Statistical tests of additional plate boundaries from plate motion inversions

Abstract

We have investigated the application of the F-ratio test, a standard statistical technique, to the results of relative plate motion inversions. The method tests whether the improvement in fit of the model to the data resulting from the addition of another plate to the model is greater than that expected purely by chance. This approach appears to be useful in determining whether additional plate boundaries are justified. We confirm previous results favoring separate North American and South American plates with a boundary located between 30°N and the equator. Using Chase's global relative motion data, we show that in addition to separate West African and Somalian plates, separate West Indian and Australian plates, with a best-fitting boundary between 70°E and 90°E, can be resolved. These results are generally consistent with the observation that the Indian plate's internal deformation extends somewhat westward of the Ninetyeast Ridge. The relative motion pole is similar to Minster and Jordan's and predicts the NW-SE compression observed in earthquake mechanisms near the Ninetyeast Ridge.