Geoscience Reference
In-Depth Information
3.3.
Time-series analysis
Further analysis was carried out using time-series data for Chl-
a
and SST
between 2002 and 2005 for October to January (Fig. 2). Based on daily data
(Fig. 2(A)), the Chl-
a
values increased close to the time of the earthquake
(up to 0.69 mg m
−
3
on December 22, 2004), and soon after fluctuations
occurred with a series of aftershocks. Nevertheless, the Chl-
a
concentrations
were appreciably higher during the period of the disaster than that for other
years (Fig. 2(B)).
While Chl-
a
increased evidently from 3 to 10 December and attained
a peak value of 0.35 mg m
−
3
between 19 and 26 December (Fig. 2(B)), it
decreased to 0.33 mg m
−
3
between 27 and 31 December 2004. A further
decrease to 0.25 mg m
−
3
occurred through the period January 1-8, 2005.
The differences in Chl-
a
values between the higher daily averages compared
to the lower 8-day averages are due to the effect of smoothing. Figure 2(C)
shows inverse relationship between SST and Chl-
a
.
Fig. 2.
(A) Daily concentrations of Chl-
a
(from box-X in Fig. 1(A)). (B) Comparison
of Chl-
a
(8-day average) between the three years (2002-2005) (from box-Y in Fig. 1(A)).
(C) Chl-
and SST (8-day average) between October 2004 and January 2005 (from box-Y
in Fig. 1(A)). The date of the tsunami is indicated by the downward pointing arrow in
all panels. Discontinuity in Fig. 2(A) is due to missing values in daily coverage.
a
