Andrew Barrett: Research

PRESTO: Stationary convective rainbands

Andrew Barrett is currently working on orographic precipitation, and specifically stationary convective orographic rain bands of precipitation as part of the PRESTO project. I am attempting to understand the physical mechanisms through which bands of orographic convection are maintained through periods of many hours. These bands are often fixed in location and produce heavy but localised rain. Due to their persistent nature these bands can significantly increase the amount of precipitation falling locally which may lead to flooding. Many recent flooding events in the UK have resulted from quasi-stationary and/or orographically enhanced rainfall (e.g. Boscastle 2004, Carlisle 2005 and Cockermouth 2009) and a better understanding of the physical mechanisms will help predict these events in the future.

PhD Title: Why can't models simulate mixed-phase clouds correctly?

Mixed-phase clouds contain both supercooled liquid water and ice particles. Traditional understanding of microphysical processes suggests that when ice particles exist in liquid saturated conditions that the ice particles will grow at the expense of the liquid droplets through the Bergeron-Findeison process. This understanding suggests that mixed-phase clouds should be rare and short-lived. Contrary to this expectation, we observe that mixed-phase clouds are common and that thin supercooled liquid layers are found at the top of stratiform ice clouds. Because of this structure, and that liquid droplets are typically smaller than ice particles, these clouds have a significant radiative impact, but worryingly they are not adequately simulated by state-of-the-art weather and climate models. Research has been conducted, initially by building a new 1-D numerical model (EMPIRE) based around the UK Met Office Unified Model, and then increasing the complexity of the model to allow mixed-phase clouds to be correctly simulated. This led to the discovery that turbulent mixing in these clouds is important to allow the supercooled liquid to persist and that these clouds are sensitivity to ice growth processes which need to be modelled correctly. Additionally, the vertical resolution of the model has a significant impact, with mixed-phase clouds occurring more frequently in high resolution models. Models with vertical grid spacing greater than 250 metres rarely form mixed-phase clouds; worryingly this is the best attainable resolution in operational forecast models, with many forecast and climate models having much coarser resolution. A parameterization that allows mixed-phase clouds to form and persist in coarse resolution models has been created and the performance of the parameterization in EMPIRE is very promising and now needs to be applied to weather and climate models.

Figure: A wordle of my thesis text. The most commonly used words are larger.