Resolution dependence of southern Atlantic Ocean stratocumulus decks
Jacopo Canton,Anurag Dipankar,Christoph Schär
DOI: https://doi.org/10.1002/qj.4874
2024-10-18
Quarterly Journal of the Royal Meteorological Society
Abstract:Depiction of the computational domains employed in the present analysis. The purple (dashed) rectangle represents the intermediate domain at 4.4 km resolution. The pink (continuous line) square represents the main study domain with resolution varying between 4.4 and 0.55 km, which is the focus of this study. The cyan (dashed) domain is used for verification. (a) A satellite picture taken on one of the simulated days (August 6, 2016) by the NASA/NOAA Suomi National Polar orbiting Partnership satellite. (b, c) Illustration of the low cloud coverage simulated by the 4.4 km model in the intermediate domain, ranging from no clouds (0%, blue) to totally covered (100%, white). Panel (b) presents an instantaneous snapshot taken on the same day as the satellite photograph in panel (a) (August 6, 2016 at 0800 UTC), and panel (c) represents the cloud coverage averaged over the seven simulated days. Oceanic stratocumulus decks of clouds are among the largest contributors to the Earth's radiation budget, covering around a fifth of the planet's surface and reflecting a large part of the incoming solar radiation. Unfortunately, these clouds are not well represented in modern climate models, resulting in one of the leading causes of uncertainty in climate change projections. This contribution analyses this issue from a novel perspective and sheds light on the mechanisms behind the misrepresentation and resolution dependence of these large clouds. The analysis is based on realistic week‐long simulations performed over a 563×563 km2 oceanic domain. Four horizontal resolutions, between 4.4 and 0.55 km, are employed, resulting in a timely investigation, especially in light of the high resolutions employed by present and near‐future climate projections. Results show that the liquid cloud water, the main contributor to the simulated grid‐scale clouds, decreases with a power‐law decay as the resolution increases, whereas the water vapour, responsible for subgrid‐scale clouds, is much less affected by the grid spacing. The leading cause is identified as an imbalance between the rates of change of the advection and turbulence parametrisation terms. In order to verify this observation and provide a possible mitigation to the issue, a second set of simulations is performed where the turbulence parametrisation is tuned. The strategy proves to be successful, confirming the hypotheses and resulting not only in a resolution‐independent radiation budget but also cloud coverage.
meteorology & atmospheric sciences