Spring frost events occurring after budburst of grapevines can damage new shoots, disrupt plant growth and cause large economic losses to the viticultural sector. Frost protection practices encompass a variety of vineyard management actions across timescales, from seasonal to decadal and beyond. The cost-effectiveness of such measures depends on the availability of accurate predictions of the relevant climate hazards at the appropriate timescales.

In this work, we present a statistical downscaling method which predicts variations in the frequency of occurrence of spring frost events in the important winemaking region of Catalunya at the seasonal timescale. The downscaling method exploits the seasonal predictability associated with the predictable components of the atmospheric variability over the Euro-Atlantic region, and produces local predictions of frost occurrence at a spatial scale relevant to vineyard management.

The downscaling method is designed to address the specific needs highlighted by a representative stakeholder in the local viticultural sector, and is expected to deliver an actionable prototype climate service. The statistical procedure is developed in perfect prognosis mode: the method is trained with large-scale reanalysis data against a high-resolution gridded observational reference, and validated against multi-model seasonal hindcast predictions.

Our work spotlights the potential benefits of transferring climate predictability across spatial scales for the design and provision of usable climate information, particularly regarding extremes.

How to cite: Roncoroni, S., Athanasiadis, P., and Gualdi, S.: Statistical downscaling of extremes in seasonal predictions - a case study on spring frosts for the viticultural sector, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4538, https://doi.org/10.5194/egusphere-egu24-4538, 2024.

The Mediterranean region has been identified as an important climate change hotspot, over the 21st century both air temperature and its extremes are projected to rise at a rate surpassing that of the global average and a significant decrease of average summer precipitation is projected, particularly for the western Mediterranean. On average, Mediterranean droughts have become more frequent and intense in recent years and are expected to become more widespread in many regions. These prolonged dry spells pose a substantial threat to agriculture and impact several socio-economic sectors. In this context, long-range weather forecasting has emerged as a promising tool for seasonal drought risk assessment. However, the interpretation of the forecasting products is not always straightforward due to their inherent probabilistic nature. Therefore, a rigorous evaluation process is needed to determine the extent to which these forecasts provide a fruitful advantage over much simpler forecasting systems, such as those based on climatology. 

In this study, we use the latest version of ECMWF’s seasonal prediction system (SEAS5) to understand its skill in predicting summer droughts. The Standardized Precipitation Evapotranspiration Index (SPEI) aggregated over different lead times is employed to mark below-normal dryness conditions in August. We use a comprehensive set of evaluation metrics to gain insight into the accuracy, systematic biases, association, discrimination and sharpness of the forecast system. Our findings reveal that up to 3 months lead time, seasonal forecasts show stronger association and discrimination skills than the climatological forecast, especially in the Southern Mediterranean, although the prediction quality in terms of accuracy and sharpness is limited. On the other hand, extending the forecast range up to 6 months lead time dramatically reduces its predictability skill, with the system mostly underperforming elementary climatological predictions. 

This approach is then extended to examine the full ensemble of seasonal forecasting systems provided by the Copernicus Climate Change Service (C3S) to test their skill in predicting droughts. Our findings can help an informed use of seasonal forecasts of droughts and the development of related climate services.

How to cite: Cerato, G., Bellomo, K., and von Hardenberg, J.: Summer drought predictability in the Mediterranean region in seasonal forecasts, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11485, https://doi.org/10.5194/egusphere-egu24-11485, 2024.

Accurate forecasts of the natural resources of renewable energy production have become not only a valuable but a crucial tool for managing the associated risks of specific events, such as wind droughts. Wind energy, alongside with solar power, now provide a substantial part to the renewable energy share of the global energy production and growth in this sector will most likely further increase. The naturally given fluctuations of wind resources, however, pose a challenge for maintaining a stable energy supply, which, at the end of the chain, can have an impact on the energy market prices.
Operational short-term forecasting products for the wind energy sector (multiple days) are already commonly available and seasonal to sub seasonal forecasting solutions (multiple months) can provide valuable skill and are gaining in popularity. On the other side of the spectrum, typically on a time scale of multiple decades, we find risk assessment based on climate change projections. In between the long and short term time scales, however, there is a gap that still needs to be filled to achieve seamless prediction of risks that are relevant for the energy sector: decadal predictions.

Here, we present the results of an evaluation study of a multi-model decadal prediction ensemble (DCPP) for a selection of wind development regions in Europe. The evaluation is based on multiple decades long hindcasts and carried out with a focus on the skill of predicting specific event types of wind resource availability in a probabilistic context, alongside with basic deterministic skill measures. We further investigate specific event constellations and their large-scale drivers that, in combination, can provide windows of opportunity with enhanced predictive skill. We conclude with a discussion on how this hybrid approach can be used to potentially increase not only forecast skill but also the trust of the end user.

How to cite: Lochbihler, K., Lopez, A., and Lizcano, G.: Skill of wind resource forecasts on the decadal time scale, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-17585, https://doi.org/10.5194/egusphere-egu24-17585, 2024.

The Mediterranean region is highly sensitive to climate change, having experienced an intense warming and drying trend in recent decades, primarily due to the increased concentrations of anthropogenic greenhouse gases. In the context of decision-making processes, there is a growing interest in understanding the near-term climate evolution of this region.

In this study, we explore the climatic fluctuations of the Mediterranean region in the near-term range (up to 10 years ahead) using two different products: projections and decadal predictions. The former are century-scale climate change simulations initialized from arbitrary model states to which were applied anthropogenic and natural forcings. A major limitation of climate projections is their limited information regarding the current state of the Earth’s climate system. Decadal climate predictions, obtained by constraining the initial conditions of an ensemble of model simulations through a best estimate of the observed climate state, provide a better understanding of the next-decade climate and thus represent an invaluable tool in assisting climate adaptation.

Using retrospective forecasts from eight decadal prediction systems contributing to the CMIP6 Decadal Climate Prediction Project (CMIP6 DCPP) and the corresponding ensemble of non-initialized projections, we compare the capabilities of the state-of-the-art climate models in predicting future climate changes of the Mediterranean region for some key quantities so as to assess the added value of initialization. 

Beyond the contribution of external forcings, the role of internal variability is also investigated since part of the detected predictability arises from internal climate variability patterns affecting the Mediterranean. The observed North Atlantic Oscillation, the dominant climate variability pattern in the Euro-Atlantic domain, as well as its  impact on wintertime precipitation over Europe are well reproduced by decadal predictions, especially over the Mediterranean, outperforming projections. We also apply a sub-sampling method to enhance the respective signal-to-noise ratio and consequently improve precipitation skill over the Mediterranean.

How to cite: Nicolì, D., Gualdi, S., and Athanasiadis, P.: Decadal predictions outperform projections in forecasting winter precipitation over the Mediterranean region, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16366, https://doi.org/10.5194/egusphere-egu24-16366, 2024.

Storm surges and elevated water levels regularly challenge coastal protection and inland water management along the low-lying coastline of the German Bight. Skillful seasonal-to-decadal (S2D) predictions of the local storm surge climate would be beneficial to stakeholders and decision makers in the region. While storm activity has recently been shown to be skillfully predictable on a decadal timescale with a global earth system model, surge modelling usually requires very fine spatial and temporal resolutions that are not yet present in current earth system models. We therefore propose an alternative approach to generating S2D predictions of the storm surge climate by training a neural network on observed water levels and large-scale atmospheric patterns, and apply the neural network to the available model output of a S2D prediction system. We show that the neural-network-based translation from large-scale atmospheric fields to local water levels at the coast works sufficiently well, and that several windows of predictability for the German Bight surge climate emerge on the S2D scale.

How to cite: Krieger, D., Brune, S., Baehr, J., and Weisse, R.: Exploring ML-based decadal predictions of the German Bight storm surge climate, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14688, https://doi.org/10.5194/egusphere-egu24-14688, 2024.

Global mean surface temperature is the key metric by which our warming climate is monitored and for which international climate policy is set. At the end of each year the Met Office makes a global mean temperature forecast for the coming year. Following on from the new record 2023, we predict a high probability of another record year in 2024 and a 35% chance of exceeding 1.5 C above pre-industrial. Whilst a one-year temporary exceedance of 1.5 C would not constitute a breech of the Paris Agreement target, our forecast highlights how close we are now to breeching this target. We show that our 2024 forecast can be largely explained by the combination of the continuing warming trend of +0.2 C/decade and the lagged warming affect of a strong tropical Pacific El Nino event. We further highlight 2023 was significantly warmer than forecast and that much of this warming signal came from the southern hemisphere and requires further understanding.

How to cite: Dunstone, N., Smith, D., Scaife, A., Hermanson, L., Colman, A., and Folland, C.: Will 2024 be the first year above 1.5 C?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10574, https://doi.org/10.5194/egusphere-egu24-10574, 2024.

The Earth has warmed significantly over the past 40 years, and the fastest rate of warming has occurred in and around the Arctic. The warming of northern high latitudes at a rate of almost four times the global average (Rantanen et al., 2022), known as Arctic amplification, is associated with sea ice loss, glacier retreat, permafrost degradation, and expansion of the melting season. Since the mid-2000s, summer sea ice has exhibited a rapid decline, reaching record minima in September sea ice area in 2007 and 2012. However, after the early 2010s, the downward trend of minimum sea ice area appears to decelerate (Swart et al., 2015; Baxter et al., 2019). This apparent slowdown and the preceding acceleration in the rate of sea ice loss are puzzling in light of the steadily increasing rate of greenhouse gas emissions of about 4.5 ppm yr⁻¹ over the past decade (Friedlingstein et al., 2023) that provides a constant climate forcing. Recent studies suggest that low-frequency internal climate variability may have been as important as anthropogenic influences on observed Arctic sea ice decline over the past four decades (Dörr et al., 2023; Karami et al., 2023). Here, we investigate how unusual this decade-long pause in Arctic summer sea ice decline is within the context of internal climate variability. To do so, we first assess how rare this is deceleration of Arctic sea ice loss is by comparing it to trends in CMIP6 historical simulations. We also use simulations from the Decadal Climate Prediction Project (DCPP) contribution to CMIP6 to determine if initializing decadal prediction systems from estimates of the observed climate state substantially improves their performance in predicting the slowdown in Arctic sea ice loss over the past decade. As the DCPP does not specify the data or the methods to be used to initialize forecasts or how to generate ensembles of initial conditions, we also assess how different formulations affect the skill of the forecasts by analyzing differences between models. This work provides an opportunity to attribute this pause in Arctic sea ice retreat to interannual internal variability or radiative external forcings, something that observation analysis alone cannot achieve.

How to cite: DeRepentigny, P., Massonnet, F., Bilbao, R., and Materia, S.: How unusual is the recent decade-long pause in Arctic summer sea ice retreat?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12969, https://doi.org/10.5194/egusphere-egu24-12969, 2024.

Accurate forecasts of the climate of the coming season and years are highly desired by many sectors of society. The skill of near-term climate prediction in winter in the North Atlantic and European region has improved over the last decade associated with larger ensembles, improving models and boosting of the prediction signal using intelligent post processing. International collaboration has improved the availability of forecasts and promoted the uptake of forecasts by different sectors. However, significant challenges remain, including summer prediction, understanding the risk of extremes within a season, multi-seasonal extremes and how best to post process the forecasts to aid decision making. This talk will summarise recent near-term climate prediction research activities at the UK Met Office and will detail our experience of providing such forecasts to the energy and water sectors.  

How to cite: Thornton, H.: The opportunities and challenges of near-term climate prediction, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19251, https://doi.org/10.5194/egusphere-egu24-19251, 2024.

Atlantic meridional overturning circulation (AMOC) is one of the mechanisms for climate predictability and one of the properties that decadal climate predictions are attempting to predict. The starting point for AMOC decadal predictions is sensitive to the underlying data assimilation and/or initialization procedure. This means that different choices during the data assimilation procedure (e.g., assimilation method, assimilation window, data sources, resolution, nudging terms and strength, full field vs anomaly initialization/assimilation, etc) can result in a different mean and even variability of reconstructed ocean circulation. How coherent the AMOC initial states should be among the CMIP-like decadal prediction experiments? How good in general should the initial AMOC be for decadal predictions? And do initialization issues of the ocean circulation influence the prediction skill of other variables that are of interest for application studies? These are the questions that we were attempting to address in our study, where we analyzed twelve decadal prediction systems from the World Meteorological Organization Lead Centre for Annual-to-Decadal Climate Prediction project. We identify that the AMOC initialization influences the quality of predictions of the subpolar gyre (SPG). When predictions show a large initial error in their AMOC, they usually have low skill for predicting the internal variability of the SPG five years after the initialization.

How to cite: Polkova, I. and the Co-Authors: Initialization shock in the ocean circulation reduces skill in decadal predictions of the North Atlantic subpolar gyre, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15358, https://doi.org/10.5194/egusphere-egu24-15358, 2024.

Skillful prediction of seasonal monsoons has been a challenging problem since the 1800s. However, significant progress has been made in Indian summer monsoon rainfall prediction in recent times, with skill scores reaching 0.6 and beyond, surpassing the estimated predictability limits. This phenomenon leads to what is known as the “Signal-to-noise Paradox.” To investigate this paradox, we utilized 52 ensemble member hindcast runs spanning 30 years.

Through the application of ANOVA and Mutual Information methods, we estimate the predictability limit globally. Notably, for the boreal summer rainfall season, the Indian subcontinent exhibited the paradox, among several other regions, while the Equatorial Pacific region, despite demonstrating high prediction skill, does not have the Signal-to-Noise paradox. We employed a novel approach to understand how sub-seasonal variability and their projection in association with predictors are linked to the paradoxical behavior of seasonal prediction skill.

We propose a new method to estimate predictability limits that is free from paradoxical phenomena and shows much higher seasonal predictability. This novel method provides valuable insights into the complex dynamics of monsoon prediction, thereby creating opportunities for expanded research and potential improvements in seasonal forecasting skill in the coming years.

How to cite: Shivamurthy, Y., Saha, S. K., Pokhrel, S., Konwar, M., and Chaudhari, H. K.: Why does the Signal-to-Noise Paradox Exist in Seasonal Climate Predictability?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4873, https://doi.org/10.5194/egusphere-egu24-4873, 2024.

How to cite: Christiansen, B., Yang, S., and Drews, A.: The Atlantic Multi-decadal Variability in observations and in a large historical multi-model ensemble: Forced and internal variability, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9100, https://doi.org/10.5194/egusphere-egu24-9100, 2024.

The European North-West shelf seas (NWS) support economic interests and provide environmental services to several adjacent populous countries. Skilful seasonal forecasts of the NWS would be useful to support decision making. Here, we quantify the skill of an operational large-ensemble ocean-atmosphere coupled dynamical forecasting system (GloSea), as well as a benchmark persistence forecasting system, for predictions of NWS sea surface temperature (SST) at 2-4 months lead time in winter and summer. We also identify sources of- and limits to NWS SST predictability with a view to what additional skill may be available in the future. We find that GloSea NWS SST skill is generally high in winter and low in summer. Persistence of anomalies in the initial conditions contributes substantially to predictability. GloSea outperforms simple persistence forecasts, by adding atmospheric variability information, but only to a modest extent. Where persistence is low – for example in seasonally stratified regions – both GloSea and persistence forecasts show lower skill. GloSea skill can be degradeded by model deficiencies in the relatively coarse global ocean component, which lacks a tidal regime and likely fails to properly fine-scale NWS physics. However, using “near perfect atmosphere” tests, we show potential for improving predictability of currently low performing regions if atmospheric circulation forecasts can be improved, underlining the importance of development of atmosphere-ocean coupled models for NWS seasonal forecasting applications.

How to cite: Atkins, J., Tinker, J., Graham, J., Scaife, A., and Halloran, P.: Seasonal forecasting of the European North-West shelf seas: limits of winter and summer sea surface temperature predictability, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3190, https://doi.org/10.5194/egusphere-egu24-3190, 2024.

Sea surface temperature (SST) changes in the Mediterranean Sea have profound impacts on both the Mediterranean regions and remote areas. Previous studies show that the Mediterranean SST has significant decadal variability that is comparable with the Atlantic multidecadal variability (AMV). However, few studies have discussed the characteristics and sources of the decadal predictability of Mediterranean SST based on observations. Here for the first time we use observational datasets to reveal that the decadal predictability of Mediterranean SST is contributed by both external forcings and internal variability for both annual and seasonal means, except that the decadal predictability of the winter mean SST in the eastern Mediterranean is mostly contributed by only internal variability. Besides, the persistence of the Mediterranean SST is quite significant even in contrast with that in the subpolar North Atlantic, which is widely regarded to have the most predictable surface temperature on the decadal time scale. After the impacts of external forcings are removed, the average prediction time of internally generated Mediterranean SST variations is more than 10 years and closely associated with the multidecadal variability of the Mediterranean SST that is closely related to the accumulated North Atlantic Oscillation forcing.

How to cite: Yan, X., Tang, Y., and Yang, D.: Study of the Decadal Predictability of Mediterranean Sea Surface Temperature Based on Observations, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1940, https://doi.org/10.5194/egusphere-egu24-1940, 2024.

The inter-annual to multi-decadal variability of recurrent, synoptic-scale atmospheric circulation patterns in the Northern Hemisphere extratropics, as represented by the Jenkinson-Collison classification scheme, is explored in reanalysis data spanning the entire 20th century, and in global climate model (GCM) data from the historical, AMIP and DCPP experiments conducted within the framework of CMIP6. The aim of these efforts is to assess the effect of coupled vs. uncoupled and initialised vs. non-initialized GCM simulations in reproducing the observed low-frequency variability of the aforementioned circulation patterns.

Results reveal that the observed annual counts of typical recurrent weather patterns, such as cyclonic or anticyclonic conditions and also situations of pronounced advection, exhibit significant oscillations on multiple time-scales ranging between several years and several decades. The period of these oscillations, however, is subject to large regional variations. This is in line with earlier studies suggesting that the extratropical atmospheric circulation’s low frequency variability is essentially unforced, except in the Pacific-North American sector where the forced variability is enhanced due to ENSO teleconnections. Neither the periods obtained from historical nor those obtained from AMIP experiments align with observations. Likewise, not even the periods obtained from different runs of the same GCM and experiment correspond to each other. Thus, in an non-initialized model setup, ocean-atmosphere coupling or the lack thereof essentially leads to the same results. Whether initialization and/or augmenting the ensemble size can improve these findings, will also be discussed.

Acknowledgement: This work is part of project Impetus4Change, which has received funding from the European Union’s Horizon Europe research and innovation programme under grant agreement No 101081555.

How to cite: Brands, S., Cimadevilla, E., and Fernández, J.: Low-frequency variability of synoptic-scale atmospheric circulation patterns in the Northern Hemisphere extratropics and associated hindcast skill of decadal forecasting systems, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10551, https://doi.org/10.5194/egusphere-egu24-10551, 2024.

The recent severe European summer heat waves of 2015 and 2018 co-occurred with cold subpolar North Atlantic (NA) sea surface temperatures (SSTs). However, a significant connection between this oceanic state and European heat waves was not yet established.

We investigate the effect of cold subpolar NA SSTs on European summer heat waves using two 100-year long AMIP-like model experiments: one that employs the observed global 2018 SST pattern as a boundary forcing and a counter experiment for which we removed the negative NA SST anomaly from the 2018 SST field, while preserving daily and small-scale SST variabilities. Comparing these experiments, we find that cold subpolar NA SSTs significantly increase heat wave duration and magnitude downstream over the European continent. Surface temperature and circulation anomalies are connected by the upper-tropospheric summer wave pattern of meridional winds over the North Atlantic European sector, which is enhanced with cold NA SSTs. Our results highlight the relevance of the subpolar NA region for European summer conditions, a region that is marked by large biases in current coupled climate model simulations.

How to cite: Bischof, S., Pilch Kedzierski, R., Hänsch, M., Wahl, S., and Matthes, K.: The Role of the North Atlantic for Heat Wave Characteristics in Europe, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9274, https://doi.org/10.5194/egusphere-egu24-9274, 2024.

As a consequence of limited reliability of dynamical forecast systems, particularly over Europe, efforts in recent years have turned to exploiting the power of Machine Learning methods to extract information on drivers of extreme temperature from observations and reanalysis. Meanwhile, the diverse impacts of extreme heat have driven development of new indicators which take into account nightime temperatures and humidity. In the H2020 CLimate INTelligence (CLINT) project, a feature selection framework is being developed to find the combination of drivers which provides optimal seasonal forecast skill of European summer heatwave indicators. Here, we present the methodology, its application to a range of heatwave indicators and forecast skill compared to existing dynamical systems. First, a range of (reduced-dimensionality) drivers are defined, including k-means clusters of variables known to impact European summer (e.g. precipitation, sea ice content), and more complex indices like the NAO and weather regimes. Then, these drivers are used to train machine learning based prediction models, of varying complexity, to predict seasonal indicators of heatwave occurrence and intensity. A crucial and novel step in our framework is the use of the Coral Reef Optimisation algorithm, used to select the variables and their corresponding lag times and time periods which provide optimal forecast skill. To maximise training data, both ERA5 reanalysis and a 2000-year paleo-simulation are used; the representation of heatwaves and atmospheric conditions are validated with respect to ERA5. We present comparisons of forecast skill to the dynamical Copernicus Climate Change Service seasonal forecasts systems. The differences in timing, predictability and drivers of daytime and nighttime heatwaves across Europe are highlighted. Lastly, we discuss how the framework can easily be adapted to other extremes and timescales.

How to cite: McAdam, R., Peláez Rodríguez, C., Hansen, F., Pérez Aracil, J., Squintu, A., Cavicchia, L., Zorita, E., Saldez-Sanz, S., and Scoccimarro, E.: Optimization-based driver detection and prediction of seasonal heat extremes, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9905, https://doi.org/10.5194/egusphere-egu24-9905, 2024.

This study focuses on applying machine learning techniques to bias-correct the seasonal temperature forecasts provided by the Copernicus Climate Change Service (C3S) models. Specifically, we employ bias correction on forecasts from five major models: UK Meteorological Office (UKMO), Euro-Mediterranean Center on Climate Change (CMCC), Deutscher Wetterdienst (DWD), Environment and Climate Change Canada (ECCC), and Meteo-France. Our primary objective is to assess the performance of our bias correction model in comparison to the original forecast datasets. We utilise temperature-based indices recommended by the Expert Team on Climate Change Detection and Indices (ETCCDI) to evaluate the effectiveness of the bias-corrected seasonal forecasts. These indices served as valuable metrics to gauge the predictive capability of the models, especially in forecasting natural cascading hazards such as wildfires, droughts, and floods. The study involved an in-depth analysis of the bias-corrected forecasts, and the derived indices were crucial in understanding the models' ability to predict temperature-related extreme events. The results of this research contribute valuable information for decision-making and planning across various sectors, including disaster risk management and environmental protection. Through a comprehensive evaluation of machine learning-based bias correction techniques, we enhance the accuracy and applicability of seasonal temperature forecasts, thereby improving preparedness and resilience to climate-related challenges. 

How to cite: Mbuvha, R. and Nikraftar, Z.: Machine Learning Approaches to Improve Accuracy in Extreme Seasonal Temperature Forecasts: A Multi-Model Assessment , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19297, https://doi.org/10.5194/egusphere-egu24-19297, 2024.

How to cite: Bachelery, M.-L., Brajard, J., Patacchiola, M., and Keenlyside, N.: Predicting Atlantic and Benguela Niño events with deep learning , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15709, https://doi.org/10.5194/egusphere-egu24-15709, 2024.

Long-term sea level rise and multiyear-to-decadal sea level variations pose substantial risks of flooding and erosion in coastal communities. The North Atlantic Ocean and the U.S. East Coast are hotspots for sea level changes under current and future climates. Here, we employ a machine learning technique, a self-organizing map (SOM)-based framework, to systematically characterize the North Atlantic sea level variability, assess sea level predictability, and generate sea level predictions on multiyear-to-decadal timescales. Specifically, we classify 5000-year North Atlantic sea level anomalies from the Seamless System for Prediction and EArth System Research (SPEAR) model control simulations into generalized patterns using SOM. Preferred transitions among these patterns are further identified, revealing long-term predictability on multiyear-to-decadal timescales related to shifts in Atlantic meridional overturning circulation (AMOC) phases. By combining the SOM framework with “analog” techniques based on the simulations and observational/reanalysis data, we demonstrate prediction skill of large-scale sea level patterns comparable to that from initialized hindcasts. Moreover, additional source of short-term predictability is identified after the exclusion of low-frequency AMOC signals, which arises from the wind-driven North Atlantic tripole mode triggered by the North Atlantic Oscillation. This study highlights the potential of machine learning methods to assess sources of predictability and to enable efficient, long-term climate prediction.

How to cite: Gu, Q., Jia, L., Zhang, L., Delworth, T., Yang, X., Zeng, F., and Li, S.: Exploring multiyear-to-decadal North Atlantic sea level predictability using machine learning and analog methods, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10539, https://doi.org/10.5194/egusphere-egu24-10539, 2024.

The implementation of adaptation policies requires seamless and relevant information on the evolution of the climate over the next decades. Decadal climate predictions are subject to drift because of intrinsic model errors and their skill may be limited after a few years or even months depending on the region. Non-initialized ensembles of climate projections have large uncertainties over the next decades, encompassing the full range of uncertainty attributed to internal climate variability. Providing the best climate information over the next decades is therefore challenging. Recent studies have started to address this challenge by constraining uninitialized projections of sea surface temperature using decadal predictions or using a storyline approach to constrain uninitialized projections of the Atlantic Meridional Overturning Circulation using observations. Here, using a hierarchical clustering method, we select a sub-ensemble of non-initialized climate simulations based on their similarity to observations. Then, we try to further refine this sub-ensemble of trajectories by selecting a subset based on its consistency with decadal predictions. This study presents a comparison of these different methods for constraining surface temperatures in the North-Atlantic / Europe region over the next decades, focusing on CMIP6 non-initialized simulations.

How to cite: Bonnet, R., Boé, J., and Sanchez, E.: Constraining near to mid-term climate projections by combining observations with decadal predictions, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8028, https://doi.org/10.5194/egusphere-egu24-8028, 2024.

Compound Dry and Hot Extremes (CDHE) have an adverse impact on socioeconomic factors during the Indian summer monsoon, and a future exacerbation is anticipated. The occurrence of CDHE is influenced by teleconnections, which play a crucial role in determining its likelihood on a seasonal scale. Despite the importance, there is a lack of studies unravelling the teleconnections of CDHE in India. Previous investigations specifically focused on teleconnections between precipitation, temperature, and climate indices. Hence, there is a need to unravel the teleconnections of CDHE. This study presents a framework combining event coincidence analysis (ECA) with complexity science. ECA evaluates the synchronization between CDHE and climate indices. Subsequently, complexity science is utilized to construct a driver-CDHE network to identify the critical drivers of CDHE. A logistic regression model is employed to evaluate the proposed drivers' effectiveness. The occurrence of CDHE exhibits distinct patterns from July to September when considering intra-seasonal variability. Our findings contribute to the identification of drivers associated with CDHE. The primary driver for Eastern, Western India and Central India is the indices in the Pacific Ocean and Atlantic Ocean, respectively, followed by the indices in the Indian Ocean. These identified drivers outperform the traditional Niño 3.4-based predictions. Overall, our results demonstrate the effectiveness of integrating ECA and complexity science to enhance the prediction of CDHE occurrences.

How to cite: Agarwal, A. and Guntu, R.: Towards the Predictability of Compound Dry and Hot Extremes through Complexity Science, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-7134, https://doi.org/10.5194/egusphere-egu24-7134, 2024.