Our new paper – published in Environmental Research Letters – systematically reviews and compares 88 scenarios of energy demand in commercial and residential buildings that include the additional energy use or savings induced by thermal adaptation in heating and cooling needs at global level. 

Our analysis then only includes the IAMs that have explicitly addressed heating and cooling needs. Our objective is to review the methodological approaches used in these resulting 14 papers and highlight their importance for the economy and the environment. 

Our novel classification makes it possible to systematically understand why the energy projections of Integrated Assessment Models’ vary depending on how changes in climatic conditions and the associated adaptation needs are modelled. 

On top of the typical categorization based on how IAMs represent the economy and the energy sectors, we analyze how these models deal explicitly with climate shocks transmitted to energy demand. Short-term demand responses to weather usually characterize both cooling and heating services in a similar way, and are called intensive margin. While long-term demand responses driven by an increase in the penetration of air conditioner appliances are called extensive margin. 

According to our classification, five different approaches can be identified based on how IAMs combine the modeling of the economy, the energy sector and the climate-energy feedback.

Our results clearly underscore the importance of incorporating in the energy demand functions the effect of climate change adaptation: across the studies reviewed, already by 2050 climate change will induce a median 30% (90%) variation of building’s energy demand for cooling and a median -8% (-24%) variation for heating, leading to an net 2% (13%) increase in thermal adaptation energy demand, under the Representative Concentration Pathway 1.9 (8.5).

Results vary widely across model Types: for instance, models lacking extensive margin adjustments highly underestimate the additional cooling needs of the building sector. Our review also highlights the much larger uncertainty that characterizes the commercial sector, which often, due to the lack of specific data or evidence, is modelled similarly to the residential sector.

This comparison exercise allows us to shed light on how much IAM’s assumptions and mechanisms influence their projections, and to identify the main modeling gaps that should be addressed by future studies. The 88 scenarios analysed are available freely for download in a unique dataset (when the definitive version of the paper is out).