Global ocean dimethyl sulfide climatology estimated from observations
and an artificial neural network
Abstract
Marine dimethyl sulfide (DMS) is important to climate due to the ability
of DMS to alter Earth’s radiation budget. However, a knowledge of the
global-scale distribution, seasonal variability, and sea-to-air flux of
DMS is needed in order to understand the factors controlling surface
ocean DMS and its impact on climate. Here we examine the use of an
artificial neural network (ANN) to extrapolate available DMS
measurements to the global ocean and produce a global climatology with
monthly temporal resolution. A global database of 57,810 ship-based DMS
measurements in surface waters was used along with a suite of
environmental parameters consisting of lat-lon coordinates, time-of-day,
time-of-year, solar radiation, mixed layer depth, sea surface
temperature, salinity, nitrate, phosphate, silicate, and oxygen. Linear
regressions of DMS against the environmental parameters show that on a
global scale mixed layer depth and solar radiation are the strongest
predictors of DMS, however, they capture 14% and 12% of the raw DMS
data variance, respectively. The multi-linear regression can capture
more (29%) of the raw data variance, but strongly underestimates high
DMS concentrations. In contrast, the ANN captures ∼61% of the raw data
variance in our database. Like prior climatologies our results show a
strong seasonal cycle in DMS concentration and sea-to-air flux. The
highest concentrations (fluxes) occur in the high-latitude oceans during
the summer. We estimate a lower global sea- to-air DMS flux (17.90±0.34
Tg S yr−1) than the prior estimate based on a map interpolation method
when the same gas transfer velocity parameterization is used.