Search Results

You are looking at 1 - 4 of 4 items for :

  • Author or Editor: Paul M. James x
  • Journal of Applied Meteorology and Climatology x
  • All content x
Clear All Modify Search
James E. Peak and Paul M. Tag

Abstract

A significant task in the automated interpretation of cloud features on satellite imagery is the segmentation of the image into separate cloud features to be identified. A new technique, hierarchical threshold segmentation (HTS), is presented. In HTS, region boundaries are defined over a range of gray-shade thresholds. The hierarchy of the spatial relationships between collocated regions from different thresholds is represented in tree form. This tree is pruned, using a neural network, such that the regions of appropriate sizes and shapes are isolated. These various regions from the pruned tree are then collected to form the final segmentation of the entire image.

In segmentation testing using Geostationary Operational Environmental Satellite data, HTS selected 94% of 101 dependent sample pruning points correctly, and 93% of 105 independent sample pruning points. Using Advanced Very High Resolution Radiometer data, HTS correctly selected 90% of both the 235-case dependent sample and the 253-case independent sample pruning points.

The strength of this approach is that artificial intelligence, that is, reasoning about the sizes and shapes of the emergent regions, is applied during the segmentation process. The neural network component can be trained to respond more favorably to shapes of interest to a particular analysis problem.

Full access
Paul M. Tag and James E. Peak

Abstract

In recent years, the field of artificial intelligence has contributed significantly to the science of meteorology, most notably in the now familiar form of expert systems. Expert systems have focused on rules or heuristics by establishing, in computer code, the reasoning process of a weather forecaster predicting, for example, thunderstorms or fog. In addition to the years of effort that goes into developing such a knowledge base is the time-consuming task of extracting such knowledge and experience from experts. In this paper, the induction of rules directly from meteorological data is explored-a process called machine learning. A commercial machine learning program called C4.5, is applied to a meteorological problem, forecasting maritime fog, for which a reliable expert system has been previously developed. Two detasets are used: 1) weather ship observations originally used for testing and evaluating the expert system, and 2) buoy measurements taken off the coast of California. For both datasets, the rules produced by C4.5 are reasonable and make physical sense, thus demonstrating that an objective induction approach can reveal physical processes directly from data. For the ship database, the machine-generated rules are not as accurate as those from the expert system but are still significantly better than persistence forecasts. For the buoy data, the forecast accuracies are very high, but only slightly superior to persistence. The results indicate that the machine learning approach is a viable tool for developing meteorological expertise, but only when applied to reliable data with sufficient cases of known outcome. In those instances when such databases are available, the use of machine learning can provide useful insight that otherwise might take considerable human analysis to produce.

Full access
Michelle M. Gierach, Mark A. Bourassa, Philip Cunningham, James J. O’Brien, and Paul D. Reasor

Abstract

Ocean wind vectors from the SeaWinds scatterometer aboard the Quick Scatterometer (QuikSCAT) satellite and Geostationary Operational Environmental Satellite (GOES) imagery are used to develop an objective technique that can detect and monitor tropical disturbances associated with the early stages of tropical cyclogenesis in the Atlantic basin. The technique is based on identification of surface vorticity and wind speed signatures that exceed certain threshold magnitudes, with vorticity averaged over an appropriate spatial scale. The threshold values applied herein are determined from the precursors of 15 tropical cyclones during the 1999–2004 Atlantic Ocean hurricane seasons using research-quality QuikSCAT data. The choice of these thresholds is complicated by the lack of suitable validation data. The combination of GOES and QuikSCAT data is used to track the tropical disturbances that are precursors to the 15 tropical cyclones. This combination of data can be used to test detection but is not as easily used to examine false alarms. Tropical disturbances are found for these cases within a range of 19–101 h before classification as tropical cyclones by the National Hurricane Center. The 15 cases are further subdivided based upon their origination source (i.e., easterly wave, upper-level cutoff low, stagnant frontal zone, etc.). The primary focus centers on the cases associated with tropical waves, because these waves account for the majority of all Atlantic tropical cyclones. The detection technique illustrates the ability to track these tropical disturbances from near the coast of Africa. Analysis of the pretropical cyclone (pre-TC) tracks for these cases depicts stages, related to wind speed and precipitation, in the evolution of a tropical disturbance within an easterly wave to a tropical cyclone.

Full access
Wayne M. Angevine, Christoph J. Senff, Allen B. White, Eric J. Williams, James Koermer, Samuel T. K. Miller, Robert Talbot, Paul E. Johnston, Stuart A. McKeen, and Tom Downs

Abstract

Air pollution episodes in northern New England often are caused by transport of pollutants over water. Two such episodes in the summer of 2002 are examined (22–23 July and 11–14 August). In both cases, the pollutants that affected coastal New Hampshire and coastal southwest Maine were transported over coastal waters in stable layers at the surface. These layers were at least intermittently turbulent but retained their chemical constituents. The lack of deposition or deep vertical mixing on the overwater trajectories allowed pollutant concentrations to remain strong. The polluted plumes came directly from the Boston, Massachusetts, area. In the 22–23 July case, the trajectories were relatively straight and dominated by synoptic-scale effects, transporting pollution to the Maine coast. On 11–14 August, sea breezes brought polluted air from the coastal waters inland into New Hampshire.

Full access