Tripole Index for the Interdecadal Pacific Oscillation (TPI)
The TPI (Tripole Index) is a metric developed to represent the tripolar pattern of Sea Surface Temperature (SST) anomalies associated with the Interdecadal Pacific Oscillation (IPO). It is calculated from the difference between the mean SST anomaly in the central equatorial Pacific and the mean anomalies in the northern and southern subtropical Pacific regions, as defined in Henley et al. (2015).
The regions used to compute the index are:
Region 1: 25°N–45°N, 140°E–145°W
Region 2: 10°S–10°N, 170°E–90°W
Region 3: 50°S–15°S, 150°E–160°W
This spatial configuration was formally introduced in Henley et al. (2015), which established the TPI as the standard representation of the IPO in the scientific literature.
Climatology from Jan/1971 to Jan/2000, following the interval used in SST datasets such as ERSST and HadISST (source of the TPI data provided by NOAA/PSL).
The calculation begins with determining the monthly SST anomalies obtained by subtracting the climatology (1971–2000) from each point of the ocean grid. This procedure removes the long-term mean and isolates the actual variability of the system, as described in Henley et al. (2015) and other studies that use the index.
After computing the anomalies, the spatial average for each of the three regions is obtained. Region 2 represents the tropical component of the IPO, while Regions 1 and 3 represent its subtropical components, as defined in Henley et al. (2015).
The monthly TPI index is then defined by the expression:
TPI = SSTRegion 2 − ( SSTRegion 1 + SSTRegion 3 ) / 2
This formula captures the characteristic tripolar pattern of the IPO, in which the tropical Pacific exhibits anomalies opposite to those found in the subtropical bands, as detailed in Henley et al. (2015).
Thus, positive TPI values indicate that the tropical Pacific is warmer relative to the subtropical regions, characterizing the positive (warm) phase of the IPO; negative values indicate the negative (cool) phase, as interpreted in climate variability studies that use the TPI to monitor the IPO.