Tropospheric gravity waves in the subtropics: Optical detection on low cloud decks

Internal gravity waves with wavelengths of tens to hundreds of kilometers are frequently seen as ripples on high-resolution geostationary satellite animations of subtropical stratocumulus decks. To systematically detect and characterize these waves, several satellite data fields are employed. Daytime visible reflectance images have high contrast, with units of reflected insolation relevant to climate impacts. IR brightness temperature is available day and night, but requires high-pass filtering and contrast enhancements. The divergence of low cloud tracking winds, retrieved via Particle Image Velocimetry (PIVdiv), is a scalar field independent of those radiative quantities. Water vapor channel time differences show wave vertical displacements at midlevels. 

In any given image array, Matlab’s Cauchy continuous wavelet transform detects packets of elongated phase crests and projects them into 10 logarithmic half-wavelength bins between 20-500 km, with angle discrimination of about 15 degrees, all on a 5 degree coarse geographical mesh. Cross-wavelet analysis probes for connections between pairs of images. Time pairs of the same field lead to estimates of wave propagation speed. Cross spectra of PIVdiv and radiative brightnesses help to quantitatively relate wave modulations of cloudiness to the vertical displacement of PBL top where the clouds reside. Connecting low level cloud signals to midlevel water vapor signals allows us to estimate vertical wavelength, allowing an independent check against propagation speed via the dispersion relation.

Preliminary wave rose maps, generated for the southeast Pacific during October–December 2023 reveal multiple source regions: synoptic jet-front disturbances in the South Pacific upper-level westerlies, intertropical convergence zone (ITCZ) convection, and orographic or thermal forcing from South America. We hypothesize that similar processes, plus tropical cyclones absent in this sample, drive similar wave activity in other basins and seasons. 

The results may have several applications. Any novel observed signal stands as a challenge or target for high-resolution models. Wave sources inferred from these observations may usefully constrain estimates of physical and nonlinear processes in the atmosphere. Low cloud dependence on vertical velocity could have climate relevance, for instance case studies of strong waves have shown they can be rectified in closed to open cell transitions. If periodic waves are trackable for much longer than their inverse frequency, they could comprise a subtle source of surprisingly long predictability of convective initiation, coastal fog/clearing, or other local effects. Like all gravity waves, these redistribute zonal momentum via meridional and vertical fluxes, a process whose contribution to larger scale flows can now be estimated quantitatively. 

By offering open-access wave data products, we hope to inspire collaborative efforts on all these application areas. By scaling up computations from 3 months in one region to many years around the globe, downgrading newer data to be comparable to older data as needed, we can build up a nearly global daily picture of tropospheric internal waves over the subtropical oceans through time. With so many degrees of freedom contributing to these high-resolution measurements, very subtle trends and differences should be detectable.