| |
|
|
What are the sources of uncertainty considered in UKCP09?
There are three sources of uncertainty that are explicitly considered in UKCP09:
1) natural internal climate variability,
2) uncertainty in the climate models, and
3) uncertainty in future emissions.
The first two of these sources are addressed by using a large ensemble of variants of the Met Office Hadley Centre HadSM3 model to run a series of perturbed physics experiments, and using single projections from other IPCC climate models. The third source is addressed by generating separate probabilistic projections for each of the three emissions scenarios.
In detail
There are various sources of uncertainty in the UKCP09 climate
projections, however, three principle sources of climate science
uncertainty are explicitly considered: 1) natural internal climatic
variability, 2) uncertainty in the climate models, and 3) uncertainty
in future emissions. In UKCP09, the Met Office Hadley Centre climate model HadSM3 (which is HadCM3
but with a simplified ocean model – the slab model) is used to address
some of these issues, whilst acknowledging that it is itself, a source
of uncertainty.
- Natural climatic variability: this source of uncertainty is incorporated in the probabilistic projections
of UKCP09 by running a series of model simulations using HadSM3, with
different initial starting conditions for the climate model parameters.
- Modelling uncertainty:
arises from an incomplete understanding of the physical processes of
the climate system, and, ability to model or represent the physical
processes adequately, leading to differences between different climate
models. In UKCP09, uncertainty arising from incomplete understanding of
the physical processes is considered, by using HadSM3 in a series of
perturbed physics ensembles (PPE), to generate a 300 member ensemble of possible future global climate. This ensemble is then used to train a climate model emulator,
with which an even greater range of variation in HadSM3 model
parameters is explored and quantified. Each member of the ensemble is
assessed as to its quality, with respect to how well it models recent
climatic observations and evolution. This assessment of model quality
is used to develop a weighting scheme, whereby the most accurate model
simulations receive a greater weighting than less accurate ones, in
establishing the probability of a particular outcome. To account for
uncertainty that arises from the ability of one model to provide an
accurate representation of the real world, thirteen additional IPCC
models, are used to generate a multi-model ensemble (MME). The uncertainty from the PPE and MME are then combined with downscaling and timescaling uncertainty in a statistical framework, to arrive at a final assignment of probability to each future climate projection.
- Emissions uncertainty: climate projections are provided in UKCP09 based on three different assumptions of greenhouse gas
emission scenarios: Low, Medium and High, which respectively correspond
to B1, A1B, and A1FI, of the IPCC special report on emissions scenarios
(SRES).
|
|
You may also be
interested in…
|
|