Climate change projections Assumptions
All modelling methods incorporate assumptions. Although assumptions are part of UKCP09, it presents a range of projections that show some of those previously hidden, and addresses others, e.g. sampling uncertainty in the carbon cycle, aerosol chemistry and ocean dynamics.
- That known sources of uncertainty not included in UKCP09 are not likely to contribute much extra uncertainty
The modelling and statistical framework developed by the Met Office Hadley Centre to produce UKCP09 includes consideration of the most important known drivers of uncertainty in 21st century climate change including physical atmospheric processes and carbon cycle feedbacks. There are other possible forcing processes and feedbacks that are not included either because their importance is thought to be small relative to those process and feedbacks included in UKCP09 (for example, warming due to black carbon), or because they remain too poorly understood (such as the feedback between climate change and the methane cycle) to be included in a credible way.
- That structural uncertainty across the current state of the art models is a good proxy for structural error
Structural error (the difference between the real world and model projections), is incorporated into the UKCP09 projections by using the difference between the HadCM3 projections and the twelve other climate models as a proxy. Including this structural uncertainty prevents the projections from being too heavily biased by the way in which one model is structured. Using multiple climate models overcomes this issue to a certain extent, however, it cannot be ruled out that there are common systematic biases in all climate models, such that the discrepancy between the real and modelled worlds, may not be fully accounted for.
More information is provided in Chapter 3 of the UKCP09 probabilistic climate projections report.
- That models that simulate recent climate, and its recent trends well, are more accurate at simulating future climate
The basic assumption has been made that climate models that more accurately reproduce observed climate, and climate changes over the past decades, are considered to be more skilful in their projection of future climate. In practice, this means that different variants of the Met Office Hadley Centre model are assigned different weights according to how well the model simulates a set of historical climate and climate change observations.
It was judged to be better that each version of the Met Office Hadley Centre model should be assessed based on its ability to simulate climate globally, because studies show that large scale processes dominate local responses to forcing and to avoid the risk of rewarding models that well simulate local climate through some convenient chance compensation of errors. This means that the weighting is carried out based on global observations of historical climate and is not based solely on observations that relate specifically to the UK.
The choice of what observations to use in the weighting process, is restricted to those climate variables that are available on a global basis and which can be simulated by the ensemble of Met Office Hadley Centre global climate model versions and other international models. The seasonal mean observations relate to the baseline period of 1961–1990 and include variables typically used to assess climate models, such as sea surface temperature, land surface air temperature, precipitation, air pressure, radiative forcing and cloud cover. Trends in recent climate change is represented by four indices of temperature which capture the main patterns of the warming trend observed over the last century.
More information is provided in Chapter 3 of the UKCP09 probabilistic climate projections report.
The UKCP09 probabilistic climate change projections include single model results from other global climate models. These were selected from those that contributed to recent model inter-comparisons (such as that for IPCC AR4 ) and it was assumed that the results from all 12 models used, were equally credible. This was partly because there is currently no widely accepted method of quantifying the quality of single runs of different global climate models, and also to avoid the risk of over-constraining projections by double-counting the observations used to weight the Met Office Hadley Centre model variants. The assumption of equal credibility does carry the risk that too much consideration is given to those models that simulate climate relatively poorly. However, the inclusion of projections from other climate models is a necessary component of the methodology, as they provide information on the structural error, which is a part of the model uncertainty which cannot be captured using variants of the Met Office Hadley Centre model alone.
More information is provided in Chapter 3 of the UKCP09 probabilistic climate projections report.
It may be misleading to think of the projected change of a particular climate variable, as being representative of the whole 30-year time period. It is the average for that 30-year time period (as are the values for the baseline (1961–1990) 30-year time period).
For example, if there is a linear trend in the way precipitation amount is projected to evolve over time, then it is likely that the average value reported for a given 30-year period, is more likely to be observed in the middle than the first or last decade, of the 30-year period. However, because we do not know for sure what the form of the trend will be (and thus cannot assume linearity across the 30-year period), it is suggested that all of the 30-year overlapping periods that include the period of interest should be used. For example, if one were interested in the 2050s we would suggest to look at the average values for the two overlapping time periods of the 2040s and 2060s, in addition to the 2050s.
- That local carbon cycle feedbacks are small compared to the impact of the carbon cycle via change in global temperature
In the UKCP09 probablistic projections, feedbacks are considered at the global scale through the change in global temperature that would occur as a result of changes in the carbon cycle. The UKCP09 projections do not consider any local changes in the carbon cycle that might occur under this changed temperature. For the UK, this is likely to be a reasonable assumption, given that there is no major component of the carbon cycle that might change dramatically under a changing climate. This is in contrast to somewhere like Brazil, where a change in temperature beyond a certain level, may lead to the die back of the Amazon rainforest, which would clearly have considerable effects on the local carbon cycle.
More information is provided in Chapter 3 of the UKCP09 probabilistic climate projections report.
- Last updated: Sunday, 11 March 2012