Search Results
You are looking at 1 - 2 of 2 items for :
- Author or Editor: David Hall x
- Journal of Applied Meteorology and Climatology x
- Refine by Access: All Content x
Abstract
Extracting valuable information from large sets of diverse meteorological data is a time-intensive process. Machine-learning methods can help to improve both speed and accuracy of this process. Specifically, deep-learning image-segmentation models using the U-Net structure perform faster and can identify areas that are missed by more restrictive approaches, such as expert hand-labeling and a priori heuristic methods. This paper discusses four different state-of-the-art U-Net models designed for detection of tropical and extratropical cyclone regions of interest (ROI) from two separate input sources: total precipitable water output from the Global Forecast System (GFS) model and water vapor radiance images from the Geostationary Operational Environmental Satellite (GOES). These models are referred to as International Best Track Archive for Climate Stewardship (IBTrACS)-GFS, Heuristic-GFS, IBTrACS-GOES, and Heuristic-GOES. All four U-Nets are fast information extraction tools and perform with an ROI detection accuracy ranging from 80% to 99%. These are additionally evaluated with the Dice and Tversky intersection-over-union (IoU) metrics, having Dice coefficient scores ranging from 0.51 to 0.76 and Tversky coefficients ranging from 0.56 to 0.74. The extratropical cyclone U-Net model performed 3 times as fast as the comparable heuristic model used to detect the same ROI. The U-Nets were specifically selected for their capabilities in detecting cyclone ROI beyond the scope of the training labels. These machine-learning models identified more ambiguous and active ROI missed by the heuristic model and hand-labeling methods that are commonly used in generating real-time weather alerts, having a potentially direct impact on public safety.
Abstract
Extracting valuable information from large sets of diverse meteorological data is a time-intensive process. Machine-learning methods can help to improve both speed and accuracy of this process. Specifically, deep-learning image-segmentation models using the U-Net structure perform faster and can identify areas that are missed by more restrictive approaches, such as expert hand-labeling and a priori heuristic methods. This paper discusses four different state-of-the-art U-Net models designed for detection of tropical and extratropical cyclone regions of interest (ROI) from two separate input sources: total precipitable water output from the Global Forecast System (GFS) model and water vapor radiance images from the Geostationary Operational Environmental Satellite (GOES). These models are referred to as International Best Track Archive for Climate Stewardship (IBTrACS)-GFS, Heuristic-GFS, IBTrACS-GOES, and Heuristic-GOES. All four U-Nets are fast information extraction tools and perform with an ROI detection accuracy ranging from 80% to 99%. These are additionally evaluated with the Dice and Tversky intersection-over-union (IoU) metrics, having Dice coefficient scores ranging from 0.51 to 0.76 and Tversky coefficients ranging from 0.56 to 0.74. The extratropical cyclone U-Net model performed 3 times as fast as the comparable heuristic model used to detect the same ROI. The U-Nets were specifically selected for their capabilities in detecting cyclone ROI beyond the scope of the training labels. These machine-learning models identified more ambiguous and active ROI missed by the heuristic model and hand-labeling methods that are commonly used in generating real-time weather alerts, having a potentially direct impact on public safety.
Abstract
The Operational Multiscale Environment Model with Grid Adaptivity (OMEGA) is a multiscale nonhydrostatic atmospheric simulation system based on an adaptive unstructured grid. The basic philosophy behind the OMEGA development has been the creation of an operational tool for real-time aerosol and gas hazard prediction. The model development has been guided by two basic design considerations in order to meet the operational requirements: 1) the application of an unstructured dynamically adaptive mesh numerical technique to atmospheric simulation, and 2) the use of embedded atmospheric dispersion algorithms. An important step in proving the utility and accuracy of OMEGA is the full-scale testing of the model using simulations of real-world atmospheric events and qualitative as well as quantitative comparisons of the model results with observations. The main objective of this paper is to provide a comprehensive evaluation of OMEGA against a major dispersion experiment in operational mode. Therefore, OMEGA was run to create a 72-h forecast for the first release period (23–26 October 1994) of the European Tracer Experiment (ETEX). The predicted meteorological and dispersion fields were then evaluated against both the atmospheric observations and the ETEX dispersion measurements up to 60 h after the start of the release. In general, the evaluation showed that the OMEGA dispersion results were in good agreement in the position, shape, and extent of the tracer cloud. However, the model prediction indicated that there was a limited spreading of the predictions around the measurements with a small tendency to underestimate the concentration values.
Abstract
The Operational Multiscale Environment Model with Grid Adaptivity (OMEGA) is a multiscale nonhydrostatic atmospheric simulation system based on an adaptive unstructured grid. The basic philosophy behind the OMEGA development has been the creation of an operational tool for real-time aerosol and gas hazard prediction. The model development has been guided by two basic design considerations in order to meet the operational requirements: 1) the application of an unstructured dynamically adaptive mesh numerical technique to atmospheric simulation, and 2) the use of embedded atmospheric dispersion algorithms. An important step in proving the utility and accuracy of OMEGA is the full-scale testing of the model using simulations of real-world atmospheric events and qualitative as well as quantitative comparisons of the model results with observations. The main objective of this paper is to provide a comprehensive evaluation of OMEGA against a major dispersion experiment in operational mode. Therefore, OMEGA was run to create a 72-h forecast for the first release period (23–26 October 1994) of the European Tracer Experiment (ETEX). The predicted meteorological and dispersion fields were then evaluated against both the atmospheric observations and the ETEX dispersion measurements up to 60 h after the start of the release. In general, the evaluation showed that the OMEGA dispersion results were in good agreement in the position, shape, and extent of the tracer cloud. However, the model prediction indicated that there was a limited spreading of the predictions around the measurements with a small tendency to underestimate the concentration values.
