A Probabilistic Framework for Present and Future Design Outdoor Temperatures in Norway

Design outdoor temperature (DOT) is a critical parameter for building design, indoor climate regulation, and energy calculations. Historically, these values in Norway were derived from station-based observations with varying methodologies for summer and winter. This presentation outlines a transition to a modernized, unified framework for calculating DOT, as detailed in recent MET Norway reports.

In the first part we introduce a new probabilistic approach for current climate conditions (1991–2020). By utilizing high-resolution (1*1 km) operational gridded climatology, the method overcomes challenges related to sparse station networks and short observation series. The framework employs a Bayesian Generalized Extreme Value (GEV-Bay) distribution to calculate design temperatures for durations of 1 to 5 days and return periods ranging from 2 to 200 years. This approach provides a full distribution of risk rather than a single value, allowing for more robust decision-making in building construction.

In the second part we address the impact of climate change on these design values. Using national climate projections (Klima i Norge 2025), the study compares current DOT values with future estimates for two periods: mid-century (2041–2070) and end-of-century (2071–2100). Analyses for selected locations representing diverse Norwegian climates, indicate that future design summer temperatures are highly sensitive to (global) model selection and emission scenarios.

To facilitate practical climate adaptation, particularly for BREEAM (Building Research Establishment Environmental Assessment Method) certification, a simplified method is proposed. Preliminary results suggest that future global warming impacts can be approximated by adopting a higher return period from the current climate statistics. This provides engineers and architects with a pragmatic tool to account for future heat stress while awaiting more comprehensive national calculations.