Assessment of the sources of uncertainty in projected changes in precipitation

What is it about?

The projected changes in precipitation differ considerably when different climate model, socio-economic scenario and different climate model runs are evaluated. These differences originate from three sources: natural variability of the climate system, incomplete knowledge of the physical processes governing the climate system or the technical limitations (model uncertainty), and the unpredictability of the socioeconomic factors determining the future emissions of greenhouse gases (scenario uncertainty). The contribution of these sources is in general different for different variables and different characteristics (e.g. it is different for mean and annual maxima). We present here a simple and flexible framework for assessing the contribution of each source of uncertainty applicable for assessment of climate model ensembles.

Featured Image

Why is it important?

- it is possible to easily apply the method to unbalanced data which is typical situation for climate model ensembles - standard ANOVA framework is difficult to apply in this situation