Search Results

You are looking at 1 - 10 of 111 items for :

  • Deep learning x
  • Weather and Forecasting x
  • All content x
Clear All
John L. Cintineo, Michael J. Pavolonis, Justin M. Sieglaff, Anthony Wimmers, Jason Brunner, and Willard Bellon

thunderstorms into a single output, we utilize a deep-learning approach that mimics expert human pattern recognition of intense convection in satellite imagery. The goal of this approach is to quantify convective intensity automatically, saving forecasters time in identifying, diagnosing, and prioritizing threats. Deep learning is a branch of machine-learning methods based on artificial neural networks with feature learning, or the ability to automatically find salient features in data (e.g., Schmidhuber

Restricted access
Jonny Mooneyham, Sean C. Crosby, Nirnimesh Kumar, and Brian Hutchinson

approach presented here. Here, we develop a deep learning model that we call Spectral Wave Residual Learning Network (SWRL Net) to improve numerical model predictions with directional wave buoy observations. Spectral wave predictions at buoy locations are used with collocated directional buoy observations to generate forecast corrections up to 24 h in the future. Frequency-directional spectra are transformed into the observed buoy moments resulting in a large feature set and large number of model

Open access
Ryan Lagerquist, Amy McGovern, and David John Gagne II

induced by wind shear and convergence between two anticyclones), the thermal method is better at detecting warm fronts (which are almost never detected by the wind-shift method). Machine learning (ML) is a process whereby computers learn autonomously from data, as opposed to an expert system like NFA, which is based on human-derived rules. Deep learning (DL) is a subset of ML, which offers the ability to encode the input data at various levels of abstraction. These abstractions are called features

Full access
Florian Dupuy, Olivier Mestre, Mathieu Serrurier, Valentin Kivachuk Burdá, Michaël Zamo, Naty Citlali Cabrera-Gutiérrez, Mohamed Chafik Bakkay, Jean-Christophe Jouhaud, Maud-Alix Mader, and Guillaume Oller

. The atmospheric research community has already taken advantage of CNN’s ability [see Reichstein et al. (2019) for an overview]. Most of the applications deal with images, for example from satellite observations to create cloud masks or derive rainfalls ( Drönner et al. 2018 ; Moraux et al. 2019 ), or from pictures for weather classification ( Elhoseiny et al. 2015 ). Often, CNNs using NWP data as predictors (predictors are also named features in the deep learning community) are used to produce

Open access
Buo-Fu Chen, Boyo Chen, Hsuan-Tien Lin, and Russell L. Elsberry

various life stages, environments, and basins. In addition, only a few features (usually less than 10) may be finally used in the regression models. This collaborative study between meteorologists and data scientists proposes a deep-learning model to address the need for an automated, objective, and end-to-end intensity estimation technique. Since AlexNet, which established the baseline architecture of convolutional neural networks for image recognition used today, was proposed in 2012 ( Krizhevsky

Full access
Kirkwood A. Cloud, Brian J. Reich, Christopher M. Rozoff, Stefano Alessandrini, William E. Lewis, and Luca Delle Monache

-scale meteorological predictors, along with information describing the atmospheric flow stability and the uncertainty in initial conditions, to predict forecast intensity error in operational prediction schemes. A18 also addressed intensity prediction with the analog ensemble method. More recently, machine learning has gained increasing prominence in postprocessing. Evolutionary programming, simple neural networks, and deep learning have shown significant promise as postprocessing tools (e.g., Gagne et al. 2014

Full access
Tao Zhang, Wuyin Lin, Yanluan Lin, Minghua Zhang, Haiyang Yu, Kathy Cao, and Wei Xue

parameterization schemes, such as convection and cloud schemes. Meanwhile, new discoveries from the machine learning framework can be further verified through physics-based interpretation, having the potential to lead to a deeper understanding of the genesis mechanism of TCs. We plan to extend such applications in separate works. Acknowledgments This work is supported by the CMDV Project to Brookhaven National Laboratory under Contract DE-SC0012704 and Brookhaven National Laboratory’s Laboratory Directed

Full access
Loris Foresti, Ioannis V. Sideris, Daniele Nerini, Lea Beusch, and Urs Germann

historical overview on deep learning, we refer to Schmidhuber (2015) . To our knowledge, the first study that tested the usage of ANNs for precipitation nowcasting is by French et al. (1992) . The authors trained an ANN to predict the evolution of synthetic rainfall fields, but did not find significantly higher skill compared to Lagrangian persistence. Grecu and Krajewski (2000) went a step further by separating the prediction problem into two steps: the estimation of the radar echo motion and the use

Open access
Gary Lackmann

that S2S studies were welcomed. In early 2019, we modified the WAF terms of reference to clarify that such studies are welcomed, with predictive horizons extending out to a few years. To accommodate such submissions, a leading researcher in this area, Prof. Ben Kirtman (University of Miami), joined the editorial board in January 2019. 2) Machine learning (ML), artificial intelligence (AI), and deep learning (DL) have experienced rapid growth in many areas, including in forecasting applications

Open access
Gary Lackmann

that S2S studies were welcomed. In early 2019, we modified the WAF terms of reference to clarify that such studies are welcomed, with predictive horizons extending out to a few years. To accommodate such submissions, a leading researcher in this area, Prof. Ben Kirtman (University of Miami), joined the editorial board in January 2019. 2) Machine learning (ML), artificial intelligence (AI), and deep learning (DL) have experienced rapid growth in many areas, including in forecasting applications

Open access