Comparison of altimeter‐derived and ship gravity anomalies in the vicinity of the Gulf of California

Abstract

This article compares four sets of altimeter‐derived gravity anomalies in the Gulf of California region (22°N to 32°N, 244° E to 255°E) to anomalies from the EGM96 geopotential model and ship gravity data. Three anomaly sets were calculated using the same basic altimeter data (Ceosat geodetic mission data and the two cycles of ERS‐1, 168‐day repeat track) while one, by Hwang, Kao, and Parsons (HKP 1998), incorporated additional data not used in the other computations. The three sets using the same date were calculated by Knudsen and Andersen (KA 1997), Andersen and Knudsen (AK 1998), and Sandwell and Smith (SS 1997). Using a 1/8° × 1/8° grid of points, the mean difference between the altimeter anomalies and the degree 360 EGM96 geopotential model was computed in deep‐ and shallow‐water areas. The differences found reflect the different reference models used in the anomaly recovery process. Comparisons were also made with NIMA‐evaluated ship data, whose average accuracy was on the order of ±10 mgal. The comparisons were done with both shallow (—100 < d < —600 m) and deep (d < —2,000 m)‐waler data points. Comparisons were always better with the deep‐water points [for the AK set: ±7.0 mgal (deep water) versus ±11.1 mgal (shallow water)]. In all comparisons made, the AK set gave the best agreement with the ship data [e.g., ±9.7 mgal (AK), ±11.4 mgal (HKP), ±12.1 mgal(SS), ±10.9 mgal (KA)). The improved AK, over the KA set, reflects the use of the EGM96 model in the AK prediction while the KA set used the OSU91A model. Contour plots of the differences between the altimeter‐derived anomalies and ship data reveal some large differences near coastal regions, with the smallest differences found with the A K values. The causes of the various differences found can be associated with geophysical corrections to the altimeter data, altimeter data editing, data gridding procedure, and anomaly recovery methods.