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 the provided gridded observational datasets are representative of the true climate
Uncertainties exist in the gridded observed climate datasets in terms of how representative the observed climate record is of our true climate, including variability and extremes.
The observed climate datasets are presented as a gridded dataset, whereby point observations of climate (the nationwide network of climatological and meteorological stations) are used to create a 5 km grid of continuous information (a process called interpolation) for the period of record. This process of interpolation, although providing advantages for analysis and presentation, introduces further uncertainties. These uncertainties arise from the uneven spatial coverage (relatively few stations are located in northern and western parts of the UK) and the lack of availability of some desirable information (relatively few stations provide hourly observations).
Furthermore, differences have been found to occur between some monthly gridded values and those derived from daily gridded values within that same month at some locations for some months. These differences appear to primarily result from the fact that the monthly values result from one (monthly) interpolation, whereas the daily values are the result of up to 31 (daily) ones. Some portion of this difference also results from the introduction into the gridding procedure of instruments that are read only monthly.
A test of selected months showed that differences for most of lowland UK are random and very small. For two of the months tested, however, larger differences in precipitation totals were noted over parts of some upland areas (e.g. West Highlands, Lake District, Pennines, and northern Wales) typically of order +/– 50mm. In the case of maximum and minimum temperature, the differences were mainly very small with only a very few isolated random larger differences (typically +/– 0.5 to 1ºC).
It was concluded that given the random nature of the differences in adjacent grid squares, addressing these differences so that the sum of the daily values agreed with the monthly ones may lead to unrealistic daily patterns (especially for rainfall). As such, this difference is noted on the Met Office web site warning users that these differences may occasionally occur.