Variability of chlorophyll-a bloom timing in NOW ... - Camille Albouy

The North Water (NOW), located between Greenland and Ellesmere. Island (Canadian Arctic), is one of the largest and most productive polynyas in the Northern ...
2MB taille 3 téléchargements 295 vues
Variability of chlorophyll-a bloom timing in NOW polynya: a preliminary analysis Christian Marchese1, Camille Albouy1, Steve Vissault1, Jean-Éric Tremblay2, Simon Bélanger1 1 Université

du Québec à Rimouski; 2 Université Laval

Abstract

Method

Discussion

The North Water (NOW), located between Greenland and Ellesmere Island (Canadian Arctic), is one of the largest and most productive polynyas in the Northern Hemisphere. Recent sea ice and oceanic observations reveal that major changes are happening in the NOW. The impact of those changes on the marine productivity is still not well understood. A satellite-based primary production (PP) model was recently used to estimate PP in the NOW for the period 1998-2010. PP time series revealed that productivity dropped dramatically after 2002. More precisely, for the whole studied area and period a decrease of 25% in PP was observed. This decrease in PP is likely due to a decrease of the biomass, as mostly observed along the Ellesmere Island. In this regard, the seasonal-to-inter-annual variability of chlorophyll-a, obtained from the GlobColour project (http://www.globcolour.info/), was monitored for the same period through the use of a Gaussian function from which a baseline of phenological characteristics was extracted. Results suggest that major environmental changes can locally have a crucial influence on phytoplankton dynamics with unknown consequences on the whole marine food web.

To quantify the characteristics of the bloom, the time series (calculated doing the median of all valid pixels included in the sub-regions) were fitted with a Gaussian curve (Figure 2).

The preliminary analysis indicated that the timing, amplitude and duration of the phytoplankton bloom differed among the eight defined sub-regions (Figure 1B). The main results (Figure 4) can be summarized as follow:

Simple Gaussian

Simple Gaussian with linear term Chl-a GlobColour Data 8-Day (4.6 Km)

Double Gaussian

Double Gaussian with linear term

•  The timing of the bloom shifted between years, showing a strong inter-annual variability. Overall, the bloom begins in late April or slightly later in May, depending of the sub-region.

Best model fit for the annual bloom

Phenology model parameters

Fig. 2 Flowchart summarizing the processes to model the annual bloom in NOW

The starting parameters were determined using the Gauss–Newton Method based on the nonlinear least-square fit (Wu et al, 2007). The Akaike information criterion (AIC) is used to provide a mean for model selection. Parameters obtained from the Gaussian fit (Figure 3):

Fig. 3 - Characteristics of the phytoplankton bloom

tp ± 2σ occupies 95.5% of the chlorophyll-a distribution. In this study, (tp - 2σ) and (tp + 2σ) were defined as start and end timing of bloom. Consequently, bloom duration was defined as 4σ (Yamada and Ishizaka, 2006)

Results 1998−2012 1.5 1.0

R squared: 0.94 Bloom timing: 4.52 Bloom duration: 4.92 Peak day: 6.98









R squared: 0.99 Bloom timing: 5.42 Bloom duration: 2.97 Peak day: 6.91

2.0



1.5











1.0





0.5

0.5











Ellesmere_and_Devon_Island

1.2

Carey_Island



R squared: 0.99 Bloom timing: 4.74 Bloom duration: 2.19 Peak day: 5.83

● ● ●

(A)

(B)

CHL(mg m3)



1.0 ●



1.00 0.75







0.4



www.PosterPresentations.com





● ●

0.50













1.5

0.25



1.0

R squared: 0.99 Bloom timing: 5.29 Bloom duration: 0.84 Peak day: 5.71





Greenland_south_east_coast

2.0





R squared: 1 Bloom timing: 4.81 Bloom duration: 1.44 Peak day: 5.53

1.5 ●



1.0







0.5

● ● ●

5

0

3.

4.





0.5





5

0

5

5.

5.

6.

0

5

6.

0 7.

5 7.

0

8.



5

0

8.







4.

ü  Evaluate the most robust method to describe phenology metrics of phytoplankton blooms.





0.2

9.

3.







5

0 9.

ü  Compare more precisely phytoplankton phenology parameters with variation of other environmental factors (e.g. atmospheric forcing, mixed-layer depth and sea ice variability).



5

4.

0

4.

5

0 5.

5 5.

0 6.

5

6.

0

7.

7.

5

0

8.

8.

Month

Fig. 4 – Climatology (1998-2012) of the phytoplankton bloom (top) and bloom period from 1998 to 2012 (below) in the eight sub-regions Kane$Basin$

Years$

Years$

Smith$Sound$

2012$ 2011$ 2010$ 2009$ 2008$ 2007$ 2006$ 2005$ 2004$ 2003$ 2002$ 2001$ 2000$ 1999$ 1998$ 125$

Start$$

160$ 195$ Days$of$the$year$ Peak$

230$

125$ 160$ Days$of$the$year$ Start$

Peak$$

90$

125$

Start$

Peak$

End$

230$

Peak$

265$

300$

End$

Melville$Bay$

Peak$

230$

2012$ 2011$ 2010$ 2009$ 2008$ 2007$ 2006$ 2005$ 2004$ 2003$ 2002$ 2001$ 2000$ 1999$ 1998$

End$

125$

Start$

90$

125$

160$ 195$ Days$of$the$year$

Start$$

Peak$$

230$ End$$

160$ 195$ Days$of$the$year$ Peak$

230$

265$

End$

Greenland$south$east$coast$

Bylot$Island$south$coast$$ 2012$ 2011$ 2010$ 2009$ 2008$ 2007$ 2006$ 2005$ 2004$ 2003$ 2002$ 2001$ 2000$ 1999$ 1998$

Melling, H., Y. Gratton, and R. G. Ingram (2001), Ocean circulation within the North Water polynya of Baffin Bay, Ocean Atmosphere, 39, 301-325. Munchow, A., K. Falkner, and H. Melling, Baffin Island and West Greenland Current systems in northern Baffin Bay. Progress in Oceanography, 27 pp., accepted 2013.

90$

Years$

Years$

160$ 195$ Days$of$the$year$

Start$

End$

160$ 195$ Days$of$the$year$

160$ 195$ 230$ Days$of$the$year$

Years$

2012$ 2011$ 2010$ 2009$ 2008$ 2007$ 2006$ 2005$ 2004$ 2003$ 2002$ 2001$ 2000$ 1999$ 1998$ 90$

Years$

125$

125$

Start$

Lancaster$Sound$

90$

Gaillard, B., Olivier, F., Thébault, J., Meziane, T., Tremblay, R., Dumont, D., Bélanger, S., Gosselin, M., Chauvaud, L., Martel, A. L., and Archambault, P.: Bathyal Arctic bivalves benefit from climate changes, submitted, 2013.

Carey$Island$

195$

2012$ 2011$ 2010$ 2009$ 2008$ 2007$ 2006$ 2005$ 2004$ 2003$ 2002$ 2001$ 2000$ 1999$ 1998$

References

End$

Years$

Years$

2012$ 2011$ 2010$ 2009$ 2008$ 2007$ 2006$ 2005$ 2004$ 2003$ 2002$ 2001$ 2000$ 1999$ 1998$

2012$ 2011$ 2010$ 2009$ 2008$ 2007$ 2006$ 2005$ 2004$ 2003$ 2002$ 2001$ 2000$ 1999$ 1998$

265$

Ellesmere$:$Devon$Island$

90$

RESEARCH POSTER PRESENTATION DESIGN © 2012





R squared: 0.97 Bloom timing: 4.63 Bloom duration: 2.6 Peak day: 5.93



•  The northern and western par of the NOW is dominated by the southward flows coming from the Arctic Ocean (Melling et al., 2001).

Objectives: quantify the characteristics of the bloom (e.g. initiation, duration and intensity) and describe regional differences in phytoplankton phenology pattern





Lancaster_Sound



R squared: 1 Bloom timing: 5.24 Bloom duration: 1.42 Peak day: 5.95





Melville_Bay 0.8

90$

•  In southeast area, where a reduced convection allows relatively low ice cover and shallow mixed layers, phytoplankton blooms can be expected to develop relatively early (ca. 2.5 months) due to increased irradiance in the upper mixed layer relative to north-western part, which is influenced by arctic cold water. (Tremblay et al., 2002).



0.5 ●

•  Recent oceanographic data reveal interactions between fresh and cold Arctic waters advected southward along Baffin Island and salty and warm Atlantic waters advected northward along western Greenland (Münchow et al., 2013).

•  Recent evidence from Arctic bivalves living at depth ∼ 600 m suggested that a dramatic increase in the export of freshly produced organic matter in the euphotic zone to the deep ocean has occurred during the last decade (Gaillard et al., 2013).

● ●





Bylot_Island_south_coast

The NOW is closely connected to Lancaster Sound (where a smaller polynya is also present) and Baffin Bay by a powerful system of ocean currents (Baffin Island and West Greenland Currents - Figure 1A) that directly affect the region’s climate and biology.

● ●





0.6

Fig. 1 - Map of the Baffin Island and West Greenland Currents (A) and the key sub-regions (B). The sub-regions approximate the distribution of the different water masses.



R squared: 1 Bloom timing: 4.68 Bloom duration: 1.51 Peak day: 5.43

1.5

0.8 0.4

The results imply that the timing and magnitude of blooms in the North Water are locally controlled by a succession of oceanic and climatic forcings. In this regard, the future work will consist to:

Smith_Sound ●

•  The peak of the bloom is the parameter that has a lower interannual variability (e.g. Smith Sound). Probably, part of the interannual variability in the start and the end of the blooms is due to the choice of phenology metrics.

Future work

Kane_Basin ●

•  In some sub-regions (e.g. Greenland south-east coast) a phytoplankton bloom advancement (timing) is evident through the years. •  As compared to other sub-regions the bloom period was longer in Kane Basin and in Smith Sound with an average duration through the 15 years of 51 and 80 days, respectively.

ü  Bloom timing (tp – 2σ) ü  Bloom peak (mg/m3) ü  Peak day (tp - in days) ü  Bloom duration (σ - in days)

Introduction The North Water Polynya is an area of open water surrounded by sea ice, which lies between Greenland and Canada in northern Baffin Bay. It is the world’s largest polynya with an area of about 80 000 km2, and it creates a warm microclimate that provides a refuge for narwhals, belugas, and bowhead whales to feed and rest.

AIC R-squared and analysis of residuals

2012$ 2011$ 2010$ 2009$ 2008$ 2007$ 2006$ 2005$ 2004$ 2003$ 2002$ 2001$ 2000$ 1999$ 1998$ 90$

125$

Start$

160$ 195$ Days$of$the$year$ Peak$

End$

230$

Tremblay J.E., Gratton Y., Fauchot J., Price N.M., 2002. Climatic and oceanic forcing of new, net and diatom production in the North Water Polynya. Deep-Sea Res II 49:4927-4946 Wu Y., I.K. Peterson, C.L. Tang, T. Platt, S. Sathyendranath, and C. Fuentes-Yaco, 2007. The impact of sea-ice on the initiation of spring bloom on the Newfoundland and Labrador Shelves. J. of Plankton Res., 29(6): 509-514 Yamada, K., and J. Ishizaka (2006), Estimation of interdecadal change of spring bloom timing, in the case of the Japan Sea, Geophys. Res. Lett., 33, L02608, doi: 10.1029/2005GL024792.

[email protected]