Geoscience Reference
In-Depth Information
board the NASA Nimbus-7 satellite during 1978-1987. Events associated
with ENSO during this time period were studied in Ref. 13 for global SST
and in Ref. 14 for SST in the southern Pacific and Indian Oceans.
In our current study, we are interested in analyzing interannual and
decadal variability in the Pacific SSTs from 1985 to 1997 using the AVHRR
satellite data. Here, we focus on both spatial and temporal distribution of
the ENSO component in the Pacific as well as the propagation pattern.
The main objective here is to examine the dynamic behavior of SST and its
implications for climate modeling and climate variability. In Sec. 2, we intro-
duce our method and derive the Pacific SST deviations. Results and dis-
cussions are given in Sec. 3. Finally, we make concluding remarks in Sec. 4.
2. Method
There is a rich spectrum of variability in the SSTs, ranging from intrasea-
sonal, seasonal, interannual to decadal. In order to enhance the observations
of the interannual to decadal events that are the subject of this paper, we
preprocess our data to derive the Pacific SST deviations by removing a
modeled seasonal cycle (see Ref. 13) from the data at each pixel. We then
apply a spatial Gaussian filter to smooth out spatial fluctuations (of about
5
◦
×
5
◦
), which are probably caused by high-frequency temporal variations.
Our Gaussian smoother matrix is 25
×
25, whose values are given by the
two-dimensional Gaussian function centered in the box with half width
of six grid elements. Notice that the AVHRR SST data has a half-degree
resolution.
In order to study propagation patterns of the SST deviations, we select
six paths in the Pacific Ocean (see Fig. 1). Five paths are latitude lines,
along the equator, latitudes 35
◦
,45
◦
and 50
◦
north and latidude 40
◦
south,
respectively. The last path is chosen in the North Pacific and follows approx-
imately the North Pacific Current.
Hoffmueller diagrams along the chosen path show the SST deviations
versus the location along the path as the
x
-coordinates and the time as the
y
-coordinates. A propagation pattern can be identified as a diagonal color
feature in the Hoffmueller diagram.
3. Results and Discussions
We now analyze Hoffmueller diagrams for the SST deviations along the six
paths defined in Fig. 1 and compare our results with previous studies. The
left panel of Fig. 2 shows the Hoffmueller diagram for the SST deviations
