A Comparison of the Pac-X Trans-Pacific Wave Glider Data and Satellite Data (MODIS, Aquarius, TRMM and VIIRS)

Four wave-propelled autonomous vehicles (Wave Gliders) instrumented with a variety of oceanographic and meteorological sensors were launched from San Francisco, CA in November 2011 for a trans-Pacific (Pac-X) voyage to test platform endurance. Two arrived in Australia, one in Dec 2012 and one in February 2013, while the two destined for Japan both ran into technical difficulties and did not arrive at their destination. The gliders were all equipped with sensors to measure temperature, salinity, turbidity, oxygen, and both chlorophyll and oil fluorescence. Here we conduct an initial assessment of the data set, noting necessary quality control steps and instrument utility. We conduct a validation of the Pac-X dataset by comparing the glider data to equivalent, or near-equivalent, satellite measurements. Sea surface temperature and salinity compared well to satellite measurements. Chl fluorescence from the gliders was more poorly correlated, with substantial between glider variability. Both turbidity and oil CDOM sensors were compromised to some degree by interfering processes. The well-known diel cycle in chlorophyll fluorescence was observed suggesting that mapping physiological data over large scales is possible. The gliders captured the Pacific Ocean’s major oceanographic features including the increased chlorophyll biomass of the California Current and equatorial upwelling. A comparison of satellite sea surface salinity (Aquarius) and glider-measured salinity revealed thin low salinity lenses in the southwestern Pacific Ocean. One glider survived a direct passage through a tropical cyclone. Two gliders traversed an open ocean phytoplankton bloom; extensive spiking in the chlorophyll fluorescence data is consistent with aggregation and highlights another potential future use for the gliders. On long missions, redundant instrumentation would aid in interpreting unusual data streams, as well as a means to periodically image the sensor heads. Instrument placement is critical to minimize bubble-related problems in the data.