Why is the tropical sky white? Numerical investigations to elucidate the shape of mesoscale vertical velocity spectra.

Vertical motions are fundamental to atmospheric dynamics and our understanding of phenomena such as moist convection. A long-standing problem in atmospheric sciences is to understand the mesoscale energy spectra. Several numerical studies show that the vertical velocity spectrum has a homogeneous energy distribution across the mesoscales with a flat spectrum. Compared to the energy spectra of horizontal motion, the mechanisms that govern the spectrum of vertical velocity are less well known. In the troposphere, most of the horizontal mesoscale energy comes from divergent motions. At large scales O(100 km), vertical velocity relates, to a good approximation, to the vertically averaged divergence of horizontal motions by continuity in the incompressible limit. Recent measurements from NARVAL-2 (Next Generation Remote Sensing for Validation Studies) campaign conducted in the tropical Atlantic, unveiled that mesoscale horizontal mass divergence profiles possess a rich vertical structure and high spatio-temporal variability. Although the premise of a radiatively-balanced circulation holds on the long-term average, instantaneous deviations from this equilibrium occur in the form of wave-like oscillations. Numerical studies show that our state-of-the-art models can reproduce the observed variability in mesoscale divergence. We ask the following question in support of the previous arguments: What controls the spectrum of coherent mesoscale vertical motion? We aim to elucidate the mechanisms determining the homogeneous energy distribution across horizontal scales of vertical velocity spectra. This study designs numerical experiments, which include mechanisms-denial simulations employing the Icosahedral Nonhydrostatic (ICON) model. We conducted numerical simulations on a limited-area domain located in the western tropical Atlantic (4°S – 18°N, 64°W – 42°W). This domain has a horizontal resolution of 1.25 km and a lid at 30 km—the analysis period spans 48 hours. The experiments include the following: (i) a control run using DWD NWP physics configuration (ii) a dry atmosphere with all moist processes excluded along with the latent heat surface fluxes (iii) clouds invisible to radiation and, (iv) effects of saturation adjustment on temperature neglected while maintaining surface heat fluxes. Preliminary results show that the divergence profiles horizontally averaged over 200 km present a clear dominance of vertical wavelengths of 3 – 6 km. We found autocorrelation time-scales of around 4 – 6 hours increasing with altitude outside convective areas and consistent among all simulations. All experiments show a systematic decrease of about 50% in the temporal autocorrelation inside convective areas; therefore, moist convective processes modulate divergence's temporal variability. Moreover, we found that local moist processes contribute the most considerable fraction to the energy spectrum at scales < 200 km. The spectral response to moist processes is broad and extends into the free troposphere. The spectral response of surface fluxes instead is confined to the subcloud layer.