Dr David Coley

University of Exeter

July 2008 to June 2011

AIM: To develop a new set of probabilistic reference years that can be understood and used by building designers.


  • Nationally agreed protocol for the creation of probabilistic future test reference years from either a weather generator or transformed historical data. Protocol to include wind direction.
  • Scientifically valid approach to the generation of wind direction and speed data for the building sector and a clear understanding of the limits of such data.
  • A large ensemble of probabilistic reference years for use mainly by the academic community/or an agreed method for using probabilistic data directly in simulation and design.
  • A clear understanding of the benefits of using probabilistic data and what form this should take (cf. Hanby et al. proposal).
  • Study whether climate predictions outside of UKCP09 ones give rise to different predictions of future building behaviour.
  • Understanding of the benefits of using data from weather generators to run multi-year simulations of buildings.
  • A smaller group of probabilistic reference years for use by practising engineers, distributed by CIBSE, compatible with common building thermal models and possibly part of future amendments to the building regulations.
  • A well researched understanding of the needs and hurdles faced by practitioners when using probabilistic results in the building sector.
  • Case-study based evaluation of opportunities for change within the buildings sector, the range of adaptation strategies probabilistic data leads to and the costs.


Probabilistic future weather files for various UK locations are available to download from the Centre for Energy and the Environment, University of Exeter.

These files have been created using the outputs of the UKCP09 Weather Generator, which uses the 2009 climate change scenario predictions and a gridded set of baseline data from the period 1961 to 1990. These files enable users to run weather through a buildings simulation model to predict behaviour in the future under a range of climate change scenarios as well as the control period. The weather files are presented in the .epw weather format, which is compatible with many building performance simulation programs.

The methodology for the creation of these files has been published in Building Services Engineering Research and Technology as On the creation of future probabilistic design weather years from UKCP09. Doi:10.1177/0143624410379934.