Post-Graduate Showcase

Date: Friday 21 February 2020

Time: 18:00 - 19:00

Location: 

University of Edinburgh
School of GeoSciences
The University of Edinburgh, Grant Institute
The Kings Buildings, West Mains Road
Edinburgh
EH9 3JW
UK

Email: 

scotland@rmets.org

SPEAKER | Ms Sarah Wilson-Kemsley, Uni. East Anglia

TITLE | Stochastic Weather Generators and the Koppen Classification System

Stochastic weather generators are statistical tools used for generating realistic, synthetic time series of weather under current or future climate conditions. Modelling precipitation accurately is of particular importance as many weather generators condition other climatic variables on precipitation status. Input parameters of these generators are calculated using past observations. Subsequently, these input parameters will be perturbed to represent weather characteristics under a changed climate.

As such, stochastic weather generators following a Markovian approach are of particular interest as their input parameters (wet and dry transition probabilities) are believed to be simpler to perturb. From initial investigation, it appears that different climate regimes may be best described by different orders of Markov chain. The accuracy of generated precipitation sequences will be assessed globally across five basic climate regimes – tropical, humid, dry, continental and cold (using the Köppen climate classification system) – to determine which order of Markov chain best suits each regime. First, second and third order two-state Markov chains will be used in this comparison.

---------------------------------

SPEAKER | Mr Chris Manktelow, Uni. Exeter

Following The Forecast: The Communication of the Met Office's 3-Month Outlook

The 3-month outlook produced by the Met Office provides an indication of possible future temperature and precipitation conditions over the whole of the UK. It is potentially useful to contingency planners in government, business and industry, who can use the outlook to assess and manage seasonal weather-related risks. However, the 3-month outlook is a probabilistic forecast and is inherently uncertain. This makes the message of the 3-month outlook difficult to interpret and communicate. Existing social scientific research also suggests that different social groups often have different ways of judging the credibility and usability of seasonal forecasts, which further complicates the communication process. This research project will therefore aim to understand how the 3-month outlook is understood, interpreted and communicated by following it from its point of production to its point of use. Interviews with Met Office staff and users of the 3-month outlook, observations of stakeholder briefings and the textual analysis of relevant documents will be used to produce a detailed, qualitative account of the communication process. This will highlight how the message of the 3-month outlook changes as it is interpreted. It will also draw attention to the different views that people have on the purpose, credibility and usability of the 3-month outlook. The results of this research project will be relevant to anyone interested improving the communication of uncertainty in climate science & meteorology.