Bridging the Gap Between Weather and Climate: a Surface Transport Case Study

Oral Presentation 

Seamless modelling, whereby a single model family can be used across a range of timescales, has been at the heart of the Met Office strategy for weather and climate prediction since 1990. The Unified Model applies this approach, applying the same dynamical core and (where appropriate) parameterization schemes across a broad range of spatial and temporal scales. However, despite this unified approach in our core weather and climate modelling, approaches to downstream impact modelling remain largely inconsistent across timescales. This is because, historically, different approaches have been required due to large differences in resolution between weather and climate models. However, the advent of local-scale climate projections matching the spatial and temporal resolution of operational weather forecast models, along with the Met Office’s unique position as both a weather and climate centre, opens up the opportunity to apply consistent methodologies across all timescales. 

One example of historically disjointed approaches across different timescales is the approach to predicting weather-related hazards on road and rail networks (surface transport forecasting). At the Met Office, operational NWP-timescale surface transport forecasts are produced by a physical surface-exchange-scheme model, which takes in forecast outputs from NWP models and evaluates the surface conditions at a list of pre-defined locations. Output from this model is routinely delivered to customers across a variety of sectors, including traditional road forecasting, aviation and defence runway forecasting, and national rail infrastructure forecasting, and is also used to support the National Severe Weather Warning Service. In contrast, climate-timescale studies have traditionally assumed a simple linear relationship between air temperature and road/rail surface temperature. Now, we are taking a new approach, applying the 2.2km UKCP18 local climate projections to our operational impact models. This talk will discuss the outcomes of this study, alongside more general discussion on the benefits of our new approach, and how this may be extended to other impact models in future.