The global distribution of high Remote-sensing reflectance (Rrs) waters
visible from satellite, likely associated with coccolithophore blooms,
has changed markedly over the past 40 years. Over that period there has
globally been an overall decrease in bloom area of 1.15 million km2 but
with notable Rrs increases in the Barents Sea and the Antarctic Ocean.
The primary drivers of these fundamental changes to ocean
biogeochemistry have been investigated using Machine Learning techniques
together with contemporaneous global multi-decadal time-series of
sea-surface temperature (SST); wind speed and stress; sea level anomaly
(SLA); photosynthetically available radiation (PAR) and; mixed layer
depth (MLD). When split into ocean provinces different drivers of
positive and negative trends in Rrs were found to dominate in different
regions, but generally increases were found to coincide with changes to
SST, PAR and reductions to wind-speed.