Investigating the controls on surface ocean dimethyl sulphide concentrations at regional to global scales

This thesis details a series of investigations into the controls on surface ocean concentrations of the climatically relevant, biogenic sulphur compound, dimethyl sulphide (DMS) at regional to global scales. The primary focus is upon the role of solar irradiance and metrics of biological activity in modulating DMS concentrations using bivariate and multivariate statistical techniques in conjunction with three different data sets from multiple spatial and temporal scales. Firstly, a statistical investigation into the proposed strong positive relationship between surface DMS concentration and the average mixed layer irradiance (solar radiation dose: SRD) was undertaken using DMS data from a series of cruise tracks from the Atlantic Meridional Transect (AMT) programme, primarily from the oligotrophic Atlantic gyres. Positive correlations were found between DMS and (a) SRD formulations using concurrently sampled in situ data (ρ=0.55 n=65 p<0.01), (b) SRD formulations based on using climatological data (ρ=0.74 n=65 p<0.01) and (c) a ultraviolet radiation dose (ρ= 0.67 n=54 p<0.01). The next analysis investigated whether the inclusion of a biological variable (chlorophyll or primary production) alongside irradiance could explain additional variance in DMS concentrations. This analysis employed a database of cruise data from a range of biogeochemical domains, latitudes and trophic conditions (AMT, the Barents Sea, the Atmospheric Chemistry Studies in the Oceanic Environment (ACSOE) research campaign and the DImethyl Sulphide biogeochemistry within a COccolithophore bloom (DISCO) study. Using multiple linear regression (MLR) analyses, it was found that the combination of, in situ rate of primary production and underwater irradiance accounted for significant variance in DMS concentrations in data from discrete depths within the euphotic zone (R2 = 0.55), from near-surface waters (R2 = 0.66) and within depth profile integrated data (R2 = 0.40). The final analysis is an investigation into global surface DMS dynamics using the global surface seawater DMS database (http://saga.pmel.noaa.gov/dms/) and satellite based retrievals of irradiance and primary production rates. A novel composite approach which combines multiple MLR models applied to Longhurst biogeochemical provinces, and using monthly averaged data, explained maximum variance. Models developed within a randomly selected training subset were able to explain significant variance within the remaining validation subset using this composite approach (predicted vs. observed ρ = 0.93, p = 0, n = 107). Previous studies had been unable to identify a strong link between DMS and indicators of the biological community (e.g. chlorophyll) at large scales. Our results suggest that a link exists between ecosystem productivity and DMS concentrations, and moderated by processes directly influenced by solar irradiance. These findings on large scale ecosystem controls on DMS, based on remote-sensing datasets, provide an advancement in the understanding and prediction of global-scale surface DMS concentrations.

Name dc:type.qualificationname

phd

Level dc:type.qualificationlevel

doctoral

Grantor dc:publisher.institution

University of East Anglia

Year dc:date.issued

2012

Author dc:creator

• Miles, Christopher James

Language dc:language

en