Pattern scaling using ClimGen: monthly-resolution future climate scenarios including changes in the variability of precipitation

Abstract

Development, testing and example applications of the pattern-scaling approach for generating future climate change projections are reported here, with a focus on a particular software application called BClimGen^. A number of innovations have been implemented, including using exponential and logistic functions of global-mean temperature to represent changes in local precipitation and cloud cover, and interpolation from climate model grids to a finer grid while taking into account land-sea contrasts in the climate change patterns. Of particular significance is a new approach for incorporating changes in the inter-annual variability of monthly precipitation simulated by climate models. This is achieved by diagnosing simulated changes in the shape of the gamma distribution of monthly precipitation totals, applying the pattern-scaling approach to estimate changes in the shape parameter under a future scenario, and then perturbing sequences of observed precipitation anomalies so that their distribution changes according to the projected change in the shape parameter. The approach cannot represent changes to the structure of climate timeseries (e.g. changed autocorrelation or teleconnection patterns) were they to occur, but is shown here to be more successful at representing changes in low precipitation extremes than previous pattern-scaling methods.

Recent United Kingdom and global temperature variations

“…We might expect CET to warm more than the global mean because of the land-ocean warming contrast mechanisms, but this is not apparent from current climate model simulations. Osborn et al (2016) diagnosed the warming simulated over central England per degC of global warming from a set of CMIP5 climate models and found ratios between 0.7 and 1.3, with a multi-model mean close to 1.0. With the caveat that the CMIP5 models are generally too coarse to simulate features such as land-sea breezes that would affect CET, this suggests that CET warming from external forcings should be close to the global-mean warming, plus or minus 30% (this is depicted in Figure 3 by the grey shading, which has scaled the multi-decadal global-mean warming by the range of CMIP5 CET/global ratios).…”

The global-scale impacts of climate change: the QUEST-GSI project

“…Climate and sea level rise scenarios were produced by pattern-scaling climate model output, and then rescaling the results to match specified changes in global mean temperature (Osborn et al 2015). Patterns for change in climate variables and sea level rise were constructed from 21 and 9 CMIP3 climate models (Meehl et al 2007) respectively, although not all models were used in all impact sectors.…”

“…The first paper (Osborn et al 2015) presents an overview of the pattern-scaling approach used to construct climate scenarios for the QUEST-GSI project, summarising the method and its assumptions. The following papers present impacts in individual sectors, considering in turn water resources (Gosling and Arnell this issue), river flooding (Arnell and Gosling this issue), coastal flooding (Brown et al this issue) and crop productivity (Rose et al this issue).…”

The Impact of Treatment of Organic Manures on Future Soil Carbon Sequestration Under Different Tillage Systems in Pakistan

“…Future monthly temperature and precipitation for four SRES emissions scenarios for the grid cell containing the site were extracted from the ClimGen Emissions Scenario dataset (Osborn et al, 2016). These scenarios were economic growth with globalization (A1B), economic growth with regionalization (A2), environmental protection with globalization (B1) and environmental protection with regionalization (B2).…”