A temperature profile estimation technique using coastal acoustic tomography experiment

초록

A remote sensing measurement system using acoustic signals, called coastal acoustic tomography (CAT) system, has been developed and applied in regions where heavy ship traffics and/or fishing activities make in-situ oceanic sensor moorings quite difficult. However, the outputs of CAT so far are mainly depth-averaged horizontal temperature and currents velocity fields. Here, we demonstrate a new technique for vertical temperature profile estimation from CAT data using additionally available measurements such as sea surface and near-bottom temperature measurements. Simulations of the new technique are conducted using data-assimilated and tide-included high-resolution ocean general circulation model results in a southern coastal region of Korea, where two CAT stations have been moored to establish a continuous coastal ocean monitoring system. Four months (February, May, August, and November) representing four seasons are chosen for the plausibility test of the new CAT data analysis technique. Ray-tracing simulations of acoustic signals reveal that tides influence on the acoustic signal propagation during summer when the seasonal thermocline is well-developed. We also tested the availability of the second detected acoustic signals from a CAT data. Timeseries of estimated temperature profiles agree well with the original ocean model outputs with monthly-mean root mean squared differences about 0.5°C even with reasonable measurement errors included, which supports that the new temperature profile estimation technique proposed in this study is applicable enough to the in-situ CAT data.