ELAPSE - Enhancing local and regional climate change projections for Switzerland

• Forschungsstelle: COST
• Projektnummer: C12.0089
• Projekttitel Englisch: ELAPSE - Enhancing local and regional climate change projections for Switzerland

Schlüsselwörter (Englisch)

climate change; COST VALUE; CH2011; model output statistics (MOS); quantile mapping; regional climate model (RCM); downscaling; pattern-scaling; model uncertainty

Forschungsprogramme (Englisch)

COST-Action ES1102 - VALUE - Validating and Integrating Downscaling Methods for Climate Change Research

Kurzbeschreibung (Englisch)

An assessment of future climate change on scales relevant for impact and adaptation studies as well as decision makers requires high spatial resolution. Application-oriented climate projections - such as the recently published Swiss Climate Change Scenarios (CH2011) - thus require the downscaling of global climate model (GCM) information to a much finer scale, using a higher-resolution regional climate model (RCM) and/or statistical methodologies. Several international research efforts, among them the recently launched COST Action VALUE, are currently underway to address the associated difficulties and limitations. One limitation is the way model and emission uncertainty is represented in regional probabilistic projections. Another limitation stems from the difficulty to appropriately account for changes in temporal variability and spatial covariance that are often crucial for the representation of extreme events. It is the aim of ELAPSE to elaborate on these two constraints by further developing downscaling and post-processing techniques, and thereby to complement the existing climate scenario information from CH2011. First, ELAPSE will investigate the potential to estimate the regional response based on GCM data only. If successful, this would allow expanding on the sample size and strengthening the associated uncertainty assessment by inclusion of many new GCMs that lack an explicit RCM downscaling step. Second, ELAPSE will implement a quantile mapping methodology to produce bias-corrected transient scenarios that account for variability changes. This would in particular allow strengthening the assessment of extremes. The new scenarios will be disseminated through a common platform.

Weitere Hinweise und Angaben (Englisch)

Full name of research-institution/enterprise: ETH-Zürich - Institute for Atmospheric and Climate Science Center for Climate Systems Modeling (C2SM) - Director

Partner und Internationale Organisationen (Englisch)

AT; BG; HR; CZ; DK; FI; FR; DE; EL; HU; IE; IT; MT; NL; NO; PL; RO; RS; ES; SE; CH; UK

Abstract (Englisch)

The primary aim of the ELAPSE project was to enhance the quality and the applicability of climate change projections for Switzerland by further developing and complementing the downscaling methodologies applied within the frame of the CH2011 climate scenarios. First, relationships between global and regional climate model (GCM and RCM) projections for the Alpine region were investigated within a probabilistic framework (ELAPSEA). If robust, such relations could potentially be used to generate regional scenarios based on large GCM ensembles that have not been explicitly downscaled by RCMs. The comparison of GCM-derived and RCMderived climate change signals provides furthermore a better assessment of uncertainty propagation along the modeling chain. Second, the applicability and robustness of distribution-based bias-correction methods over Switzerland were assessed (ELAPSE B). If applied in a downscaling context, these methods can be used to produce transient climate scenarios at the site-scale based on coarse resolution and potentially biased RCM output. They thereby complement existing delta-change-based local scenarios such as those provided by the CH2011 initiative. The results obtained in ELAPSE A indicate some modification of GCM-simulated large-scale climate change signals by RCMs over the Alpine region, especially for the summer season when RCMs tend to dampen the GCM-derived drying signal and consequently yield a less pronounced warming. In other seasons, the joint multimodel projections based on RCM simulations and based on the simulations from their driving GCMs are close to each other over the analyzed regions in Switzerland. This in turn points to an important role of the specific GCM and its associated circulation and circulation changes in affecting the calculation of regional climate change signals. Indeed, the impact of differently selecting GCMs from a large ensemble, such as CMIP5, can largely modify the mean change signal and the corresponding uncertainty estimate over the Alpine region. It is hence important to compare dynamically downscaled multi-model projections with just a few GCM drivers such as those produced in ENSEMBLES or EURO-CORDEX against the larger set of available GCM projections. Such a comparison can subsequently be used as prior information in a Bayesian framework to better specify model uncertainty. Concerning the performance assessment of bias-correction methods carried out in ELAPSE B, 21 implementations of the quantile-mapping (QM) method plus a simple correction of mean biases were investigated in both a historical cross validation and a future pseudo-reality framework. The results strongly suggest the applicability of QM for the generation of local climate scenarios at the site-scale and for the region of Switzerland. QM is superior to a simple correction of mean biases, and non-parametric QM variants mostly outperform a parametric implementation. Furthermore, a comparison of QM-derived scenarios to the previous delta-change-based products (that were used within CH2011) yields a considerable added value of QM especially when considering changes in temporal variability, multi-day indices and extremes. Overall, the original objectives of ELAPSE could be achieved and the results obtained (1) complement the existing CH2011 scenario products and (2) provide important input for the setup of the next generation of Swiss Climate Change Scenarios currently in preparation and to be released in 2018. On an international level, ELAPSE results will also contribute to the inter-comparison, consolidation and final assessment of climate downscaling methods carried out within the framing COST-Action VALUE.

Datenbankreferenzen (Englisch)

Swiss Database: COST-DB of the State Secretariat for Education and Research Hallwylstrasse 4 CH-3003 Berne, Switzerland Tel. [phone] Swiss Project-Number: C12.0089