The Norwegian Meteorological Institute (Norwegian: Meteorologisk institutt), also known internationally as MET Norway, is Norway's national meteorological institute. It provides weather forecasts for civilian and military uses and conducts research in meteorology, oceanography and climatology. It is headquartered in Oslo and has offices and stations in other cities and places. It has around 500 full-time staff and was founded in 1866..
This study introduces a new unstructured computational mesh for hydrodynamic simulations of the Oslofjord. The mesh was created with global bathymetry and shoreline data, using OceanMesh2D. It contains 70,410 nodes, with a resolution at the coastline of 50 m. We use the new mesh to create an advanced circulation model of the fjord. The model is run for four time periods with different characteristics, and validated against the current state of the art and elevation gauges in the fjord.Initial analyses show that the model effectively propagates tides through the narrow channels of the fjord, while requiring minimal computation time. Three different combinations of tidal constituents are used to force the model, and we analyze the cost and benefits of using additional constituents. In order to assess the model in a realistic scenario, the water surface elevation output from a coarse global model is used to force the model for an extreme weather event. The results demonstrate the mesh’s ability to reproduce high-resolution total water levels in the fjord, with a maximum root mean square error of just over 10 cm.
Many forecast applications require high frequency temporal resolution, yet most state-of-the-art data-driven weather forecasting systems operate at 6-hourly resolution. Although direct hourly forecasting is possible, it suffers from error accumulation and temporal inconsistency. We introduce HourGlass, a probabilistic data-driven temporal downscaling method that reconstructs the evolution between forecast states. HourGlass is trained using variants of the continuous ranked probability score (CRPS) preserving small-scale spatial variability while encouraging temporal consistency. Unlike existing deterministic temporal downscaling approaches, which tend to produce overly smooth fields, HourGlass generates realistic probabilistic forecasts. Training on forecast trajectories rather than reanalysis or analysis data also avoids the temporal inconsistencies present in datasets used by previous methods. We evaluate HourGlass in two settings: AIFS-HourGlass, applied globally to ECMWF's AIFS-Single and AIFS-ENS forecast systems, and Bris-HourGlass, applied regionally to MET Norway's high-resolution stretched-grid ensemble model, Bris. Verification against observations shows that both models retain the skill of their underlying forecasting systems while producing temporally coherent hourly forecasts with realistic small-scale variability. Case studies demonstrate physically consistent evolution during rapidly developing weather events, including extratropical cyclones and organised convection. Hourly precipitation remains challenging: HourGlass improves the spatial realism of precipitation fields but still underestimates the most intense extremes, a common limitation of data-driven weather forecasting models. These results demonstrate that HourGlass effectively bridges the gap between 6-hourly data-driven forecasts and the hourly products required for operational regional and global forecasting.
Sparse point observations are increasingly available for precipitation nowcasting, but it is unclear how much they improve dense radar-field forecasts. We partially address this question with a multimodal graph neural network nowcasting system over the Nordic radar domain. The model predicts rain rate every five minutes up to two hours ahead and is trained with different combinations of radar history, MEPS numerical weather prediction, Netatmo surface observations, MSG satellite channels, stochastic noise, and CRPS-based ensemble losses. The study is designed as an ablation of operationally relevant information sources and training objectives. We compare radar-only, NWP-informed, station-informed, satellite-informed, noise-augmented, and CRPS-based configurations using complementary diagnostics on the radar grid, at station locations, for rain onset, and through oracle, displacement, and amplitude scores. The results show that each source improves a different part of the forecast problem. MEPS stabilises radar-only extrapolation, Netatmo observations improve local station and onset diagnostics, and satellite predictors reduce some station-level biases but may activate rain too early when used deterministically. CRPS-based configurations provide the most consistent radar-grid gains, while the combined satellite and CRPS setup gives the best overall oracle/DAS score. These results do not support the conclusion that point observations are uninformative for nowcasting, but they show that local observational skill and spatially coherent radar-field skill are distinct targets. The practical implication is that sparse observations can provide useful local constraints, but their benefit for radar-like fields depends on the training loss, uncertainty representation, and how observation support is encoded in the model.
Land-atmosphere coupling is an important process for correctly modelling near-surface temperature profiles, but it involves various uncertainties due to subgrid-scale processes, such as turbulent fluxes or unresolved surface heterogeneities, suggesting a probabilistic modelling approach. We develop a copula Bayesian network (CBN) to interpolate temperature profiles, acting as alternative to T2m-diagnostics used in numerical weather prediction (NWP) systems. The new CBN results in (1) a reduction of the warm bias inherent to NWP predictions of wintertime stable boundary layers allowing cold temperature extremes to be better represented, and (2) consideration of uncertainty associated with subgrid-scale spatial variability. The use of CBNs combines the advantages of uncertainty propagation inherent to Bayesian networks with the ability to model complex dependence structures between random variables through copulas. By combining insights from copula modelling and information entropy, criteria for the applicability of CBNs in the further development of parameterizations in NWP models are derived.
Abstract As scientific instruments and the literature generate ever larger volumes of data, machine learning (ML) has become essential for organizing, analyzing and interpreting complex information. This Perspective examines how ML accelerates discovery across disciplines, with examples such as brain mapping and exoplanet detection. It also considers situations with different levels of prior knowledge about the underlying phenomenon, outlining strategies to address limitations and exploit ML effectively. Although growing reliance on ML raises challenges for research practice and validation, it is reshaping scientific methods and expanding what can be studied. We also highlight foundation models as a promising route to faster, broader scientific discovery.
