Workshop on physical modeling supporting a “storyline

Workshop on physical modeling supporting a “storyline approach”

Date: 24 - 26 April 2019Place: Oslo Science Park in Oslo, Norway. Toppsenter, Meeting room Faros

Organizers: Jana Sillmann, Ted Shepherd, Bart van den Hurk, Wilco Hazeleger, Olivia Martius, Jakob Zscheischler

Local Host: CICERO Center for International Climate Research

Phot

o: Is

elin

Røn

ning

sbak

k

The workshop is funded by the COST Action (CA17109) DAMOCLES “Understanding and modeling compound climate and weather events” and by the Norwegian Research Council as part of the TWEX project “Translating Weather Extremes into the Future” (grant # 255037). The workshop is further supported by the WCRP Grand Challenge on Weather and Climate Extremes as contribution to the implementation of the WCRP Strategic Plan 2019-2028, in support of innovation in the generation of decision-relevant information and knowledge about the evolving Earth system.

Scope

In this workshop, we will review the current approaches to generate event-based storylines using physi-cal modelling, and explore how they can be most effectively integrated into the wider climate information landscape in a way that helps climate communication and policies. The workshop is divided into thematic sessions, consisting of a series of short talks followed by breakout sessions, in which we will discuss a number of questions and formulate recommendations for future application of event-based storylines. The range of topics covered includes:

• What are the challenges and advantages of this storyline approach (basically picking a particular event) compared to the full probabilistic approach?

• How can storylines or narratives of specific events be combined with probabilistic approaches?

• What kind of models (GCM, RCM, convection-permitting, impact models) are required to support the storyline approach, and are useful to include in a model chain approach?

• What should be considered when designing the model experiments for localized events (e.g., resolu-tion, ensemble size and design, choice of scenario/forcing/initial conditions)?

• What are useful approaches to combine physical with statistical modeling and the knowledge from other domains (e.g., social sciences, impact modeling)?

Practical information

The workshop will be held at Oslo Science Park, which is easily accessed by tram from the city centre. Please use westbound lines 4 or 5.

Page 1

List of participants

Name Affiliation Country BOG_a BOG_b BOG_c

Adam Sobel Columbia University USA xAglaé Jézéquel LMD-IPSL France xAlex Hall UCLA USA xAlexandre M. Ramos Instituto Dom Luiz, University of Lisbon Portugal xAndreas Paul Zischg University of Bern Switzerland xBart Van den Hurk KNMI / Deltares Netherlands

Ben Clarke Environmental Change Institute, University of Oxford UK xChristopher White University of Strathclyde UK xDetlef van Vuuren PBL Netherlands xDouglas Maraun University of Graz Austria xElisabeth Lloyd Indiana University USA xEmanuele Bevacqua University of Reading UK xFiachra O’Loughlin University College Dublin Ireland xFrancesco Ragone ENS-Lyon France xGeert Lenderink KNMI Netherlands xGordon Woo RMS UK xJakob Zscheischler University of Bern Switzerland

Jana Sillmann CICERO Norway

Julia Slingo Cabot Institute, University of Bristol UK xKatharine Reich UCLA Center for Climate Science USA xKarin van der Wiel KNMI Netherlands xKatja Frieler Potsdam Institute for Climate Impact Research Germany xLiese Coulter B.C. Ministry of Agriculture Canada xMarisol Osman Centro de Investigaciones del Mar y la Atmósfera Argentina xMelissa Bukovsky National Center for Atmospheric Research USA xOlivia Martius University of Bern Switzerland

Pascal Yiou Laboratoire des Sciences du Climat et de l’Environment

France x

Pedro Soares Instituto Dom Luiz, Universidade de Lisboa Portugal xRita M Cardoso Instituto Dom Luiz, Universidade de Lisboa Portugal xSarah D’haen Climate Analytics Germany xTed Shepherd University of Reading UK

Trine Jahr Hegdahl Norwegian Water Resources and Energy Directorate (NVE)

Norway x

Wilco Hazeleger Netherlands eScience Center Netherlands

William Ball IAC/ETH Zurich & PMOD/WRC Switzerland xPage 2

Agenda

Wednesday April 24th

12:00 - 13:30 Registration and lunch. Lunch is included

13:30 - 14:00 Jana Sillmann and Jakob Zscheischler Welcome and introduction

Session 1: Event-based storyline approaches Conveners: Olivia Martius and Jana Sillmann 14:00 - 14:15 Julia Slingo A flood risk story for the UK

14:15 - 14:30 Douglas Maraun Storylines of a future precipitation extreme event - what is the event, what are the storylines, and are they sensible?

14:30 - 14:45 Andreas Paul Zischg Identifying precipitation events leading to worst case flood damages in mountainous river basins

14:45 - 15:00 Geert Lenderink Cases of extreme precipitation in pseudo-global warming experiments: what can we learn from it?

15:00 - 15:30 Discussion

15:30 - 16:00 Coffee break

16:00 - 17:30 Breakout groups (BOGs) Q1: Reasons for working with event-based storylines? Q2: Is a categorization or typology useful and can we work with that? Q3: What are the limitations of starting from a physical modeling /hazard perspective?

19:00 - Dinner at Ekeberg Restaurant (Map and description on how to get there at page 13)

Page 3

Thursday April 25th

09:00 - 09:30 BOG Outcomes

Session 2: Generic/sector-based approachesConveners: Bart van den Hurk and Jakob Zscheischler

09:30 - 09:45 Alex Hall The Great Flood of 1861-1862 and Its Implications for the Vulnerability of California’s Water Infrastructure to Climate Change

09:45 - 10:00 Gordon Woo Storylines for insurance insight into compound climate events

10:00 - 10:15 Trine Jahr Hegdahl Event based flood evaluation for western Norway – under present and future climate



Page 4

11:00 - 12:30 BOGs Q1: What are good examples of where storylines have been useful in decision making contexts? Q2: Which criteria do storylines need to fulfill to “work”, e.g. ability to include local information and reach out to other aspects than the hazard element? Q3: How can storylines be linked to some sense of probability (conditional probability) by e.g. dependence on external drivers?

12:30 - 13:30 Lunch


Session 3: Storylines as evidenceConveners: Ted Shepherd and Olivia Martius

14:00 - 14:15 Liese Coulter Linking lived experience and future climate knowledge

14:15 - 14:30 Elisabeth Lloyd The Storyline Account goes to Court in Juliana v US

14:30 - 14:45 Adam Sobel Cyclone landfall in Mumbai: historical analogs, urban myths, and real risks

14:45 - 15:00 Pascal Yiou A data-based approach to simulate realistic ensembles of the most extreme climate events



15:45 - 17:15 BOGs Q1: Do we need to associate storylines with probabilities? Is plausibility enough? Q2: How can probabilities be added to storylines to define their relevance for decision making? Q3: Will people trust storylines? If yes, when and why? Q4: What counts as evidence?

17:15 - 18:00 Early Career Presentations

William Ball Continued decreases in the ozone layer: unfortunate timing of events or a serious threat?

Marisol Osman Activity of the Southern Annular Mode during the Strong El Niño 2015-2016: less unusual than it used to be?

Karin van der Wiel Extreme impacts from moderate meteorology

Francesco Ragone Studying extreme climatic events with rare event algorithms applied to numerical climate models

Aglaé Jézéquel The storyline approach for extreme event attribution

Friday April 26th


Session 4: Wider perspectives on event-based storylinesConvenors: Wilco Hazeleger and Ted Shepherd

09:30 - 09:45 Melissa Bukovsky A Storyline for Precipitation in the U.S. Southern Great Plains Supported by Dynamic and Thermodynamic Changes Projected by Climate Model Intercomparison Ensembles

09:45 - 10:00 Detlef van Vuuren Storylines and integrated assessment modeling

10:00 - 10:15 Katja Frieler Extreme impact events across sectors – New raw material for storyline approaches



11:00 - 12:30 BOGs Q1: How can (physical climate) storylines be connected to the wider perspective of scenario storylines and narratives? Q2: Can storylines be useful within scientific assessments, e.g. IPCC? Can they serve as an effective bridge between WGI science and the science of WG’s II and III? Q3: Can storylines provide a means of combining information from multiple, conflicting sources (Bruce Hewitson’s “distillation” challenge)? Q4: How to make use of existing large data pools (ISIMIP, CORDEX, CMIP) and non-climate data sources for storyline development?

12:30 - 13:30 Lunch


14:00 - 14:30 Discussion & Wrap up

Page 5

Full titles and abstracts

Wednesday April 24th

Session 1: Event-based storyline approaches

14:00 - 14:15 Julia Slingo A flood risk story for the UK

Abstract In December 2015, the UK suffered some of the most serious flooding in living memory associated with unprecedented rainfall amounts. In the aftermath, the UK government asked the Met Office how bad rainfall and hence flooding could be in the future so that they could be confident that flood risk policies were adequate. This talk will tell the story of how we attempted to answer this challenging question and what gaps in our capabilities, both scientific and technical, were subsequently identified.

14:15 - 14:30 Douglas Maraun Storylines of a future precipitation extreme event - what is the event, what are the storylines, and are they sensible? Abstract In June 2009 a cut-off low over the Adriatic brought large amounts of moisture into Austria, causing a 4-day episode of repeated thunderstorms which in turn resulted in 3000 landslides in southeastern Styria. The EASICLIM project set out to construct past and future storylines of the observed event, for conditional event attribution, and for developing a narrative of future changes in landslide risk. But de fining both the event and the storylines is not straightforward. Does the event encompass the whole synoptic situation including the remote moisture sources, or just the local precipitation event? The series of storms themselves, or also the preceding wet spring responsible for the wet soil? And how are the storylines defined? Which variables are kept fixed, which are adjusted? And how are the changes defined? Based on long-term climatologies, or on composites of similar extreme events? We present different choices, consequences of these choices, including the relevance of these choices for the generalisation and relevance of the storylines.

14:30 - 14:45 Andreas Zischg Identifying precipitation events leading to worst case flood damages in moun tainous river basins Abstract Modelling flood impacts and risks at river basin scale is challenging. Especially in mountainous watersheds, extreme precipitation can be distributed spatially and temporally with remarkable variability. Depending on the topography of the river basin and the topological characteristics of the river network, some rainfall patterns can lead to the synchronization of the flood peaks between tributaries and the main river. Variability in the rainfall pattern has been demonstrated as one of the main uncertainties in determining flood impacts in mountainous river basins. This has implications for identifying the precipitation event that leads to the flood event with the highest flood losses. The knowledge of such an event is a prerequisite for stress test analyses and insurance portfolio analyses. In this talk, I will show two different approaches for identifying rainfall patterns leading to extreme flood damages. The first approach is a model experiment with an ensemble of probable maximum precipitation scenarios created by Monte Carlo simulations. For each rainfall pattern, we computed the flood losses with a model chain, consisting of a hydrological model, a routing model, a flood inundation model, and a flood impact model at street level. The second approach bases on extracting extreme precipitation events from a long-term global climate model simulation and dynamically downscaling these events with a regional climate model. The resulting rainfall events are fed into the same model chain. Both approaches allow identifying extreme flood events and losses. In addition to the flood hazard component, the other components of flood risk, exposure and vulnerability, contribute remarkably to the overall variability. This leads to the conclusion that the estimation of the probable maximum expectable flood losses in a river basin should not be based exclusively on one probable maximum precipitation scenario. Consequently, the basin-specific sensitivities to different precipitation patterns and the spatial organization of the settlements within the river basin need to be considered in the analyses. Future research might focus on the implementation of an inverse modelling approach, which considers the vulnerabilities of the different floodplains.

Page 6

Page 7

14:45 - 15:00 Geert Lenderink Cases of extreme precipitation in pseudo-global warming experiments: what can we learn from it? Abstract It is commonly accepted that precipitation extremes will increase due to increases in humidity of the air in a warming climate. To first order the latter follows from the Clausius-Clapeyron relation, predicting increases of approximately 7 % per degree warming. Based on this simple physical principle, experiments have been designed whereby cases of extreme precipitation have been re-simulated for perturbed warmer climates (using a pseudo-global warming approach). Likewise, cooling experiments have been performed in order to answer question concerning attribution. But, how representative are the results of these cases? and how should the cases be chosen? Here, we illustrate that a selection bias could affect the results, and this may lead to underestimation of the response to warming, but reversely also to too strong attribution statements.

Thursday April 25th

Session 2: Generic/sector-based approaches

09:30 - 09:45 Alex Hall The Great Flood of 1861-1862 and Its Implications for the Vulnerability of California’s Water Infrastructure to Climate Change Abstract The largest flood in California’s recorded history occurred from December 1861 to January 1862. The flood was the result of several atmospheric rivers hitting the region in succession, and likely produced the equivalent of a few meters of precipitation over a period of 43 days. The combined effects of rain and subsequent snowmelt flooded nearly all agricultural valleys and inundated many cities, including the new State capital of Sacramento. If such an event were to occur now, it would easily overwhelm the State’s more recently constructed and highly engineered water resource infrastructure, producing a similar scale of flooding as the historical event. It could also cut off the water supply for most of the State’s urban areas, including the megalopolises of Los Angeles and San Francisco. The paleoclimate record indicates that such events occur roughly every 200 years. Meanwhile, future climate simulations predict an increase in the frequency of such events, to the point where one occurrence is a near certainty before the year 2100. Future warming would cause the precipitation to fall almost exclusively as rain rather than snow, greatly exacerbating the flooding. Here we explore how the Great Flood of 1861-1862 could be used as the basis of a storyline to support climate change adaptation planning.

09:45 - 10:00 Gordon Woo Storylines for insurance insight into compound climate events Abstract Climate change is a major concern of insurers of weather risks. Insurance risk management of weather risks in a changing climate environment requires insight into compound climate events that might result in extreme economic loss. Risk insight is enhanced by the recognition that even if such events may have not already happened, they might have happened before. A counterfactual risk analysis approach is well suited to addressing compound risks, and can be embedded within a storyline framework. Examples are given to illustrate the value of counterfactual risk analysis in developing storylines for climate change insurance risk management.

Page 8

10:00 - 10:15 Trine Hegdahl Event based flood evaluation for western Norway – under present and future climate Abstract In a majority of future climate studies several projections of global and regional climate models makes up an ensemble of simulations used to obtain a range of probabilities for e.g. future floods. In this study, we provide complementary event-based information, and compare flood impact caused by high intensity precipitation events in a warmer future climate to present climate. From EC-Earth simulations, representing 30 years of both present and future climate, four high intensity precipitation events over western Norway were chosen. A modelling chain similar to the operational flood forecasting chain enabled us to assess the flood impact caused by each of the chosen events. First, the EC-Earth simulations were downscaled with the regional weather prediction model AROME-MetCoOp, allowing for a better spatial representation of high intensity precipitation. Thereafter, catchments average values of temperature and precipitation were used to run the hydrological flood-forecasting model. Flood levels for 37 catchment in western Norway were evaluated. The results showed that in the future climate compared to the present climate, more catchments were affected by floods during the events, and in addition, the floods in the future climate were larger.

Session 3: Storylines as evidence

14:00 - 14:15 Liese Coulter Linking lived experience and future climate knowledge

Abstract The quality of stationarity in the climate system allowed human civilization to develop within the highly variable, but bounded limits of a slowly changing envelope of extremes. However, the climate system has started to surpass the edge of those stationary boundaries and will reach new environmental limits which are a struggle to incorporate into personal and social narratives of the future. Researchers and decision-makers themselves frame climate narratives within their personal, cultural and employment communities, even while a non-stationary climate system presents previously unlikely stories, populated with black swans, extreme events and the need for heroic efforts. Quantitative information is essential to calculate what climate trajectories are most likely however, narrative communication is crucial to understand the implications of future impacts, if they happen. In contrast to traditional risk management that calculates a probability of occurrence, storylines can account for unprecedented events by assuming a 100% probability and imagining the event as real. Even so, where expected climate impacts are locally unprecedented, lived experience does not provide supportive links to imagine future climate. In the human brain, the foundation of future thinking is memory; anchoring images, metaphors and analogues that can be informed by climate information to represent the imagined future. When the future does not offer significant mirrors of the past, it requires substantial mental and creative effort to reflect new knowledge in research design as well as in adaptation decision-making. The challenges inherent in picturing an unprecedented future affects imagining problems, devising research methods and contextualising findings. Linking probabilistic and imaginative communication can support the application of future climate knowledge by drawing on both projected changes and lived experience.

14:15 - 14:30 Elisabeth Lloyd The Storyline Account goes to Court in Juliana v US

Abstract I examine the logic of the research questions facing the Plaintiffs and court in the Juliana v US federal court case involving climate change and the Federal government’s contributions to warming. In examining the Federal expert testimony of John Weyant, who criticizes Keven Trenberth’s storyline testimony, I discover some interesting features of Weyant’s logic and argument. Briefly, Trenberth argues on behalf of the Plaintiffs that they suffered and will suffer harms using storyline accounts of extreme events. Weyant mischaracterizes the storyline methods, and applies faulty standards of evidence in order to conclude that the storyline approach fails to meet the necessary standards. By claiming that highly detailed causal accounts, “with sufficient certainty”, are required, Weyant moves the goalposts outside the field of play, thus inappropriately demanding finely-tuned causal accounts of degrees that are only necessary if the plaintiffs are demanding compensation, which they are not.

Page 9

14:30 - 14:45 Adam Sobel Cyclone landfall in Mumbai: historical analogs, urban myths, and real risks Abstract Could Mumbai be struck by a major tropical cyclone, and if so what would the impacts be? I will tell the story of some recent inquiry into this question, and the surprisingly rich and complex set of storylines, in multiple senses, that have come out of it. A few colleagues and I were led to ask the question by the writer and anthropologist Amitav Ghosh, who himself was provoked to think about it by the impacts of Hurricane Sandy on New York City, and the several historical and geographical commonalities between the two post-colonial coastal megacities. We first wondered if it had happened before, and quickly found accounts, in many online and published sources, of a terribly destructive (death toll ~100,000) cyclone making landfall in Mumbai (then Bombay) in 1882. Investigation into primary historical sources, however, revealed this story to be false, an “urban myth”; we still do not know where it came from. This doesn’t mean it couldn’t happen, however, and we went on to conduct a small research project, using a newly developed hybrid statistical-dynamical tropical cyclone risk model, to determine the probability that such an event actually could occur. Our calculations show that a major cyclone landfall in Mumbai is certainly possible: the return period (depending on the intensity threshold chosen) is perhaps a few hundred years, comparable to that od Sandy in NYC, based on the historical climate. Global warming is probably increasing the risk, though at an uncertain rate. While a severe cyclone may still seem a small and remote risk to many Mumbai residents and government officials, the city has recently experienced major, very damaging floods – most severely in 2005, but also in 2017 – due to monsoon rains not associated with cyclones. These events have clearly demonstrated the city’s vulnerability, and sparked new planning, though not yet to the point of seriously considering a major cyclone landfall; in fact, new development continues in the most exposed and vulnerable coastal zones. I will try to bring together these different threads to find lessons for storyline approaches more generally.

14:45 - 15:00 Pascal Yiou A data-based approach to simulate realistic ensembles of the most extreme climate events Abstract Simulating ensembles of extreme events is a necessary task to evaluate their probability distribution and analyse their meteorological properties. Algorithms of importance sampling have provided a way to simulate trajectories of dynamical systems (like climate models) that yield extreme behavior, like heat waves. We present an adaptation of an importance sampling algorithm to provide a data-based algorithm that simulates the most extreme events in a realistic way. This presentation outlines the methodology on European heatwaves and illustrates the spatial and temporal properties of simulations.

17:15-18:00 Early Career Presentations

William Ball Continued decreases in the ozone layer: unfortunate timing of events or a serious threat? Abstract Recent observational evidence suggests that the ozone layer continues to display highly probable decreases above highly populated regions, in contradiction to model projections. This may be due to multiple modes of natural variability conspiring to produce a spurious signal. However, this has not been conclusively demonstrated, so that a genuine long-term decrease is a plausible inference from the data. The latter possibility is more serious for society than the former, but overconfidence in models may lead to a delay in dealing with a potential threat. I will make the case that, in light of the risk and uncertainty of the driving cause, the evidence should lead us to assume a serious problem in need of investment to understand and resolve, rather than the assumption that we should continue to wait for additional data to be accumulated.

Page 10

Marisol Osman Activity of the Southern Annular Mode during the Strong El Niño 2015-2016: less unusual than it used to be? Abstract In this work we analyzed the strong positive phase of the Southern Annular mode (SAM) observed during the El Niño event of 2015-2016 (EN15-16) and impact on climate anomalies in the western portion of the Southern Hemisphere. Previous studies documented that El Niño (EN) events are in general associated with negative phases of the SAM. However, despite EN15-16 was one of the three strongest events ever recorded, it was associated with SAM positive phases of extreme intensity. Furthermore, while the negative linear relationship between ENSO and SAM during the most recent period (1986-2014) was significant and associated with a narrow uncertainty band, the combined condition of both climate patterns in the EN15-16 event was an outlier. In addition, while SAM anomalies during 2015 austral spring (September-October-November) can be attributed to the extreme negative ozone anomalies in the Southern Hemisphere (SH), the reasons for the strong positive anomalies in austral summer (December-January-February) are still unclear. The EN15-16 influence on the austral summer circulation anomalies at the extratropical and polar regions of the SH was considerably altered by the strong SAM positive phase, which was evident not only at the troposphere but also at the stratosphere. The analysis of the geopotential heights anomalies at 200hPa reveals that in the EN15-16 event, large negative anomalies were well discernible over the Antarctica, that are typical of SAM positive phases. In addition, negative centers located over central and eastern portions of the south Pacific middle latitudes and the positive center extended to the west of the Antarctic Peninsula are shifted equatorward in the EN15-16 than in the other EN composites. Such circulation changes resulted in unusual regional impacts. The unusual behaviour of the SAM during the last EN and the projected changes of SAM during the 21st period, as a consequence of GHG emissions and Ozone recovery, open up multiple questions around the future of ENSO-SAM relationship and its implication in the climate anomalies observed in the SH.

Karin van der Wiel Extreme impacts from moderate meteorology Abstract The investigation of risk due to weather and climate events is an example of policy relevant science. Risk is the result of complex interactions between the physcial environment and societal factors. The societal impact of two similar meteorological events at different times or different locations may therefore vary widely. Despite the complex relation between meteorological conditions and impacts, most meteorological research is focused on the occurrence or severity of extreme meteorological events. Here I will argue that an approach of ensemble climate-impact modelling is required to adequately investigate the relationship between meteorology and extreme impact events. By means of a simple case study I will demonstrate that extreme weather conditions do not always lead to extreme impacts; in contrast, extreme impacts may result from (coinciding) moderate weather conditions. Explicit modelling of climate impacts, using the complete distribution of weather realisations, is thus necessary to ensure that the most extreme impact events are identified. The approach allows for the investigation of high-impact meteorological conditions and provides higher accuracy for consequent estimates of risk.

Page 11

Friday April 26th

Francesco Ragone Studying extreme climatic events with rare event algorithms applied to numerical climate models Abstract A reliable quantification of the risk associated with extreme climatic events is crucial for policymakers, civil protection agencies and insurance companies. Studying extremes on a robust statistical basis with complex numerical climate models is however computationally challenging, since extreme events are rare, and thus very long simulations are needed to sample a significant number of them. I will discuss how the problem of sampling extremes in climate models can be tackled using rare event algorithms. Rare event algorithms are numerical tools developed in the past decades in mathematics and statistical physics, dedicated to the reduction of the computational effort required to sample rare events in dynamical systems. Typically they are designed as genetic algorithms, in which a set of cloning rules are applied to an ensemble simulation in order to focus the computational effort on the trajectories leading to the events of interest. I will present a rare event algorithm developed in the context of large deviation theory, and I will show how it can be used to sample very efficiently extreme European heat waves in simulations with the climate model Plasim. This allows to characterise the statistics of heat waves with return times up to millions of years, with computational costs three orders of magnitude smaller than with direct sampling. The algorithm samples a large number of trajectories leading to very rare events, which can be used to study their characteristic dynamics, and also to observe ultra rare events that would have never been observed in a normal simulation. I will then discuss how these techniques can be applied to study a wide range of different processes with complex climate models.

Aglaé Jézéquel The storyline approach for extreme event attribution

Abstract The “storyline approach” for extreme event attribution has recently been developed, in contrast to more traditional “risk-based approach”. Rather than asking how climate change the probability of occurrence of an event (risk-based approach), the idea is to better understand how climate change affects the physical processes leading to the event. I will present both approaches, their main differences and how they feed a debate within the extreme event attribution community regarding the social usefulness of extreme event attribution results.

Session 4: Wider perspectives on event-based storylines

09:30 - 09:45 Melissa Bukovsky A Storyline for Precipitation in the U.S. Southern Great Plains Supported by Dynamic and Thermodynamic Changes Projected by Climate Model Intercomparison Ensembles Abstract In this presentation, I will give an overview of the projections for the driving factors of precipitation in the southern Plains, their physical and/or mechanistic plausibility, and explain how they lead to a potentially credible storyline for precipitation change. A strong consensus for an amplification of the warm-season cycle of precipitation is present in regional climate model (RCM) and global climate model (GCM) ensemble simulations of precipitation over the southern U.S. Plains. Consensus, however, does not guarantee credibility or usefulness. Thus, to assess the credibility of these projections, the consistent parts of the ensembles’ responses were assessed and the physical arguments for their plausibility in a warming climate were established. In the end, the overall storyline for the precipitation changes from most of the RCMs, in particular, is physically plausible and consistent, lending credibility, and therefore may be used with more confidence relative to the details of the changes in precipitation (e.g. exact magnitude and timing of change). This work currently l everages simulations completed for the North American Regional Climate Change Assessment Program (NARCCAP, a precursor to North American CORDEX), CMIP3, and CMIP5. The RCMs project larger precipitation changes than their GCM counterparts and the CMIP3 and CMIP5

Page 12

ensembles, and simulate historical precipitation with more fidelity. The ensembles suggest that warm-season precipitation will increase during the early spring wet season, but shift north earlier in the summer, intensifying late summer drying. Both larger-scale dynamical processes and local land-atmosphere feedbacks drive these precipitation changes. In the simulations that credibly simulate the historic climate (and most do), the mechanisms for change are plausible and consistent. This leads to greater certainty in the projections, and a clear storyline for potential future changes.

09:45 - 10:00 Detlef van Vuuren Storylines and integrated assessment modeling

Abstract Storylines form a critical part of scenarios developed using integrated assessment models. This method has become popular since the publication of the IPCC SRES report in 2000. In context, storylines describe possible future evolutions of the world (or region, country). Storylines are mostly used to capture elements that are difficult to ‘model’ such as future developments with respect to governance, international relationships and policy. Storylines also help to harmonize assumptions for different model components (e.g. technology in energy and land-use). In the presentation, we will discuss the IAM storyline approach and look into the question how IAMs can contribute to storyline approaches in climate modelling.

10:00 - 10:15 Katja Frieler Extreme impact events across sectors – New raw material for storyline approaches

Abstract Global warming has already reached 1°C above pre-industrial levels. However, the public is still mostly discussing climate change and its impacts as a problem not of the current but of future generations. That is partly due to the fact that the extent and impact of crop failures, river floods, tropical cyclones, heatwaves, wildfires, and droughts depend on the underlying weather events as well as on human factors such as land use, water and agricultural management, and population density and historical variations in these human factors impede the detection of climate change signals in observational records of hazards and exposure. Here we show how harmonized climate impact simulations from the Inter-Sectoral Impact Model Intercomparison Project (www.isimip.org) can be used to quantify these signals for historical and future time periods. According to the simulations climate change has a significant contribution to historical changes in the areas affected by or the number of people exposed to the different types of extreme events going beyond the considered socio-economic forcing. We also provide cross-sectoral consistent projections of extreme events for RCP2.6 and RCP6.0 and would like to use the opportunity to discuss different options to translate them into storylines.

Page 13

Dinner 24th April - Ekeberg Restaurant

The main dinner will be held at Ekeberg Restaurant, located in the hills just above Oslo, with a stunning view of the inner Oslo fjord, the city and the Operahouse.

The restaurant is easily accessed by the blue tram. 18 or 19 to Ekebergrestauranten, which goes every ten minutes from Oslo Central Station. The stop is called Ekebergparken (Former Sjømannsskolen).

See map below.

Documents

Workshop on physical modeling supporting a “storyline