Smithsonian at the Poles: Contributions to International Polar
Smithsonian at the Poles: Contributions to International Polar
Smithsonian at the Poles: Contributions to International Polar
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
310 SMITHSONIAN AT THE POLES / SMITH AND COMISO<br />
sp<strong>at</strong>ial scales through time. At present, <strong>the</strong> only means <strong>to</strong><br />
accomplish this on <strong>the</strong> appropri<strong>at</strong>e scales is via s<strong>at</strong>ellite<br />
oceanography.<br />
S<strong>at</strong>ellites presently have <strong>the</strong> capability <strong>to</strong> accur<strong>at</strong>ely<br />
map <strong>the</strong> distributions of ice (Comiso, 2004), sea surface<br />
temper<strong>at</strong>ure (SST; Comiso, 2000; Kwok and Comiso,<br />
2002), and pigment concentr<strong>at</strong>ions (Moore and Abbott,<br />
2000), as well as o<strong>the</strong>r parameters such as winds, b<strong>at</strong>hymetry,<br />
cloud cover, and some gas concentr<strong>at</strong>ions such<br />
as ozone (Comiso, 2009). Some measurements use visible<br />
wavelengths and refl ectance from <strong>the</strong> surface, and<br />
<strong>the</strong>refore <strong>the</strong> d<strong>at</strong>a returned are reduced in space and time<br />
because of clouds; o<strong>the</strong>rs are ei<strong>the</strong>r passively detected or<br />
use o<strong>the</strong>r wavelengths <strong>to</strong> determine <strong>the</strong> distribution of <strong>the</strong><br />
variable. In biological oceanography a major variable of<br />
interest is ocean color, which is converted in<strong>to</strong> quantit<strong>at</strong>ive<br />
estim<strong>at</strong>es of pigment (chlorophyll) concentr<strong>at</strong>ions. While<br />
<strong>the</strong> estim<strong>at</strong>es include signifi cant error terms (because of<br />
<strong>the</strong> dependence of pigment estim<strong>at</strong>es as a function of l<strong>at</strong>itude,<br />
<strong>the</strong> limit<strong>at</strong>ion of refl ectance <strong>to</strong> <strong>the</strong> optical surface<br />
layer r<strong>at</strong>her than <strong>the</strong> entire euphotic zone, and <strong>the</strong> interference<br />
in some w<strong>at</strong>ers of dissolved organic m<strong>at</strong>ter), <strong>the</strong>se<br />
estim<strong>at</strong>es remain, and will remain, <strong>the</strong> only means <strong>to</strong> obtain<br />
synoptic assessments of phy<strong>to</strong>plank<strong>to</strong>n distributions<br />
over large areas as well as <strong>the</strong>ir temporal changes over<br />
rel<strong>at</strong>ively short (e.g., days) periods.<br />
Two s<strong>at</strong>ellites have provided nearly all of <strong>the</strong> d<strong>at</strong>a in <strong>the</strong><br />
past three decades on pigment distributions in <strong>the</strong> Sou<strong>the</strong>rn<br />
Ocean. The fi rst was <strong>the</strong> Nimbus 7 s<strong>at</strong>ellite, launched<br />
in 1978, which carried <strong>the</strong> Coastal Zone Color Scanner<br />
(CZCS). While questions concerning <strong>the</strong> d<strong>at</strong>a quality and<br />
coverage from CZCS have been voiced, <strong>the</strong> d<strong>at</strong>a were used<br />
<strong>to</strong> investig<strong>at</strong>e both <strong>the</strong> large-scale distributions of pigments<br />
in rel<strong>at</strong>ion <strong>to</strong> oceanographic variables (Sullivan et al., 1993;<br />
Comiso et al., 1993) and also <strong>the</strong> specifi c processes and<br />
regions (e.g., Arrigo and McClain, 1994). However, given<br />
<strong>the</strong> orbit, <strong>the</strong> frequency of d<strong>at</strong>a collection in <strong>the</strong> Sou<strong>the</strong>rn<br />
Ocean was quite restricted, and when compounded by <strong>the</strong><br />
loss of d<strong>at</strong>a from cloud cover, <strong>the</strong> temporal frequency was<br />
far from optimal. In 1996 <strong>the</strong> ORBView-2 s<strong>at</strong>ellite was<br />
launched, which included <strong>the</strong> Sea-viewing Wide Field-ofview<br />
Sensor (SeaWiFS). This s<strong>at</strong>ellite proved <strong>to</strong> be an extremely<br />
useful <strong>to</strong>ol for biological oceanographers, as <strong>the</strong><br />
sampling frequency was much gre<strong>at</strong>er and <strong>the</strong> d<strong>at</strong>a return<br />
in polar regions was far gre<strong>at</strong>er. For example, Moore et al.<br />
(1999) were able <strong>to</strong> detect a short-lived bloom in <strong>the</strong> Pacifi c<br />
sec<strong>to</strong>r of <strong>the</strong> Sou<strong>the</strong>rn Ocean th<strong>at</strong> was only infrequently<br />
sampled by ships. Dierssen et al. (2002) assessed <strong>the</strong> variability<br />
of productivity in <strong>the</strong> West Antarctic Peninsula region<br />
and found (based on a model) th<strong>at</strong> pigment concentr<strong>at</strong>ions<br />
were <strong>the</strong> dominant variable cre<strong>at</strong>ing vari<strong>at</strong>ions in space and<br />
time. Smith and Comiso (2008) assessed <strong>the</strong> productivity of<br />
<strong>the</strong> entire Sou<strong>the</strong>rn Ocean and found th<strong>at</strong> <strong>the</strong> “hot spots”<br />
of production were limited <strong>to</strong> continental shelf regions,<br />
and suggested th<strong>at</strong> this was a result of low iron concentr<strong>at</strong>ions<br />
coupled with deeper mixing in <strong>the</strong> offshore regions.<br />
The interaction of low iron and low irradiance (Sunda and<br />
Huntsman, 1997; Boyd and Abraham, 2001) gives rise <strong>to</strong> a<br />
large sp<strong>at</strong>ial limit<strong>at</strong>ion over broad areas.<br />
It is <strong>the</strong> purpose of this manuscript <strong>to</strong> look <strong>at</strong> <strong>the</strong><br />
scales of variability in <strong>the</strong> Sou<strong>the</strong>rn Ocean as a whole and<br />
<strong>to</strong> determine where such vari<strong>at</strong>ions are large by using primary<br />
production derived from SeaWiFS ocean color and<br />
advanced very high resolution radiometer (AVHRR) SST<br />
d<strong>at</strong>a in conjunction with a bio-optical model. We also will<br />
compare <strong>the</strong> modeled productivity with observed values,<br />
where those d<strong>at</strong>a are available <strong>to</strong> test <strong>the</strong> robustness of <strong>the</strong><br />
model. Finally, some aspects of <strong>the</strong> temporal p<strong>at</strong>terns of<br />
productivity in <strong>the</strong> Sou<strong>the</strong>rn Ocean are reviewed.<br />
MATERIALS AND METHODS<br />
Primary productivity was estim<strong>at</strong>ed using various d<strong>at</strong>a<br />
derived from s<strong>at</strong>ellites and a bio-optical model. The model<br />
was a vertically generalized production model (Behrenfeld<br />
and Falkowski, 1997b), in which primary productivity<br />
(PPeu, in units of mg C m �2 d �1 ) was estim<strong>at</strong>ed from <strong>the</strong><br />
following equ<strong>at</strong>ion:<br />
eu<br />
B<br />
opt<br />
o<br />
o<br />
PP = 0. 66125 × P<br />
E<br />
C × Z × D<br />
E + 41 .<br />
S<strong>at</strong> eu Irr<br />
B<br />
where Popt is <strong>the</strong> optimal r<strong>at</strong>e of pho<strong>to</strong>syn<strong>the</strong>sis within<br />
<strong>the</strong> w<strong>at</strong>er column (mg C (mg chl) �1 h�1 ) and is a function<br />
of temper<strong>at</strong>ure, Eo is <strong>the</strong> surface daily pho<strong>to</strong>syn<strong>the</strong>tically<br />
active radi<strong>at</strong>ion (PAR, mol pho<strong>to</strong>ns m�2 d�1 ), Cs<strong>at</strong> is <strong>the</strong><br />
surface chlorophyll concentr<strong>at</strong>ion (mg chl m�3 ) determined<br />
by s<strong>at</strong>ellite, Zeu is <strong>the</strong> depth of <strong>the</strong> euphotic zone<br />
B<br />
(m), and DIrr is <strong>the</strong> pho<strong>to</strong>period (h). Popt was estim<strong>at</strong>ed<br />
from sea surface temper<strong>at</strong>ures by <strong>the</strong> polynomial equ<strong>at</strong>ion<br />
B<br />
of Behrenfeld and Falkowski (1997b), and all Popt values<br />
<strong>at</strong> temper<strong>at</strong>ures less than �1.0ºC were set <strong>to</strong> 1.13.<br />
Temper<strong>at</strong>ure, PAR, ice concentr<strong>at</strong>ions, and chlorophyll<br />
concentr<strong>at</strong>ions were derived from different s<strong>at</strong>ellite<br />
d<strong>at</strong>a sets. Different s<strong>at</strong>ellite d<strong>at</strong>a were mapped <strong>to</strong> <strong>the</strong><br />
same grid as described below. We arbitrarily defi ned <strong>the</strong><br />
Sou<strong>the</strong>rn Ocean roughly as <strong>the</strong> region impacted by seasonal<br />
ice movements and hence set <strong>the</strong> nor<strong>the</strong>rn bound-