The quiet revolution of numerical weather prediction

By Peter Bauer, A. Thorpe, Gilbert Brunet. Nature, 2015

2,657 citations recorded by the source corpus.

Subfields: Atmospheric Science, Global and Planetary Change

View Paper: https://www.semanticscholar.org/paper/35e4b2d62ff35ade25f89c09e22a65b1279756e9

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Atmospheric Science

In Global and Planetary Change

Citations in the corpus

Earth System Modeling 2.0: A Blueprint for Models That Learn From Observations and Targeted High‐Resolution Simulations (2017, methodology): “Just as weather forecasts have made great strides over the past decades thanks to improvements in the assimilation of observations [Bauer et al., 2015], climate projections can advance similarly by harnessing observations and modern computational capabilities more systematically.”
The North Atlantic Waveguide and Downstream Impact Experiment (2018, background): “58 Additionally, numerical weather prediction (NWP) has undergone a revolution in recent 59 years, with the development and widespread use of ensemble prediction systems (EPS) to 60 represent forecast uncertainty (Bauer et al. 2015).”
A review of machine learning applications in wildfire science and management (2020, background): “Additionally, improvements in numerical weather prediction and climate models are simultaneously offering smaller spatial resolutions and longer lead forecast times [Bauer et al., 2015] which potentially offer improved predictability of extreme fire weather events.”
The impact of Aeolus wind retrievals on ECMWF global weather forecasts (2021, result): “The improvements described in this article compare favourably with historical improvements from introducing new observation types, see Bauer et al. (2015).”
Distributional regression forests for probabilistic precipitation forecasting in complex terrain (2018, background): “While this approach alone is already highly effective in the plains, it typically does not perform as well in complex terrain due to unresolved effects in the NWP system (Bauer et al. 2015).”
GraphCast: Learning skillful medium-range global weather forecasting (2022, methodology): “NWP methods scale well with compute: accuracy generally increases with increased compute resources, because additional compute can be allocated to higher-resolution simulations, with more sophisticated parametrization schemes (Bauer et al., 2015; Benjamin et al., 2019).”
FourCastNet: Accelerating Global High-Resolution Weather Forecasting Using Adaptive Fourier Neural Operators (2022, background): “The accuracy of these models has been steadily improving due to increasing resolution, better characterization of physical processes, and improvements in atmospheric observations leading to better model initialization via data assimilation [1].”
WeatherBench 2: A Benchmark for the Next Generation of Data‐Driven Global Weather Models (2023, background): “Global, medium‐range (1–14 days) numerical weather prediction (NWP) is a key component of modern weather forecasting (Bauer et al., 2015).”
Can deep learning beat numerical weather prediction? (2021, background): “While there has been steady progress in the development of NWP (cf. [ 2]), the history of ML (cf. [3 ]) has been more disruptive: the ﬁrst neural network (NN) was proposed in 1943 by McCulloch & Pitts [4].”
Potential applications of subseasonal‐to‐seasonal (S2S) predictions (2017, background): “Accurate climate prediction requires a good representation of weather phenomena as well as the underlying physical laws that apply to all prediction timescales (Bauer et al., 2015).”
Machine learning–accelerated computational fluid dynamics (2021, neutral): “To put these results in context, if applied to numerical weather prediction, increasing the duration of accurate predictions from 4 to 7 time units would correspond to approximately 30 y of progress (56).”