitude MODIS imagery) to significantly less. Remote sensing applications, such as harmful algae bloom detection (Stumpf et al., 2003; Tomlinson et al., 2004), have critical vulnerability to such data gaps. Remote Sensing of Environment 124 (2012) 564–580 ⁎ Corresponding author at: Marine Physical Laboratory, Scripps Institution of Ocean- ography, University of California, San Diego, La Jolla, CA 92093-0238, USA. Tel.: + 1 858 534 7841. Fax: +1 858 534 7641. E-mail addresses: gneukermans@ucsd.edu (G. Neukermans), k.ruddick@mumm.ac.be (KG. Ruddick), naomi.greenwood@cefas.co.uk (N. Greenwood). 0034-4257/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.rse.2012.06.003 Contents lists available at SciVerse ScienceDirect Remote Sensing of Environment journal homepage: www.elsevier.com/locate/rse