Search

The intensification of extreme events, storm surges and coastal flooding in a climate change scenario increasingly influences human processes, especially in coastal areas where sea-based activities are concentrated. Predicting sea level near the coasts, with a high accuracy and in a reasonable amount of time, becomes a strategic task. Despite the developments of complex numerical codes for high-resolution ocean modeling, the task of making forecasts in areas at the intersection between land and sea remains challenging. In this respect, the use of machine learning techniques can represent an interesting alternative to be investigated and evaluated by numerical modelers. This article presents the application of the Long-Short Term Memory (LSTM) neural network to the problem of short-term sea level forecasting in the Southern Adriatic Northern Ionian (SANI) domain in the Mediterranean sea. The proposed multi-model architecture based on LSTM networks has been trained to predict mean sea levels three days ahead, for different coastal locations. Predictions were compared with the observation data collected through the tide-gauge devices as well as with the forecasts produced by the Southern Adriatic Northern Ionian Forecasting System (SANIFS) developed at the Euro-Mediterranean Center on Climate Change (CMCC), which provides short-term daily updated forecasts in the Mediterranean basin. Experimental results demonstrate that the multi-model architecture is able to bridge information far in time and to produce predictions with a much higher accuracy than SANIFS forecasts.

In recent decades, the complexity of scientific and engineering problems has increased considerably. New applications and domains that use high performance computing systems have been introduced. These trends are projected to continue for the foreseen future (Reed and Dongarra, 2015) [ 1 ]. In many areas of engineering and science, High-Performance Computing (HPC) and Simulations have become determinants of industrial competitiveness and advanced research. In fact, advances in HPC architectures, storages, networking, and software capabilities are leading to a new era in HPC and simulations, along with new challenges both in computing and systems modeling (Geist and Lucas, 2009) [ 2 ]. These developments are especially critical considering that HPC systems continue to scale up in terms of nodes, cores, and accelerators, as well as software, infrastructure and tools, which in turn are expediting the move on the path toward Exascale (Reed and Dongarra, 2015; Geist and Lucas, 2009; Dongarra and Beckman, 2011; Dosanjh et al., 2014; Engelmann, 2014) [ [1] , [2] , [3] , [4] , [5] ]. Scalability and availability represent two of the main requirements that need to be considered before conceiving of these large-scale systems (ASCAC Subcommittee on Exascale Computing, 2010). The scalability feature allows the system to proportionately grow when service demand increases, whereas availability means the system continues to provide their services despite hardware and software failures (Theodoropoulos et al., 2014; Tang et al., 2014) [ [7] , [8] ]. The goal in large-scale HPC is to accommodate both availability and scalability while staying under strict constraints on performance (e.g., processing time) and cost metrics (e.g., power consumption). This special issue is envisioned to provide examples of research work on topics related to recent advances in High Performance Computing and Simulations. It briefly addresses and explores challenges toward Exascale computing, current state-of-the-art in HPC and simulation, and the path forward in the domains of large-scale HPC systems.

The evolution of High-Performance Computing (HPC) platforms enables the design and execution of progressively larger and more complex workflow applications in these systems. The complexity comes not only from the number of elements that compose the workflows but also from the type of computations they perform. While traditional HPC workflows target simulations and modelling of physical phenomena, current needs require in addition data analytics (DA) and artificial intelligence (AI) tasks. However, the development of these workflows is hampered by the lack of proper programming models and environments that support the integration of HPC, DA, and AI, as well as the lack of tools to easily deploy and execute the workflows in HPC systems. To progress in this direction, this paper presents use cases where complex workflows are required and investigates the main issues to be addressed for the HPC/DA/AI convergence. Based on this study, the paper identifies the challenges of a new workflow platform to manage complex workflows. Finally, it proposes a development approach for such a workflow platform addressing these challenges in two directions: first, by defining a software stack that provides the functionalities to manage these complex workflows; and second, by proposing the HPC Workflow as a Service (HPCWaaS) paradigm, which leverages the software stack to facilitate the reusability of complex workflows in federated HPC infrastructures. Proposals presented in this work are subject to study and development as part of the EuroHPC eFlows4HPC project., This work has received funding from the European High-Performance Computing Joint Undertaking (JU) under grant agreement No 955558. The JU receives support from the European Union’s Horizon 2020 research and innovation programme and Spain, Germany, France, Italy, Poland, Switzerland and Norway. In Spain, it has received complementary funding from MCIN/AEI/10.13039/501100011033, Spain and the European Union NextGenerationEU/PRTR (contracts PCI2021-121957, PCI2021-121931, PCI2021-121944, and PCI2021-121927). In Germany, it has received complementary funding from the German Federal Ministry of Education and Research (contracts 16HPC016K, 6GPC016K, 16HPC017 and 16HPC018). In France, it has received financial support from Caisse des dépôts et consignations (CDC) under the action PIA ADEIP (project Calculateurs). In Italy, it has been preliminary approved for complimentary funding by Ministero dello Sviluppo Economico (MiSE) (ref. project prop. 2659). In Norway, it has received complementary funding from the Norwegian Research Council, Norway under project number 323825. In Switzerland, it has been preliminary approved for complimentary funding by the State Secretariat for Education, Research, and Innovation (SERI), Norway. In Poland, it is partially supported by the National Centre for Research and Development under decision DWM/EuroHPCJU/4/2021. The authors also acknowledge financial support by MCIN/AEI /10.13039/501100011033, Spain through the “Severo Ochoa Programme for Centres of Excellence in R&D” under Grant CEX2018-000797-S, the Spanish Government, Spain (contract PID2019-107255 GB) and by Generalitat de Catalunya, Spain (contract 2017-SGR-01414). Anna Queralt is a Serra Húnter Fellow., With funding from the Spanish government through the ‘Severo Ochoa Centre of Excellence’ accreditation (CEX2018-000797-S).