[Previous Article] [Next Article]
Forecast of Northern Tropical Atlantic SST Using
Linear Inverse Modeling
contributed by Cécile Penland and Ludmila Matrosova
NOAA-CIRES/Climate Diagnostics Center, Boulder, Colorado 80309-0449
Using the methods described in Penland and Matrosova (1997a,b), sea-surface temperature
anomalies (SSTAs) in the north tropical Atlantic Ocean (NTA: Fig. 1) and the Caribbean (CAR:
Fig. 1) are predicted. Global tropical SSTAs are used as predictors; that is, a prediction at lead
time is made by applying a statistically-obtained Green function G() to an observed initial
condition consisting of global tropical SSTAs. Three-month running means of the temperature
anomalies are used, the COADS 1950-79 annual cycle has been removed, and the data have been
projected onto the 20 leading empirical orthogonal functions (EOFs) explaining about 67% of the
variance. The NTA region has an RMS temperature anomaly of about 0.3 oC; the CAR region has
an RMS temperature anomaly of about 0.2 oC. Data have been provided by NCEP, courtesy of
R.W. Reynolds.
As shown in the March 1997 issue of this Bulletin (see Fig. 4 on p. 43), the prediction method has
been estimated to have a root mean square (RMS) error which approaches the RMS value at lead
times of about 15 months. The lead time up to which correlations between time series of
predictions and observations are significantly positive is also about 15 months. The reader should
be aware that these predictions are experimental and that only modest skill can be expected.
The inverse modeling prediction for the NTA SSTA is given in Fig. 2 for lead times of 3, 6, 9 and
12 months, with a slight decrease expected for the NTA SSTA. Figure 3 shows a prediction for
slightly increasing CAR SSTA.
Penland, C. and L. Matrosova, 1997a: On the prediction of tropical Atlantic sea surface
temperatures. Proceedings, 21st Climate Diagnostics and Prediction Workshop, October
28-November 1, 1996, Huntsville, Alabama, 203-206.
Penland, C. and L. Matrosova, 1997b: Prediction of tropical Atlantic sea surface temperatures
using linear inverse modeling. J. Climate, 10, submitted.
Fig. 1. Map showing the North Tropical Atlantic (NTA) and Caribbean (CAR) regions within
which the average SSTA is predicted.
Fig. 2. Time series of linear inverse modeling (LIM) predictions (light solid line) of NTA SSTA
for lead times of 3, 6, 9 and 12 months. Also shown are the verification series (heavy solid line)
and the one standard deviation confidence interval appropriate to the LIM forecast (dotted lines).
Fig. 3. As in Fig. 2, but for CAR SSTA.
