Still weighting to break the model democracy

Collins, Matthew

doi:10.1002/2017gl073370

Cited by 10 publications

(9 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Much has been written about the validity of probability distributions derived from climate model ensembles in other contexts (e.g., temperature and precipitation fields), given that models are not independent or equally plausible (e.g., Tebaldi & Knutti, ). Techniques have been proposed to deal with issues of model independence and quality (e.g., Knutti et al, ; Sanderson et al, ), but it is unclear whether these techniques are applicable to GCM‐derived sea level change projections (Collins, ). Some assessments have introduced expert judgement‐based broadening of GCM‐based probability distributions in order to account for these issues; for example, AR5 interpreted CMIP5‐based central 90% ranges as “likely” (at least 66% probability).…”

Section: Projections Of Relative Sea Level Changementioning

confidence: 99%

Usable Science for Managing the Risks of Sea‐Level Rise

et al. 2019

View full text Add to dashboard Cite

Sea-level rise sits at the frontier of usable climate climate change research, because it involves natural and human systems with long lags, irreversible losses, and deep uncertainty. For example, many of the measures to adapt to sea-level rise involve infrastructure and land-use decisions, which can have multigenerational lifetimes and will further influence responses in both natural and human systems. Thus, sea-level science has increasingly grappled with the implications of (1) deep uncertainty in future climate system projections, particularly of human emissions and ice sheet dynamics; (2) the overlay of slow trends and high-frequency variability (e.g., tides and storms) that give rise to many of the most relevant impacts; (3) the effects of changing sea level on the physical exposure and vulnerability of ecological and socioeconomic systems; and (4) the challenges of engaging stakeholder communities with the scientific process in a way that genuinely increases the utility of the science for adaptation decision making. Much fundamental climate system research remains to be done, but many of the most critical issues sit at the intersection of natural sciences, social sciences, engineering, decision science, and political economy. Addressing these issues demands a better understanding of the coupled interactions of mean and extreme sea levels, coastal geomorphology, economics, and migration; decision-first approaches that identify and focus research upon those scientific uncertainties most relevant to concrete adaptation choices; and a political economy that allows usable science to become used science.Plain Language Summary The impacts of sea-level rise pose growing threats to coastal communities, economies, and ecosystems, and decisions made today-in areas like land-use policies, coastal development, and infrastructure investment-will affect exposure and vulnerability for generations to come. Thus, the usability of sea-level science is a pressing concern. Ensuring usability requires grappling with deep uncertainty in long-term sea-level projections, the relationship between long-term trends and the impacts of short-lived extreme events, and the ways in which the physical coast, as well as people and ecosystems along the coast, respond to increasingly frequent flooding. At the same time, it also requires more extensive and deliberate stakeholder engagement throughout the scientific process, as well as cognizance of the political economy of linking stakeholder-engaged science to action.

show abstract

Section: Projections Of Relative Sea Level Changementioning

confidence: 99%

Usable Science for Managing the Risks of Sea‐Level Rise

et al. 2019

View full text Add to dashboard Cite

show abstract

“…On the other hand, if all surrounding empirical evidence is met, and a particular data point or subset of data cannot be replicated by the model, this may warrant re-evaluation of the data in question (Figure 8). In an analogous manner to climate modelling (Collins et al, 2017), it remains open as to whether all models which pass a threshold acceptance barrier should be incorporated into an acceptable set of reconstructions (i.e. a model democracy; Knutti, 2010) or whether a "best-fit" model which performs best against all constraints should be identified and used for further research.…”

Section: An Approach To Measuring Model-data Fitmentioning

confidence: 99%

Recent progress on combining geomorphological and geochronological data with ice sheet modelling, demonstrated using the last British–Irish Ice Sheet

Ely

Clark

Hindmarsh

et al. 2019

J Quaternary Science

View full text Add to dashboard Cite

show abstract

“…Also, whilst the CCAM simulations brought considerable improvements to their driving GCMs in representing the spatial and temporal variations of temperature in the country (see Section 3.1), they were generally outperformed by their driving GCMs in capturing the rainfall characteristics of the country (see Section 3.2). Given that there is still no widely accepted robust formal method of weighting different simulations in an ensemble, despite current advances and increased understanding of the problem (e.g., Collins, ), the projected changes in temperature and rainfall over the country presented in this section are derived from all the dynamically downscaled model simulations listed in Table given equal weights, one vote for each of the simulations, following IPCC (). We recognize that a more complex approach to weighting the simulations could be taken, based on an assessment of model performance and/or the amount of independent information provided to the ensemble by each simulation, and some implications for the obtained projected values from the ensemble mean are discussed.…”

Section: Mid‐21st Century (2036–2065) Climate Projections For the Phimentioning

confidence: 99%

Projected changes in rainfall and temperature over the Philippines from multiple dynamical downscaling models

Villafuerte¹,

Macadam

Daron

et al. 2019

Intl Journal of Climatology

View full text Add to dashboard Cite

To help meet increasing demands for high‐resolution climate change projections in the Philippines, this study provides the results of multiple dynamically downscaled climate model simulations for projected changes in rainfall and temperature over the country by the mid‐21st century (2036–2065) relative to the baseline period (1971–2000), under the RCP8.5 scenario. The model‐simulated seasonal means of temperature, rainfall, and low‐level wind patterns were first compared with observations during the baseline period. Comparisons made between the model‐derived and APHRODITE observation‐based gridded temperature and rainfall data indicate that the dynamically downscaled simulations provide an overall improvement from their driving global climate models in capturing the spatial patterns of rainfall over the country, and the spatial and temporal characteristics of the country's mean temperature. Future climate projections show that the country's climate is expected to become warmer by the mid‐21st century, with a multi‐model ensemble mean increase of 1.2 to 1.9°C, relative to the baseline period, projected for many parts of the country and across most seasons. Slightly higher increases are projected during the country's hottest season, March–April–May. However, there are large differences in the models' projected rainfall changes by the mid‐21st century across seasons and regions. For most parts of the country, the multi‐model ensemble includes simulations that show increases and simulations that show decreases in rainfall. Nevertheless, there is a tendency of model projections towards wetter conditions over northern and central sections of the country (particularly in the December–January–February season) and drier conditions in the southern region of the country in almost all seasons. The results demonstrate the need for communities in the Philippines to adapt to a future warmer climate and prepare for a range of possible future changes in rainfall and temperature.

show abstract

Still weighting to break the model democracy

Abstract: Ensembles of climate model simulations are employed to project how climate might change in the future. How do these ensemble projections relate to what will happen to the real‐world climate?

Cited by 10 publications

References 9 publications

Usable Science for Managing the Risks of Sea‐Level Rise

Usable Science for Managing the Risks of Sea‐Level Rise

Recent progress on combining geomorphological and geochronological data with ice sheet modelling, demonstrated using the last British–Irish Ice Sheet

Projected changes in rainfall and temperature over the Philippines from multiple dynamical downscaling models

Contact Info

Product

Resources

About