Refining projected multidecadal hydroclimate uncertainty in East-Central Europe using CMIP5 and single-model large ensem
- PDF / 10,158,428 Bytes
- 21 Pages / 595.276 x 790.866 pts Page_size
- 93 Downloads / 159 Views
ORIGINAL PAPER
Refining projected multidecadal hydroclimate uncertainty in East-Central Europe using CMIP5 and single-model large ensemble simulations Dániel Topál 1,2
&
István Gábor Hatvani 1
&
Zoltán Kern 1
Received: 11 February 2019 / Accepted: 20 August 2020 # The Author(s) 2020
Abstract Future hydroclimate projections of global climate models for East-Central Europe diverge to a great extent, thus, constrain adaptation strategies. To reach a more comprehensive understanding of this regional spread in model projections, we make use of the CMIP5 multi-model ensemble and six single-model initial condition large ensemble (SMILE) simulations to separate the effects of model structural differences and internal variability, respectively, on future hydroclimate projection uncertainty. To account for model uncertainty, we rank 32 CMIP5 models based on their predictive skill in reproducing multidecadal past hydroclimate variability. Specifically, we compare historical model simulations to long instrumental and reanalysis surface temperature and precipitation records. The top 3–ranked models—that best reproduce regional past multidecadal temperature and precipitation variability—show reduced spread in their projected future precipitation variability indicating less dry summer and wetter winter conditions in part at odds with previous expectations for Central Europe. Furthermore, not only does the regionally best performing CMIP5 models belong to the previously identified group of models with more realistic landatmosphere interactions, their future summer precipitation projections also emerge from the range of six SMILEs’ future simulations. This suggests an important role for land-atmosphere coupling in regulating hydroclimate uncertainty on top of internal variability in the upcoming decades. Our results help refine the relative contribution of structural differences between models in affecting future hydroclimate uncertainty in the presence of irreducible internal variability in East-Central Europe. Keywords GCM . Ranking . Model evaluation . Large ensemble . Uncertainty . Hydroclimate
1 Introduction Human-induced changes in the climate system are already detectable on daily-to-decadal timescales (Santer et al. 2018; Sippel et al. 2020). Changing weather patterns (Rosenzweig et al. 2008), expanding dryland regions
Electronic supplementary material The online version of this article (https://doi.org/10.1007/s00704-020-03361-7) contains supplementary material, which is available to authorized users. * Dániel Topál [email protected] 1
Institute for Geological and Geochemical Research, Research Centre for Astronomy and Earth Sciences, MTA Centre for Excellence, Budapest H-1112, Hungary
2
Earth Research Institute, University of California, Santa Barbara, Santa Barbara, CA, USA
(Huang et al. 2016) and the dramatic reduction of Arctic sea ice (Screen and Simmonds 2010; Dai et al. 2019), are just a few from the vast set of climatic changes attributed to rising greenhouse gas concentrations. Besides global impacts, al
Data Loading...
