Does downscaling increase uncertainties?
In general, yes it does to a certain extent.
In detail
The uncertainty in UKCP09 is already quite wide at the 300 km scale
because there are several sources of uncertainty that contribute to the
spread e.g. internal variability, time scaling, modelling uncertainty,
structural uncertainty, and carbon cycle. Generally all components
contribute and none dominate.
Uncertainty accumulates by adding up the variances – a
consequence of this is that if there is already a large
variance and another variance of a similar magnitude is added the
uncertainty increases appreciably.
As the spatial scale gets more detailed, the interactions between the climate variables get
harder to predict and hence more uncertainty is potentially
introduced. However, downscaling from 300 km to >25 km makes a similar
uncertainty contribution to each of the other sources of uncertainty.
This uncertainty should be kept in mind when using the
weather generator which enables the user to analyse the 25 km
projections at a 5 km resolution. The outputs from the weather
generator are very uncertain and should be seen as only an indication
of the potential trend in daily climate for the future.
Find out more
|
