Detailed Analysis of Marine Gravity Survey Data from Panama Canal Transits: Improving Error Models and Signal Processing for BGM-3 Marine Gravimeter Survey Systems

Abstract:

Our prior work (Sailor et al., 2015) showed that the Panama Canal locking operations impart a peak vertical acceleration of about 60 mGal (6 x 10^-4 m/sec²) to ships as the individual lock chambers are filled or emptied. During a period of 8 to 12 minutes the ship’s elevation changes by over 8 meters. This motion is very repeatable, since it is driven by gravity-fed hydraulics backed up by a huge mass of water. The novelty of the prior work was to demonstrate that the lock-driven vertical acceleration is significant, of relatively long duration, easily observed by the BGM-3 accelerometer/gravimeter, and is equivalent tothe gravity anomaly caused by a moderately-sized seamount. Thus, the lock-induced vertical acceleration is a known external acceleration input that falls within the amplitude and time duration band of interest for marine gravity as well as airborne gravity survey systems.

Here we report an extension to the prior work, using BGM-3 gravimeter data from the RV Marcus G Langseth and the RV Melville, in addition to the previously-used two datasets from the RV Knorr. The new analysis allows us to compare the quality of the gravity data from these three ships in two ways, using: 1) Differences along nearly perfectly coincident gravity anomaly data profiles collected underway, during passage through calm and narrow channels with little or no vertical ship motion; and 2) Observed vertical-motion-induced accelerations, with no horizontal motion, experienced during lock operations. We use the raw 1-Hz output of the BGM-3 gravimeter and compare various filtering methods. Furthermore, good quality vertical channel GPS is used to compare to the output of our solution of a boundary value problem: Given the observed outputs of the gravimeter, solve for h(t), the elevation of the ship vs time and also for two parameters: initial gravity value prior to vertical motion in the lock and apparent vertical gravity gradient. This method will be further validated in Fall 2015 during the R/V Neil Armstrong's Pacific to Atlantic transit. A BGM-3 gravimeter will be permanently installed on the Armstrong prior to the transit and, during the transit, we will compare the results of our reported method with the previous canal transit data as well as from a variety height measurements made during lock operations.