There is a range of uncertainties associated with future climate conditions. These include:

  • how world development will progress, so we do not know future levels of greenhouse gases and aerosols in the atmosphere
  • the Earth’s climate system’s sensitivity to increasing levels of greenhouse gases and aerosols in the atmosphere
  • local climate response to global climate change
  • how natural climate variability (associated with events such as volcanic activity and variations in the amount of incoming solar radiation) will vary, for instance, from year-to-year and decade-to-decade.

Impact assessments may also have to consider other uncertainties, such as those related to environmental and socio-economic factors. These factors may contribute to risk as stressors and modifiers of impact, with resulting consequences.

The ‘top-down’ approach to developing and/or applying future anthropogenic climate change information introduces uncertainty at each step of the process.

If we assume the four representative concentration pathways (RCPs) and 40 global climate model (GCM) simulations are all plausible, this means there will be 160 possible climate futures to account for. In this context, the practice of using just one GCM simulation run by one RCP, for instance, will not suffice to cover the range of uncertainty and therefore is not recommended.

It is recommended to consider all sources of modelling uncertainties, as much as possible, in developing or applying future climate information. To assess uncertainty in projections, it is necessary to understand the assumptions made in the chosen scenarios and to validate the models’ ability to simulate current climate. Evaluation of the climate projections selected for the study can also reveal if the fields are consistent and seem reasonable in light of knowledge of physical processes that occur in the atmosphere and biosphere.