The effects of seasonal temperature variation on sound speed contrasts at the seafloor are usually considered negligible in the analysis of seismic data but may be significant at large incidence angles (offsets) important for inversion of sediment elastic properties, or long-range acoustic transmission. In coastal areas, the maximum annual seafloor temperature variation can be several degrees Celsius or more, corresponding to a sound speed variation of 30 m/s or more. Thermal pulses propagate via conduction several meters into the seafloor resulting in a damped quasi-sinusoidal temperature profile with predictable wave number characteristics. The oscillating seasonal and spatial character of this signal creates a time-and frequency-dependent effect on the elastic seafloor reflectivity. Results of numerical simulations show that the expected temperature profile for most sediment types and porosities will have the strongest affect on frequencies between about 60 and 600 Hz, at incidence angles greater than about 50°.
BackgroundSeafloor temperature in coastal areas can vary over the course of a year by several tens of degrees Celsius. This seasonal temperature variation (STV) is well documented in the World Ocean Atlas compiled for 2009 [Locarnini et al., 2010], which lists mean ocean temperatures as a function of geospatial position, depth, and month of the year. Figure 1 shows the global magnitude of the variation (warmest minus coolest monthly average) in gray scale; white to black corresponding to 0 to 8°C, respectively. The inset graph positioned over Asia shows the average and maximum STV for the entire seafloor as a function of water depth. Note that even at depths of 200 m, the variation can be as much as 8°C (±4°C about the mean). The other insets in Figure 1 show the actual variation at selected locations (arrows in Figure 1) in°C versus the 12 months of the year. The insets show that the seafloor requires somewhat longer to warm than to cool, and in the Northern Hemisphere, the highest temperature is found in September to October. Most importantly, Figure 1 shows significant geographical variability, suggesting that each location is unique. There appears to be no simple pattern that would allow accurate estimates of STV as functions of, for example, water depth or latitude.Everywhere on the seafloor the local STV propagates into the seafloor, causing a damped oscillatory temperature and sound speed profile [Jackson and Richardson, 2001]. In high-porosity coastal sediments, the STV penetrates several meters. The effects of temperature on sound speed in the water column are extremely well known [Fofonoff and Millard, 1983], as are the physics of heat and sound propagation in common sediment types [Carslaw and Jaeger, 1959;Kennett, 1983;Dvorkin et al., 1999]. Our intent here is to couple these very mature techniques to make new predictions of acoustic seafloor interaction.We will show later that the frequencies and incidence angles that are affected most by seasonal temperature variation are about ...