Convection‐permitting models: a step‐change in rainfall forecasting. (29th March 2016)

• Record Type: Journal Article
• Authors: Clark, Peter; Roberts, Nigel; Lean, Humphrey; Ballard, Susan P.; Charlton‐Perez, Cristina
• Is Part Of: Meteorological applications. Volume 23:Number 2(2016)
• Page Start: 165
• Page End: 181
• Publication Date: 2016-03-29
• Subjects: convection‐permitting models -- high‐resolution NWP -- rainfall forecasts/forecasting -- explicit convection -- high‐resolution data assimilation -- resolved processes -- convective ensembles; Meteorology -- Periodicals; Meteorological services -- Periodicals; 551.5
• DOI: 10.1002/met.1538
• Languages: English
• ISSNs: 1350-4827

Abstract

ABSTRACT: Convection‐permitting models (CPMs) have provided weather forecasting centres with a step‐change in capabilities for forecasting rainfall. They are now used operationally to forecast precipitation in many parts of the world, including the UK. CPMs are models in which the dynamics of atmospheric convection is treated with sufficient accuracy in order to make it viable to switch off convection parametrization. This review describes the current state‐of‐the‐art in operational CPM‐based numerical weather prediction (NWP), primarily within the UK, and the historical development of CPMs. The characteristics of CPM systems and forecasts are highlighted and placed in an international context to recognize similar trends and highlight some differences. It is shown that the realism of CPM‐based forecasts can provide improved subjective guidance on convection, and, when measured on appropriate scales, can improve rainfall forecasting skill compared to coarser‐resolution NWP. Data assimilation techniques used with operational CPMs are reviewed and given historical context. Examples of new types of observations that may increase the skill of forecasts from improved initial conditions are discussed. CPM‐based nowcasting systems are shown to provide considerable improvements in short‐range forecasts of rapidly developing, intense systems. As a result, these CPM‐based systems provide a new forecasting capability. Finally, the development of CPMs has also required new techniques to verify forecasts and define their skill. These have revealed that the lack of predictability of the smallest scales involving convection means that ensemble techniques are required to represent forecast uncertainty, resulting in a new capability to provide objective forecast probabilities of local precipitation.