Project

Discipline

Climate modeling; Convection; Large-eddy simulation; Turbulence closure; Turbulence ; Clouds ; Subgrid-scale parameterization

Lead
Die grössten Unsicherheiten in Klimaprojektionen stammen daher, dass es unklar ist, wie sich Wolken mit dem Klima verändern. Niedere Wolken modulieren den Strahlungshaushalt der Erde primär indem sie Sonnenlicht reflektieren. Vergrösseren sich niedere Wolkenfelder, dämpft das die globale Erwärmung. Verkleinern sie sich, verstärkt das die globale Erwärmung. Ziel dieses Projektes ist es, die Repräsentation von Wolken in Klimamodellen zu verbessern und so Unsicherheiten in Klimaprojektionen zu reduzieren.
Lay summary
Wolken in Klimamodellen zu repräsentieren ist schwierig, weil die Grössenskalen der Wolkendynamik (zwischen Metern und Kilometern) viel kleiner sind als die Auflösung von Klimamodellen (ungefähr 100 km). Deshalb muss man Wolkendynamiken vereinfacht und empirisch in Klimamodellen darstellen. Grosse Fortschritte dabei sind jetzt möglich: Dank moderner Hochleistungsrechner ist es jetzt einerseits möglich, rechnerisch aufwändigere Wolkendarstellungen in Klimamodellen zu benutzten als die derzeit üblichen. Andrerseits ist es auch möglich, Wolkendynamiken detailgetreu und zuverlässig in begrenzten Regionen (etwa 100 km breit) zu simulieren. Dieses Projekt wird neue und effizientere Methoden entwickeln, Wolken im Detail zu simulieren und wird diese Simulationen nutzen, verbesserte Darstellungen von Wolken für Klimamodelle zu entwickeln.

Direct link to Lay Summary

Last update: 26.09.2014

Self-organised

Number	Title	Start	Funding scheme

The largest contributor to uncertainties in climate change projections are clouds and particularly low clouds. If low cloud cover increases as the climate warms, the increased planetary albedo implies a mitigating feedback on climate changes; if low cloud cover decreases as the climate warms, the reduced albedo implies an amplifying feedback. Existing theories and models do not agree on the sign or magnitude of low cloud cover changes as the climate warms. Representing low clouds in climate models is difficult primarily because the turbulent dynamics governing them have scales of meters, while the resolution of climate models typically is of order hundred kilometers. Thus, cloud-scale turbulent dynamics must be represented parametrically in terms of the resolved large-scale dynamics of climate models. Approaches to do so generally represent second-order turbulent fluxes in terms of first-order quantities, often using second-order equations to estimate closure parameters that arise in the flux closures. For example, when turbulent fluxes are represented as diffusion down mean gradients, diffusivities are often estimated with the help of a (second-order) turbulent kinetic energy equation. But the second-order equations typically are local in space, disregarding non-local spatial correlations that we know to be important, for example, in convective plumes.Now is the time for fundamental progress on the problem of representing cloud-scale turbulent dynamics in climate models, for three reasons. (1) We are in a golden age of Earth observations, with space-based measurements that give us unprecedentedly detailed data. (2) Large-eddy simulations (LES) have become reliable enough to faithfully simulate cloud-scale turbulent dynamics (modulo microphysical processes such as raindrop formation). (3) High-performance computing (HPC) has reached the critical point at which we can conduct statistically steady LES of cloud-scale dynamics in domains the size of a typical climate model grid box. The research we propose will leverage these three advances to develop new and unified representations of turbulence in clouds and boundary layers, with the long-term goal to reduce uncertainties about cloud feedbacks in climate change projections. We are proposing a three-pronged approach: (1) We will develop a new LES code and validate it with observational data. (2) We will then use LES with representations of large-scale dynamics to obtain statistically steady states of clouds and boundary layers. We will vary the climate and region represented systematically, to study physical mechanisms of how clouds are maintained and respond to climate changes. (3) We will develop new and unified closures of the turbulent dynamics of clouds and boundary layers for large-scale climate models, using LES to test them systematically. First, the development of closures will focus on eddy diffusion/mass flux approaches, which hold promise for unified representations of all subgrid-scale turbulence in climate models, from boundary layers to deep convection. Then, we will use more sophisticated approaches in which the closure of turbulent dynamics is achieved at second order and non-local spatial correlations are retained explicitly. This obviates the need, for example, to specify entrainment rates---which are uncertain and contribute to uncertainties about the cloud response to climate changes. Such approaches have the promise to deliver more accurate closures, with reduced needs to specify arbitrary closure parameters, at the expense of increased computational cost, which, however, can be borne with today's HPC platforms.This research will be directed by Professor Tapio Schneider and will involve one PhD student and a postdoctoral fellow with extensive experience in high-performance computational fluid dynamics.