Cascade: Cloud System Resolving Modelling of the Tropical Atmosphere

Principal Investigator: Dr Doug Parker
Co-Investigators:  Dr Nick Dixon

This project uses state-of-the-art numerical modelling to explore the processes by which convective clouds are organized, on the time and space scales that dominate the weather and climate in the tropics. 

Convective clouds - typically very deep and intense cumulonimbus storms -constitute the fundamental building block of tropical weather systems and control the interactions of these weather systems with the global climate system. 

These storms contain organization on a wide variety of space scales, ranging from individual clouds on the scales of a few kilometres, to cloud clusters associated with synoptic-scale disturbances, such as tropical cyclones and African Easterly Waves on scales of 3000km, up to super-clusters or ensembles of clusters associated with planetary scale phenomena, such as the Madden Julian Oscillation and El Nino. 

Within these spatial scales, convection also varies on a range of timescales from the growth of individual clouds, over a few hours within the diurnal cycle, though synoptic timescales of a few days, to intraseasonal, seasonal and interannual modes of variability.

It is generally acknowledged that the influence of small spatial scales and high frequency time scales on the large scale, low frequency variability and vice versa, may be fundamental to the tropical climate system, and yet we know that these scale interactions are poorly captured in many weather and climate models. This upscale cascade of energy from the cloud-system and mesocale (from a few to a few hundreds of kilometres) to the synoptic and planetary scale (from a few hundreds to thousands of  kilometres) – in other words how convection organizes itself – remains poorly understood and is not addressed in any current convective parameterization.

It has now been possible, for the first time, to take a comprehensive approach to solving these problems. With the delivery of substantial increases in computer power, high resolution cloud-system resolving models  (CSRM), integrated over domains large enough to encompass synoptic and planetary scale systems, can be envisaged. At the same time, new remotely sensed (space borne and surface based) and in situ observations of clouds and precipitation are challenging our traditional views of the vertical heating profiles and structures of rain-producing clouds in tropical convective systems, as well as providing opportunities for evaluating these high resolution simulations. Both factors contribute to the expectation that radically different approaches to the representation of organized convection in weather and climate prediction models will emerge over the next few years, which will lead to major advances in model skill and in the potential for extended predictability on timescales from weeks to seasons.

This project is exploiting these opportunities, by drawing on the UK's expertise, including the Met Office, in modelling and observing tropical convection, in understanding continental and oceanic tropical weather systems, and in developing and testing new approaches to parameterization.

We are now performing simulations of convective clouds at sufficiently high resolution to resolve the individual cloud systems (1.5 km and 4 km), and over large enough domains (e.g. 4000 km in longitude by 1000 km in latitude) to allow the large-scale circulation to evolve in response to the convection. Our research focus is on specific processes and phenomena, such the the diurnal evolution of the clouds, and the way in which the winds and circulations, forced by the storms, feed back on the larger-scale weather.

Our results are also being informed by new observations, notably through remote sensing techniques being developed by colleagues at the University of Reading.

The modelling framework developed within this consortium has already provided the UK with a computational laboratory for exploring a much wider range of problems within atmospheric science: we are already applying these techniques to other projects such as AMMA and Fennec.

Sponsor: NERC

Start Date: 1 2007 
End Date: 31 January 2010

Value to Leeds: £330,514

Project page.