Grant: Discovery Grants Program - Individual

Climate models are used to predict changes to increases in greenhouse gases, for seasonal weather forecasting, understanding past climate change, and as a platform for improving weather forecast models. Their greatest weaknesses are associated with processes involving clouds that they cannot resolve. Most climate models divide the atmosphere into boxes of roughly 100 km in the horizontal and 1 km in the vertical. These boxes are too large to capture motions within most clouds. Convective clouds such as thunderstorms produce most of the rain in the tropics and in mid-latitudes during summer. Though large in the upper troposphere, they are fed by updrafts whose diameters are less than a few kilometers. As a result, the horizontal and vertical motions of these clouds are not resolved by climate models. Instead, climate models have used what are called parameterizations. Parameterizations are necessary because a climate model would fail or become unphysical in their absence. However, the way that parameterizations represent small scale processes such as clouds is incomplete. The problem of how to include convective clouds in climate models has continued since the first climate models fifty years ago. Although it will never be fully solved, progressively better approaches are evolving as computer models improve and more observations of convective clouds become available. The focus of my research program is on the use of convective organization variables in convective parameterizations. This involves taking some aspect of the resolved flow of the climate model, such as the upward motion in the lower troposphere, and using this variable to calculate how much convective rain is produced in a column. It is a short cut which exploits known observed relationships between larger scale motion and convective rainfall. This approach helps simulate how organized rainfall patterns propagate. This research has implications for our ability to address some of the most important challenges in climate modelling. These includes changes in the formation and intensification of hurricanes, the seasonal patterns of droughts and monsoons in the tropics, climate shifts associated with ENSO, and the Madden Julian Oscillation (MJO). The MJO is a large region of enhanced rainfall that moves eastward parallel to the equator in the Pacific Ocean. Improvements in our ability to forecast the MJO would result in significant improvements in midlatitude weather prediction.