Regionalizing the sea-level budget with machine learning techniques

Camargo, Carolina M.L.; Riva, Riccardo; Hermans, Tim H. J.; Schütt, Eike M.; Marcos, Marta; Hernández‐Carrasco, Ismael; Slangen, Aimée B. A.

doi:10.5194/os-19-17-2023

Cited by 11 publications

(3 citation statements)

References 100 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…where 𝛳 𝑖 is the latitude of the cell 𝑖. δ-MAPS has already been successfully applied to SST datasets in a multi-model assessment of CMIP6 models 30 and to 6000-year long transient simulations [31][32][33] . Moreover, the algorithm has a relatively fast execution time, on the order of a one or two hours with our tuning (Supplementary Figure 5).…”

Section: δ-Maps Methodsmentioning

confidence: 99%

netEC: an emergent constraint on Climate Sensitivity based on network analysis

Nenes,

Ricard,

Falasca

et al. 2023

Preprint

View full text Add to dashboard Cite

The 2015 Paris agreement was established to limit Greenhouse gas global warming below 1.5°C above preindustrial era values. Knowledge of climate sensitivity to greenhouse gas levels is central for formulating effective climate policies, yet its exact value is shroud in uncertainty. Climate sensitivity is quantitatively expressed in terms of Equilibrium Climate Sensitivity (ECS) and Transient Climate Response (TCR), estimating global temperature responses after an abrupt or transient doubling of CO2. Here, we use a holistic approach to summarize the behavior of modelled climate systems. Network analysis for Emergent Constraint (netEC) infers networks from complex and highly-dimensional datasets, by first reconstructing meaningful components describing regional subprocesses, and secondly inferring the causal links between these to construct causal networks. We apply this methodology to Sea Surface Temperature (SST) simulations and investigate two different metrics in order to derive an emergent constraint that yields likely ranges of ECS (2.34-4.79°C) and TCR (1.53-2.50°C). These ranges are narrower than the unconstrained distributions and consistent with the ranges of the Intergovernmental Panel on Climate Change (IPCC). More importantly, our simplified representation of simulations suggests the low-sensitivity models (ECS < 3.0°C) and high sensitivity models (ECS ≥ 4.5°C) remain two groups of plausible models.

show abstract

Section: δ-Maps Methodsmentioning

confidence: 99%

netEC: an emergent constraint on Climate Sensitivity based on network analysis

Nenes,

Ricard,

Falasca

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…The potential application of artificial intelligence or machine learning (AI/ML) to geophysical problems offers an alternative platform for storm surge simulation, mitigating computation costs and complex model setup requirements. Recent studies have utilized AI/ML approaches, such as random forests and linear regression, for sea level and storm surge simulations globally 18 . A data-driven model, termed DSSRT (Data-Driven Model Storm Surge Reconstruction 19 ), outperforms in extratropical regions but exhibits poor simulations in the tropics (average correlation of 0.45).…”

Section: Mainmentioning

confidence: 99%

Rapid Emergence of the Indian Ocean Extreme Sea Level

PUTHIYADATH,

P.,

Singh

et al. 2024

Preprint

View full text Add to dashboard Cite

Mean sea level rise (MSLR) and weather extremes can drive extreme sea level (ESL) variations locally. In the Indian Ocean, ESL estimates under global warming are either absent or limited by MSLR alone or biased storm surge model simulations. Using tide gauge, machine learning, and numerical models, we identify a rapid emergence of ESL in the Indian Ocean, particularly in the equatorial region compared to off-equatorial areas. Equatorial islands will experience the one-in-a-hundred-year ESL (ESL100) of the present-day annually by 2030-40 under a high-emission scenario, with delay until 2050 for the Arabian Sea coastline and the south subtropical regions. MSLR will mainly contribute to future ESL changes, with tide and surges contributing less than 10%. A median rise of 60–80 cm in ESL is anticipated by 2100, demanding coastal planning and climate adaptation strategies for a resilient coastal population.

show abstract

“…At regional scale, closing the SL budget is more challenging due to the higher variance of the signal, although there have been some attempts to explore the closure from local (Royston et al, 2020) to large basin-wide scales (Purkey et al, 2014). Instead, Camargo et al (2023) used unsupervised machine learning techniques to identify regions of coherent sea level variability.…”

Section: Sea Level Budgetmentioning

confidence: 99%

Sea Level Rise in Europe: Observations and projections

Melet,

van de Wal,

Amores

et al. 2023

Preprint

View full text Add to dashboard Cite

Abstract. Sea level rise (SLR) is a major concern for Europe, where 30 million people live in the historical 1-in-100-year event flood coastal plains. The latest IPCC assessment reports provide a literature review on past and projected SLR, and their key findings are synthesized here with a focus on Europe. The present paper complements IPCC reports and contributes to the Knowledge Hub on SLR European Assessment Report. Here, the state of knowledge of observed and 21st century projected SLR and changes in extreme sea levels (ESLs) are documented with more regional information for European basins as scoped with stakeholders. In Europe, satellite altimetry shows that absolute sea level trends are on average slightly above the global mean rate, with only a few areas showing no change or a slight decrease such as central parts of the Mediterranean Sea. The spatial pattern of absolute SLR in European Seas is largely influenced by internal climate modes, especially the North Atlantic Oscillation, which varies on year-to-year to decadal timescales. In terms of relative sea level rise (RSLR), vertical land motions due to human induced subsidence and glacial isostatic adjustment (GIA) are important for many coastal European regions, leading to lower or even negative RSLR in the Baltic Sea, and to large rates of RSLR for subsiding coastlines. Projected 21st century local SLR for Europe is broadly in-line with projections of GMSLR rise in most places. Some European coasts are projected to experience a relative SLR by 2100 below the projected GMSLR, such as the Norwegian coast, the southern Baltic Sea, the northern part of the UK and Ireland. A relative sea level fall is projected for the northern Baltic Sea. RSLR along other EU coasts is projected to be slightly above the GMSLR, for instance the Atlantic coasts of Portugal, Spain, France, Belgium and the Netherlands. Higher-resolution regionalized projections are needed to better resolve dynamic sea level changes especially in semi-enclosed basins, such as the Mediterranean Sea, North Sea, Baltic Sea and Black Sea. In addition to ocean dynamics, GIA and Greenland ice mass loss and associated Earth gravity, rotation and deformation effects are important drivers of spatial variations of projected European RSLR. High-end estimates of SLR in Europe are particularly sensitive to uncertainties arising from the estimates of the Antarctic ice mass loss. Regarding ESLs, the frequency of occurrence of the historical centennial event (HCE) level is projected to be amplified for most EU coasts, except along the northern Baltic Sea coasts where a decreasing probability is projected because of relative sea level fall induced by GIA. The largest HCE amplification factors are projected for the southern European Seas (Mediterranean and Iberian Peninsula coasts), while the smallest amplification factors are projected in macro-tidal regions exposed to storms and induced large surges such as the south-eastern North Sea. Finally, an emphasis is given on processes that are especially important for specific regions, such as waves, tides in the north-eastern Atlantic; vertical land motion for the European Arctic and Baltic Sea; seiches, meteo-tsunamis and medicanes in the Mediterranean Sea; non-linear interactions between drivers of coastal sea level extremes in the shallow North Sea.

show abstract

Regionalizing the sea-level budget with machine learning techniques

Cited by 11 publications

References 100 publications

netEC: an emergent constraint on Climate Sensitivity based on network analysis

netEC: an emergent constraint on Climate Sensitivity based on network analysis

Rapid Emergence of the Indian Ocean Extreme Sea Level

Sea Level Rise in Europe: Observations and projections

Contact Info

Product

Resources

About