Implementation of Ionosphere and Troposphere Models for High-Precision GPS Positioning of a Buoy During Hurricane Katrina

摘要

In December of 2004, the University of Southern Mississippi (USM) deployed an ocean observing data buoy in the northern Gulf of Mexico. To facilitate the measurement of long-period vertical movement, such as tides, dual-frequency GPS receivers were installed on the buoy and at three shore stations. The shore stations were located at ranges of 20 km, 56 km and 100 km. The 20 km baseline was established to determine the true buoy position; the other two baselines were established to facilitate investigations into the mitigation of atmospheric effects (both from the ionosphere and the neutral atmosphere) on long-baseline, high-accuracy vertical positioning with ambiguity resolution. On August 29 of 2005, Hurricane Katrina came ashore along the northern Gulf of Mexico. The USM buoy survived the storm and continued to collect data until it was recovered, approximately 13 km from its original mooring. Of the three base stations, two collected data until they lost power due to storm effects, and the third (56 km baseline) was completely destroyed and all storm data was lost. The dry conditions (low tropospheric wet delay) of the week prior to the storm provided optimum GPS positioning conditions, which was in stark contrast to the high atmosphere moisture content (high tropospheric wet delay) leading up to and during the storm. Previous studies by the authors have focused on evaluating the individual utility of new tropospheric models (e.g., NOAA) or new ionospheric models (e.g., NOAA MAGIC and NOAA USTEC) to improve float ambiguity, static baseline positioning. In the current study, improved tropospheric and ionospheric models are used in combination to improve baseline processing. Kinematic buoy data are processed from the week of August 23 through 30, 2005, and long-baseline integer ambiguity resolution (AR) is attempted. An important question to answer is "Can decimeter-level AR positioning be achieved with the aid of improved atmospheric modeling? To further place these results into the context of contemporary positioning accuracy, processing solution comparisons are made between in-house software ingesting the external atmospheric models and COTS processing software, for short and long baselines, under static and dynamic conditions.