In the aftermath of the large Tohoku earthquake on 11 March 2011, a tsunami was launched across the Pacific Ocean. As the wave approached the islands of Hawaii, a wide-angle imaging system on the Haleakala Volcano operated by the University of Illinois in collaboration with Cornell University obtained the first-ever optical images of waves in the ionosphere/ thermosphere system linked to an underlying perturbation on the ocean. The ocean-ionospheric coupling associated with tsunamis had been predicted as early as the 1970s and had been observed previously by various radio frequency techniques, including large arrays of groundbased Global Positioning System (GPS) receivers [e.g., Liu et al., 2006; Lognonné et al., 2006; Rolland et al., 2010]. However, the observations of Makela et al. [2011] confirmed that these waves have an optical signature, opening up exciting new avenues for studying oceanatmosphere coupling mechanisms. These observations represented a breakthrough for two primary reasons. First, using a single wide-angle imaging system, a 1000°—1000 km2 region of the ionosphere/thermosphere system can be observed. In contrast, previous results using the GPS technique required arrays of receivers, and the quality of the resultant GPS “image” was directly related to the number of receivers in the array. As such, the coverage over islands with only a few GPS receivers, such as Hawaii, resulted in very poor coverage. Secondly, the demonstration of the ability to observe the tsunami signature in the airglow layers opens up the possibility of creating a detection system utilizing a satellite-based imaging system, which would provide continuous monitoring capabilities for large portions of the globe for tsunamis in real time.

Although these observations confirmed the existence of the tsunami airglow signature first predicted by Hickey et al. [2010], much work is still needed in order to fully understand the coupling mechanism and allow for the development of an effective satellite-based tsunami detection/warning system. These challenges include the need for: 

additional observations of the tsunami ionospheric signature to better constrain the conditions under which ocean-atmospheric coupling is effective;
creation of efficient detection algorithms to analyze optical data in (near) real time in order to detect tsunami signatures in the airglow; and
development of a more realistic ocean-atmosphere coupling model that can be used to study the physical mechanisms responsible for the coupling as well as to simulate potential viewing modalities and test detection algorithms.

The observations described in Makela et al. [2011] were made using a wide-angle imaging system located atop the Haleakala Volcano on Maui, HI. The pertinent observations were obtained using a narrowband (2.0-nm FWHM) filter centered at 630.0 nm, an emission sensitive to density and height fluctuations of the ionospheric layer. A detailed analysis utilizing a block of FIR filters was performed to extract the tsunami-related waves in the images, such as the one shown to the right, which reveals the presence of several tsunami-related waves as greyscale features. Overlain on this figure is a red curve showing the location of the ocean tsunami at the time of the image. The arrival directions, timing, periods, and velocities of the waves in the ionosphere all matched those of waves present in the underlying tsunami wave, confirming their linkage [Makela et al., 2011]. In addition, several detailed structures matched those predicted from a simple, non-viscous, three-dimensional gravity wave model [Occhipinti et al., 2011]. Through this single observation, we have confirmed that the wide-angle optical imaging systems we typically deploy to study ionospheric irregularities, such as equatorial plasma bubbles and medium scale traveling ionospheric disturbances, do indeed have the sensitivity to observe the small (5%) fluctuation in the 630.0-nm emission caused by AGWs generated by a tsunami. In addition, we have developed filtering techniques that can be applied to analyze the properties of the waves and compare them to the underlying properties of the source tsunami. However, this single example is not enough to effectively constrain the physics-based models required to understand the complex physical process at play, nor can it inform the development of more efficient analysis algorithms. Additional observations and more detailed data-model comparisons using more realistic models are, thus, required.