Inferring submesoscale energy spectra in the Gulf of Mexico from surface drifters

Abstract Calculating energy spectra in the ocean is impeded by the resolution limitations of satellite and shipboard ADCP data. An alternate approach is to use data from GPS-tracked surface drifters, which offer resolution to finer than 10 m. While direct spectral estimates from drifter velocities have been unsuccessful, indirect estimates, based on the dispersion of pairs of drifters, are possible. Using drifter data from the Gulf of Mexico, we infer wavenumber spectra by comparing pair dispersion to that expected with different spectral slopes. The measures include both time- and separation-based metrics, and establish consistency between the two using the recently proposed Cumulative Inverse Separation Time (CIST), a separation-based measure similar to the Finite Scale Lyapunov Exponent (FSLE). The results indicate a K −2 dependence from 100 m to several kilometers. The spectral slope is the same up to the deformation radius (~ 50 km) in winter, while in summer it is steeper, close to a K −3 dependence. Additional separation-based metrics suggest a shallower K −5/3 spectrum in the 10-100 m range, but this cannot be confirmed with the time-based metrics. The results are consistent with previous drifter studies and help resolve differences between them. Differences between “original” and “chance” pairs are seen at the smallest scales (< 1 km), where the dispersion is affected by the initial conditions. Furthermore, the FSLE fails to capture the smallest scale dispersion, while a modified version, the Finite Amplitude Growth Rate (FAGR), is more successful. The findings will help guide future studies of submesoscale dispersion.