Strong influence of tomographic models on geoid prediction: Case studies from Indian Ocean and Ross Sea geoids

Abstract

The geoid over the north Indian Ocean has a significantly large negative amplitude, even if the excess flattening of the Earth beyond its equilibrium shape is considered. The various mechanisms proposed for this geoid anomaly vary and acceptable geoid predictions are obtained for specific tomographic models only. In this study, we identified model-independent features in the mantle beneath the Indian Ocean and Ross Sea region by analyzing eight recent global tomographic models. We standardized each of the models and applied cluster analysis to regionalize geophysically significant features in various depth ranges. These regionalizations are compared grid-by-grid to construct vote-maps. They highlight the anomalous features consistent across models and their approximate dimensions. Low velocity anomalies of dVS ∼-1.1% in the ∼400–680 km depth range are consistent in almost all the models beneath the Indian Ocean and Ross Sea. High velocity anomalies of dVS≥1% at depths below 1600 km, incoherent in dimension and orientation, are also observed. High velocity anomalies are most likely subducted slabs while low velocity anomalies could be partial melts generated by hydration of mantle. Additionally, a consistent low velocity structure is seen throughout the mantle beneath the southwestern Indian Ocean and east Africa. It is mostly likely a plume rising from the African LLSVP. It connects to the probable partial melts beneath the Indian Ocean via a remnant trail. Forward modelling of the geoid using vote-maps reveals that the E-W extent of the Indian Ocean Geoid Low is precisely reproduced by the consistent low velocity anomalies in the upper mantle. However, the N-S extent, most likely dependent on lower mantle anomalies, is not expressed since the dimension, orientation and dVS characteristics of high velocity structures are inconsistent in different models. The inter-model agreement is insufficient to identify structure(s) seen across models that can explain the geoid low over the Ross Sea.