SMG logo


29-31 October 2014, Santander SPAIN


The Objective 

With the term "Distillation" we indicate that area of research on how to address, integrate and analyze the multi-model multi-method multi-scale (M5S) climate data from the use-perspective of the sector needs of the Vulnerability Adaptation and Impact (VIA) communities.  This issue falls at the core of the WCRP Working Group on Regional Climate (WGRC) terms of reference, and has the potential to substantially advance the responsible adoption of climate information for adaptation and policy.

The workshop seeks to develop a research agenda on innovation in the analysis and interpretation of M5S climate data, in the context of an uncertain observational history (and un uncertain range of future human climatic influences). This is distinct from the foundational scientific questions of model performance, or of characterizing uncertainty, and instead places the focus on addressing the contradictions, differences, and data interpretation as a function of the user’s application context. An in turn the outcomes of this exercise will help shaping the scientific questions.

The workshop will be the forum for the experts from the relevant communities to help identify strategic research priorities and formulate associated scientific questions in order to bring about a step advance in the relevance of the scientific output for uptake by broader community interests.

Framing the Approach

Users face three challenges in the adoption of climate data;

a) Access to very large and evolving data sets; this is an infrastructure issue.

b) Inadequate tailoring and communication of data for purposes of specific applications; this is largely the responsibility of the emerging climate services communities.

c) The confusion and contradictions inherent across the suite of M5S data products; this is a research challenge for the scientific community. 

This workshop addresses the last aspect (c), and while rooted in the user’s interests, it is not focused on the specifics of meeting the user’s tailored needs.  Rather, this challenge is for scientists look across the diversity of M5S data to address the sources of downstream confusion that the user is faced with.  By this we mean that the user community is frustrated in their decisions to respond to climate stresses by the uncertainty and contradictions evident within and between climate data sets emerging from the collective of GCM, RCM, ESD, and observation communities’ activities.   While it is clear that some of the uncertainty and contradiction is irreducible (in that there is inherent system stochastic behavior as a function of scale in time and space), there are also additional sources of structural and scientific uncertainty that become conflated with the true system uncertainty in the communication to users.  Likewise, the limits to skill as a function of method, spatial scale, time scale of prediction, and geographic location all contribute to a difficulty in understanding the mixed messages from the plethora of products.  This further undermines confidence in the added value of the scientific outputs. 

Thus we frame the workshop objectives by this use-case question: “How do I know what credence your data product has for my decision context?”  The scientist’s answer is likely “I don’t know”, and the resulting conflict of the multiplicity of messages either leaves the user inadequately prepared to incorporate the information into their decision making, or else opens the door to maladaptation if a message from only one subset of the data is adopted

Workshop themes

The workshop themes will draw on the input from key representatives of different communities, backed by substantial allocation of discussion time in working groups and in plenary to address the following:

a) Review of the current activities by the different relevant research communities, and assess barriers and limitations that constrain the robustness of regional information, and particularly which inhibit the activities of the VIA community

  • observational data
  • Downscaling (RCMs / ESD/time-slice GCMs?/stretched-grid GCMs?)
  • Climate services

b) Explore our understanding of the contradictions between models, observations, GCMs, RCMs, and ESD methods; the sources and reasons.

c) Prioritize the research frontiers which offer the greatest return for investment as measured by benefit to the VIA community.

d) Assess the emerging approaches to method-independent user-relevant metrics of skill, and identify appropriate options to advance these activities in a cohesive and systematic approach.

e) Outline possible research agenda that would explicitly address the contradictions and uncertainty inherent in the disparity of multi-model multi-method multi-scale simulation and observational data sets.

f) Identify potential approaches that would best advance the development of robust regional information on climate change and variability that are scale relevant and geographically transferrable.

Anticipated outcomes

a) Initiate the writing of a position paper based on the workshop outcomes to highlight the issues to a wider audience and so foster broader research engagement.

b) Frame a research guidance paper for further development and adoption under the WCRP WGRC research agenda and as contribution to the WCRP Grand Challenge on Regional Climate Information

c) Foster future collaboration and communication between relevant communities: CMIP, CORDEX, GFCS, GCOS, FUTURE EARTH, PROVIA etc...

d) Establish a community of practice for addressing the distillation challenge