This splinter meeting is inviting all modeling groups working with global or regional single model large ensembles. The purpose of this meeting is to establish a forum to discuss future plans for experiments, data sharing, collaboration across disciplines as well as exchange of experiences and information about existing simulations.
In the recent years, a growing number of modelling groups have created large initial condition ensembles for global climate models as well as high-resolution downscaling of such ensembles using regional climate models. These large ensembles offer new opportunities to quantify a model's true forced response and internal variability and improve our understanding of projected future changes. The ensembles also allow a reinterpretation of the observed record and help in interpreting sparse observations on the one hand, the assessment of model biases on the other hand. Furthermore, the large sample size enables the exploration of extreme events in a changing climate, the detection of compound events, as well as a robust assessment of a model’s variability.
These large ensembles offer unexploited opportunities for a wide range of disciplines, such as hydrology, biogeosciences and open new perspectives for climate dynamics. We want to introduce these new opportunities and explore how a combined analysis of the existing large ensembles can advance our knowledge in different fields.
The configuration of the existing ensembles varies between different modeling groups in their choice of ensemble size, forcing scenarios and initialisation. In this splinter meeting, we will share lessons learned from running the ensembles and coordinate future ensemble experiments to use resources efficiently and make the experiments from different models more comparable. Furthermore, we will discuss how the existing ensemble simulations can be made accessible to the community.
We also encourage anyone who is not yet involved in large-ensemble modeling but interested in the topic and this meeting to contact us.
Schedule
16:15 – 17:00 Part I: current state
17:00 – 17:45 Part II: applications and future plans
17:45 – 18:00 Concluding remarks and next steps
Objectives:
- Establish a forum for communication among modelling groups
- Create an overview of existing large ensembles (both RCM and GCM)
- Identify gaps in RCM / GCM ensemble modelling that should be addressed in the future
- Identify topics that need further discussion and decide whether a dedicated workshop should be held in autumn 2019
Agenda:
- Introduction
- Part I - current state
Which large ensembles do exist, why were they created and what have we learned?
-- Presentation of GCM large ensembles 4' (Nicola Maher, Andrea Dittus)
-- Presentation of a RCM large ensemble 3' (Raul Wood)
-- Round table introduction of all participants
-- Open discussion to share information on existing large ensembles and lessons learned from running these simulations.
- Part II - applications and future plans
What are the near-term plans (1-2 years) for large ensembles and which challenges are we facing? What are our expectations for large ensemble modelling in the wider climate science context?
-- Presentation: future plans for a GCM large ensemble 3' (Keith Rodgers)
-- Presentation: future directions for LE: towards a better quantification of changes in extremes 3' (Erich Fischer)
-- Round table introduction of future plans from different groups
-- Open discussion on topics where large ensembles could introduce a new perspective, gaps in the existing large ensembles, and the scientific and technical challenges that we will face in this context
- Conclusions
Based on the previous discussion:
-- identify potential synergy effects both within and between GCM and RCM communities
-- identify topics that need further discussion and decide if a European large ensemble workshop should be held in autumn 2019 to continue this discussion