Breaking the tropical convection "dead-lock": Scale interactions of deep convection and tropical circulation

This project will analyse the properties of tropical convective storms and larger-scale tropical waves, in state of the art models and in observations. The work will advance our understanding of the Earth's climate system, through analysis of scale interactions in tropical atmospheric dynamics. Working closely with the Met Office and making use of their new "k-scale" modelling capability, we will address long-standing problems in the understanding of climate circulation.

The interaction of moist convection with larger-scale circulations is a limiting factor for predictions across time scales. The coupling of convection with circulation under climate change is highlighted as a grand challenge in the "Clouds, Circulation and Climate Sensitivity" Grand Challenge of the World Climate Research programme (WCRP). Past projects have shown the value of explicitly modelling convection over large domains in allowing new insights to this difficult problem.

The tropics are the "engine room" of Earth's climate: solar heating is maximised in equatorial regions at the land and ocean surface, and transmitted higher into the atmosphere by convection. Tropical convection is dominated by cumulonimbus storms that not only deliver rainfall, but generate intense atmospheric heating, which is then communicated to the wider atmosphere by waves. For decades the representation of these moist convective storms has provided a "deadlock" in weather and climate modelling, as it has been impossible to model them directly in a global model due to their small scale. This is one major reason why skill of numerical weather prediction is low in the tropics, and climate change projections of rainfall are very uncertain, for major climatic systems like the monsoons. Increased computer power is now removing this "dead-lock".

Tropical convection is intrinsically chaotic, but some predictability and organisation of the convection is provided by larger-scale propagating waves, by interaction with continental-scale circulations such as monsoons, and through processes at the land and sea surface. Understanding how moist convection interacts with larger scale flows is made much more challenging by the fact that the cumulonimbus storms that generate ascent and rainfall in the tropics are not explicitly resolved in global models, which have grid-spacings of approximately 10s or 100s of kilometres. This means their effects are normally represented by simplifications known as parametrisations, which lead to significant errors in the way that storms interact with their environment. For the first time, computational power enables us to run global or pan-tropical simulations that have a small enough grid-spacing (approximately 1 to 5 km) to explicitly capture these storms.

The project will address the following scientific objectives:
(1) How moist convection interacts with tropical waves, how this is affected by the representation of convection, and the implications for predictions.
(2) How moist convection influences the water and energy balance of large regions.
(3) Evaluating the role of clouds and convective processes in a changing climate.

We anticipate that the work will include:

Analysis of cloud and storm properties in the k-scale simulations, and their relationship to "drivers" in large-scale winds and thermodynamics;
Comparison of modelled behaviour with satellite and surface-based observations;
Diagnosis of tropical wave modes and their modulation by the convection;
Use of properties such as vorticity to quantify "scale interactions" between convection and waves;
Focus on key aspects of storm life-cycles, such as initiation, intensification, propagation and circulation.

The results will guide the development and use of the weather and climate prediction models of the future. By capitalising on unique new simulations from the Met Office's operational model.