Acoustic Detection of Greenhouse-induced Climate Changes in the Presence of Slow Fluctuations of the Thermohaline Circulation

View More View Less
  • 1 Max-Planck-Institut für Meteorologie, Hamburg, Germany
  • | 2 Scripps Institution of Oceanography, La Jolla, California
© Get Permissions Rent on DeepDyve
Restricted access

Abstract

Munk and Forbes have proposed to detect greenhouse gas-induced climate changes in the World Ocean with an array of long-range acoustic transmissions from Heard Island in the southern Indian Ocean. We estimated–assuming a continuously monitorable simplified axial ray propagation–the signal-to-noise ratio for such an experiment in an environment of slow fluctuations of the thermohaline circulation on a decadal time scale. The signal and noise are obtained from two coarse-resolution ocean general circulation model simulations. In the first, prescribed greenhouse atmospheric anomalies forced the ocean and yielded rough estimates of ocean response to greenhouse warming. In the second, some aspects of low-frequency internal variability of the ocean were obtained by stochastic forcing of the same ocean model. By this technique, no oscillations of the coupled ocean-atmosphere system Eke, for instance, El Niño-Southern Oscillation (ENSO) could be stimulated. Both signal and internal variability proved to be strongest at high latitudes, where the depth of the sound channel is small. At lower latitudes the signal is relatively weak, except for the western Atlantic. An array with an acoustic source near Heard Island would monitor primarily temperature changes in the near-surface layers of the Southern Ocean rather than in low-latitude intermediate depths.

The trend detection probability for any single path came out to be weak, at least for a one-decade measuring interval. But using information from at least a two-decade interval and an array of receivers improved the detection probabilities substantially. Two different pattern detection strategies were tested: projecting the natural variability on the expected greenhouse signal and projecting the greenhouse signal onto the major components of the natural variability. Both techniques proved to give almost identical results.

Abstract

Munk and Forbes have proposed to detect greenhouse gas-induced climate changes in the World Ocean with an array of long-range acoustic transmissions from Heard Island in the southern Indian Ocean. We estimated–assuming a continuously monitorable simplified axial ray propagation–the signal-to-noise ratio for such an experiment in an environment of slow fluctuations of the thermohaline circulation on a decadal time scale. The signal and noise are obtained from two coarse-resolution ocean general circulation model simulations. In the first, prescribed greenhouse atmospheric anomalies forced the ocean and yielded rough estimates of ocean response to greenhouse warming. In the second, some aspects of low-frequency internal variability of the ocean were obtained by stochastic forcing of the same ocean model. By this technique, no oscillations of the coupled ocean-atmosphere system Eke, for instance, El Niño-Southern Oscillation (ENSO) could be stimulated. Both signal and internal variability proved to be strongest at high latitudes, where the depth of the sound channel is small. At lower latitudes the signal is relatively weak, except for the western Atlantic. An array with an acoustic source near Heard Island would monitor primarily temperature changes in the near-surface layers of the Southern Ocean rather than in low-latitude intermediate depths.

The trend detection probability for any single path came out to be weak, at least for a one-decade measuring interval. But using information from at least a two-decade interval and an array of receivers improved the detection probabilities substantially. Two different pattern detection strategies were tested: projecting the natural variability on the expected greenhouse signal and projecting the greenhouse signal onto the major components of the natural variability. Both techniques proved to give almost identical results.

Save