The Implications of Long-Term Local Climate Change for the Energy Performance of an nZEB Residential Building in Volos, Greece

Energies, Journal Year: 2025, Volume and Issue: 18(5), P. 1032 - 1032

Published: Feb. 20, 2025

The construction of nearly zero-emission buildings in Europe and internationally has become mandatory by legislation. In parallel with these developments, the non-reversible increase ambient temperatures stresses buildings’ energy systems during summer months extreme temperatures, their severity varying according to local microclimate. These phenomena result an cooling loads. Thus, HVAC system’s performance needs more careful study, especially for residential sector wherever night effect is no longer capable releasing stress. present work, impact climate change on a building’s studied through simulations. future increases intensity duration heat waves assessed exploiting long-term forecasting capabilities transformer neural network model, trained existing meteorological data period 2007–2023. Based forecasted climatic conditions 2030 2040 produced this way, projected effects are assessed. forecast was aided 43 years temperature Europe, available ERA5 Copernicus program datasets. respective predictions electricity consumption wave episodes long durations point necessity special measures keep internal grid’s autonomy reduce unwanted interactions external grid. Moreover, further improvements nZEB building design improved would be critical success policy next two decades.

Language: Английский

---

Digital twin model for analyzing deformation and seepage in high earth-rock dams

Automation in Construction, Journal Year: 2025, Volume and Issue: 173, P. 106079 - 106079

Published: Feb. 22, 2025

Language: Английский

---

Selected Topics in Time Series Forecasting: Statistical Models vs. Machine Learning

Entropy, Journal Year: 2025, Volume and Issue: 27(3), P. 279 - 279

Published: March 7, 2025

Machine learning forecasting methods are compared to more traditional parametric statistical models. This comparison is carried out regarding a number of different situations and settings. A survey the most used models given. methods, such as convolutional networks, TCNs, LSTM, transformers, random forest, gradient boosting, briefly presented. The practical performance various analyzed by discussing results Makridakis competitions (M1–M6). I also look at probability via GARCH-type modeling for integer time series continuous Furthermore, comment on entropy volatility measure. Cointegration panels mentioned. paper ends with section weather potential machine in context, including very recent GraphCast GenCast forecasts.

Language: Английский

---

Precipitation Nowcasting in Dar es Salaam: Comparative Analysis of LSTM and Bidirectional LSTM for Enhancing Early Warning Systems

Journal of Geoscience and Environment Protection, Journal Year: 2025, Volume and Issue: 13(04), P. 327 - 342

Published: Jan. 1, 2025

Language: Английский

---

SwinNowcast: A Swin Transformer-Based Model for Radar-Based Precipitation Nowcasting

Remote Sensing, Journal Year: 2025, Volume and Issue: 17(9), P. 1550 - 1550

Published: April 27, 2025

Precipitation nowcasting is pivotal in monitoring extreme weather events and issuing early warnings for meteorological disasters. However, the inherent complexity of precipitation systems, coupled with their nonlinear spatiotemporal evolution, poses significant challenges traditional numerical prediction methods capturing multi-scale details effectively. Existing deep learning models similarly struggle to simultaneously capture local features global long-term dependencies. To tackle this challenge, we propose SwinNowcast, a model based on Swin Transformer architecture. Through novel design feature balancing module (M-FBM), dynamically integrates local-scale Specifically, convolutional block attention (MSCBAM) captures features, while gated fusion unit (GAFFU) adaptively regulates intensity, thereby enhancing spatial structure temporal continuity synergistic manner. Experiments were performed dataset from Royal Netherlands Meteorological Institute (KNMI) under thresholds 0.5 mm, 5 10 mm. The results indicate that SwinNowcast surpasses six state-of-the-art approaches regarding critical success index (CSI) Heidke skill score (HSS), markedly reducing false alarm rate (FAR). proposed holds substantial practical value applications such as short-term heavy rainfall urban flood warning, offering effective technological support disaster mitigation.

Language: Английский

---

DRN-DSA: A Hybrid Deep Learning Network Model for Precipitation Nowcasting using Time Series Data

Knowledge-Based Systems, Journal Year: 2024, Volume and Issue: 306, P. 112679 - 112679

Published: Nov. 10, 2024

Language: Английский

---