Evaluation of chlorophyll-a and other ocean colour products to predict the organic mass fraction in sea spray aerosol

Initial efforts towards developing a combined organic-inorganic sea-spray source function parameterisation for use in large-scale models utilised a combination of chlorophyll-a and wind speed as primary input parameters to combine the biological and dynamical contributions. The initial studies [1,2], performed using low time resolved (monthly averages) satellite derived chlorophyll-a fields, reported only a modest correlation coefficient between Chlorophyll-a concentration ([Chl-a]) and organic matter (OM) enrichment in sea-spray (r = 0.55). This suggested that Chlorophyll-a is only partially suitable as a biological activity surrogate for predicting the organic mass fraction of sea spray. In this work aerosol chemical composition data collected at Mace Head Station, on the Irish West coast, from 2002 to 2009 (52 samples), have been used to calculate the organic mass fraction of submicron sea spray (OM%). Satellite ocean colour data from 1998 to 2010, with 1° resolution, were obtained from ESA GlobColour. Specifically, concentration data of chlorophyll-a, derived combining SeaWiFS, MODIS and MERIS data, coloured dissolved and detrital organic materials ([CDM]) and particulate organic carbon ([POC]) were used. Multichannel singular spectrum analysis (MSSA) was used to fill gaps in time series of satellite data due to cloud cover or other environmental factors. For each aerosol sample available at Mace Head, a map of average [Chl-a], [CDM] and [POC] has been produced. The correlation coefficient between OM% and ocean colour products at each grid point over the North-East Atlantic Ocean was computed, obtaining correlation maps. The new approach allowed the individuation of the oceanic region exerting the maximum influence on the chemical composition of sea spray aerosol measured at Mace Head, that is the area in which the maximum correlation was observed. This region is located west of Mace Head, between 47 and 57° N and 14 and 24° W. With the new approach an improved linear relation between [Chl-a] and OM% was obtained with a correlation coefficient r = 0.73. Following the same approach maximum correlations of 0.68 and 0.69 were obtained for [CDM] and [POC], respectively. These results demonstrate that, to date, chlorophyll-a is marginally the best biological activity surrogate available from satellite measurements for predicting the organic mass fraction of submicron sea-spray. Moreover, it is demonstrated that the use of daily data reduces the dispersion of the data around the best fit line and increases the correlation coefficient of the mixed organic-inorganic sea spray source function. [1] O'Dowd et al., GRL, 35, L01801, 2008. [2] Vignati et al., Atm. Env., 44, 670-677, 2010.