Monitoring a Sentinel Species from Satellites: Detecting Emiliania huxleyi in 25 Years of AVHRR Imagery - Satellite-based Applications on Climate Change - page 276

bloom area in all regions, but there was generally no consistent relationship
between the ENSO indices and the fraction of area covered by E . huxleyi . The
monthly and annual PDO indices were also not significantly correlated to E . huxleyi
blooms over the 25-year time series, although it is worth noting that the summer-
time PDO had a higher correlation for all regions (0.2-0.3) compared to wintertime
PDO (0.05-0.06). Although the NAO wintertime index was correlated with a 2-year
lag in bloom fractional area between 0.3 and 0.4 for all regions, they were not
statistically significant.
The lack of significant correlations between bloom fractional area and the
relevant environmental variables and climatic indices examined above suggest
that a nonlinear combination of these or others factors, such as changing ocean
chemistry, are responsible for the decreasing trend in the surface area of E . huxleyi
blooms documented in the four regions of our study. For example, recent studies
suggest that decreasing alkalinity in ocean surface layers could lead to changes in
coccolith size, shell thickness, species distribution, calcification rate, and carbon
fixation (Barcelos e Ramos et al. 2010 ; Charalampopoulou et al. 2010 ) which could
conceivably decrease the calcite concentration within blooms and impair our ability
to detect them in AVHRR imagery. Further study, especially with in situ data, is
required to ascertain the factors responsible for this decreasing trend in the
E . huxleyi bloom surface area.
18.4 Conclusion
The well-established ocean color satellite measurements of E . huxleyi blooms
around the world were extended back to 1982 by identifying them in AVHRR
imagery. The observed trend over 25 years toward decreasing bloom areal extent in
the four regions examined at the same time that temperatures have been rising
suggests that the hypothesized relationship between E . huxleyi bloom events and
warming temperatures does not hold over the long term. Other than the weak
relationship between bloom fractional area and mixed layer depth, the paucity of
robust correlations with physical variables and climate indices suggest a nonlinear
combination of those or other factors are responsible for the decreasing trend in the
surface area of E . huxleyi blooms documented here. Other possible factors, such as
ocean acidification or salinity anomalies, need to be studied further.
Acknowledgments This study benefited from many discussions with Mete Uz. The AVHRR data
were specially processed for R rs by Aleksander Jelenak and William Straka. This research was
supported by the Office of Global Program's NOAA Climate and Global Change Program. The
views, opinions, and findings contained in this paper are those of the authors and should not be
construed as an official National Oceanic and Atmospheric Administration or US Government
position or decision.