On the calibration and use of in situ ocean colour measurements for monitoring algal blooms
D. G. BOWERS†, S. KRATZER¶, J. R. MORRISON‡, P. S. D. SMITH†, P. TETT§, A. W. WALNE d and K. WILD-ALLEN§
†University of Wales Bangor, Menai Bridge, Anglesey LL59 5EY, Wales, UK ‡Bermuda Biological Station for Research, Ferry Reach, St.George’s, GE01, Bermuda §Department of Biological Sciences, Napier University, 10 Colinton Road, Edinburgh EH10 5DT, Scotland, UK d Sir Alistair Hardy Foundation for Ocean Science, The Laboratory, Citadel Hill, Plymouth Pl1 2PB, England, UK ¶Department of Physical Geography, University of Stockholm, S-106 91, Stockholm, Sweden (Received 6 January 1999 ) Abstract. Simultaneous in situ measurements ofchlorophyll concentration and water colour are reported at three diverse sites: a sea loch, the west Scottish shelf and the north Atlantic. A good (R2 5 0.97) log–log relation exists between the ratio, c, of the irradiance re ection coe cients at 490 and 570 nm and the chlorophyll concentration, C, over a range of chlorophyll concentrations from 0.01 to nearly 50 mg mÕ 3 . This relation can be expressedas C 5 1.87 cÕ 1 .8 1 . The equation is very similar to that used in remote sensing algorithms for the conversion of ocean colour data into chlorophyll concentrations. The root mean square (RMS) variation of observed chlorophyll about the values predicted by this equation is 33%. The robustness of this algorithm implies that this colour ratio can be used to monitor chlorophyll concentrations incase 1 waters with a minimum of additional calibration information. As an illustration, a time series is presented of colour-ratioderived chlorophyll concentration during the spring bloom in a Scottish sea loch in 1994. The data show reasonable agreement with the chlorophyll time series measured by a recording uorometer, and a comparison of the two time series serves to highlight the advantages anddisadvantages of each instrument.
Introduction It has long been recognized that the colour of the sea contains information on chlorophyll concentrations and potential primary productivity (Steemann Neilsen 1963 ). Early measurements of ocean colour were made by eye using colour-coded chemical mixtures for comparison (see the work of Schott in Sverdrup et al. (1942, p. 784)). Maps of thecolour of the Atlantic Ocean obtained in this way show remarkable similarity to much more recent maps of chlorophyll distribution prepared from satellite imagery. The rst quantitative suggestion for measuring ocean colour was made by Jerlov (1973 ) who suggested making simultaneous measurements of blue and green sealeaving irradiance. Because chlorophyll absorbs more blue light than green, theInternational Journal of Remote Sensing ISSN 0143-1161 print/ISSN 1366-5901 online © 2001 Taylor & Francis Ltd http://www.tandf.co.uk/journals
D. G. Bowers et al.
ratio of these irradiances —the ‘blue/green ratio’—decreases as the chlorophyll concentration increases. Jerlov tested an instrument based upon this principle in the Mediterranean Sea. The use of a colour ratio as a measure ofoceanic chlorophyll concentration received a great boost with the launch of the rst ocean colour satellite instrument, the Coastal Zone Color Scanner (CZCS) in 1978 (Morel and Prieur 1977, Gordon et al. 1983 ). For the rst time it became possible to measure colour ratios over the world’s oceans, cloud cover permitting. Gordon and Morel (1983 ) provided an algorithm for converting the blue/greenratio to chlorophyll concentration over a wide range of chlorophyll concentrations in optical case 1 waters. These are waters where variations in chlorophyll are the main cause of colour variations: they usually exclude water near the coast where sediment and dissolved organic material in freshwater input from the land can have a signi cant eŒ on the water colour. The ect Gordon–Morel algorithm...