Uncertainties involved in such CC projections
As
mentioned briefly in the previous post, certain degree of uncertainties in
Climate Change projections is inevitable. Although enhancements are being
made continuously, it is difficult to say that there is an absolute and/or
objective ways to assess different climate change projection models - so each
model has equal probability of getting the projection 'right'.
Some scientists suggest that the uncertainties in CC projections
should be discussed further in detail, since it can influence massively in the
process of forming climate mitigation and adaptation policies around the world.
For instance, IPCC Technical Report VI (Bates et al., 2008) discuss about the
uncertainties involved in climate projections and how it can impact the results
indicated in the report; it includes major uncertainty sources related to
hydrological cycle such as limits in climate models produced by spatial
resolution and ensemble size achieved by present computer resources (Bates et al., 2008). However, such
approach of uncertainty based on ensembles of different GCM projections is
often criticised due to lack of rigour in mathematical aspect since the
probability function is not conditioned on measured values of the variables (Rougier, 2007 in Taylor et al., 2009).
Particularly in climate projections of African region,
uncertainty in projected precipitation is a major concern because it can affect
the water resources management significantly. Not only the impact of CC on
atmospheric water-holding capacity derived from Clausius-Clapeyron relationship
and precipitation intensities, but also the socioeconomic changes in land
use/cover in the area can influence the uncertainties in future climate
projections. Furthermore, many regions in African continent lack sufficient
observational data, which is essential in climate model validation
process. Hence, there are more diverse factors, which could be considered
more extensively in uncertainty discussions of CC projections.
No comments:
Post a Comment