Search Results

You are looking at 11 - 20 of 21 items for

  • Author or Editor: Paul M. Tag x
  • Refine by Access: All Content x
Clear All Modify Search
Paul M. Tag and Steven W. Payne

Abstract

Cloud top entrainment instability, as a mechanism for the breakup of marine stratus, is examined with a three-dimensional, planetary boundary layer (PBL) model. Specifically, we examine the criterion developed by Randall and Deardorff; this criterion states that stratus will break up if the equivalent potential temperature gradient at cloud top becomes less than a critical value. To examine this hypothesis, we simulate a horizontally uniform stratus layer which is excited from above by small random temperature perturbations. The buoyancy instability ratio (BIR), defined as Δθe(Δθe)crit and computed at cloud top, is calculated locally across the domain and also averaged to define a mean value. Six cases, involving different wind speeds and above-cloud soundings, produce different initial BIRs and different breakup sequences. In general, we find that a mean BIR greater that one is a necessary condition for stratus breakup; however, we also find that the timing of breakup following achievement of the critical ratio is different from run to run. The low wind speed cases, initially most stable at cloud top, are the first to break up, while the higher wind speed (most unstable) cases require longer time to break up. We conclude that an additional mechanism is necessary to stimulate vertical motion in order to take advantage of the cloud-top entrainment instability. In our simulations, that additional stimulation comes from vertical motions generated by Rayleigh-type instability in the PBL.

Full access
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
Paul M. Tag and Thomas E. Rosmound

Abstract

Accuracy and energy conservation are examined in a three-dimensional (3D) anelastic model. For both dry and moist (noncondensing) atmospheres, we prescribe analytic solutions of momentum, potential temperature and mixing ratio for both periodic and closed boundaries. Accuracy is assessed by comparing amplitudes and phase speeds from both the numerical and analytic solutions. Kinetic and potential energies and enthalpy (including air, vapor, liquid and latent) are calculated for both the mean

and perturbation states. To assess the energetics involved in phase changes, we examine a separate cloud simulation. Two-dimensional (2D) and hydrostatic experiments are also conducted using the cloud simulation.

For the linear analytic wave solutions, phase speeds as a function of time step for our semi-implicit model are compared to both implicit and explicit linear stability generated speeds. We show that an explicit scheme enhances the phase speed up to the CFL cutoff while an implicit scheme retards the phase speed. For the quasi-Lagrangian method of moisture advection, we find that a water conservation algorithm is necessary to maintain conservation of total perturbation energy. Similarly. the correct inclusion of moisture in the computation of density is most critical to energy conservation. In comparing a 2D forced cloud to the 3D simulation, only 17% of the perturbation energy which changes form in the 3D case does so in the 2D experiment-in direct relation to the larger cloud in the 3D simulation. And finally, comparing experiments both with and without the hydrostatic assumption, we verify earlier 2D findings that the magnitude of the vertical motion is larger in a hydrostatic model.

Full access
Richard L. Bankert and Paul M. Tag

Abstract

An automated method to estimate tropical cyclone intensity using Special Sensor Microwave Imager (SSM/I) data is developed and tested. SSM/I images (512 km × 512 km) centered on a given tropical cyclone (TC), with a known best-track intensity, are collected for 142 different TCs (1988–98) from the North Pacific, Atlantic, and Indian Oceans. Over 100 characteristic features are computed from the 85-GHz (H-pol) imagery data and the derived rain-rate imagery data associated with each TC. Of the 1040 sample images, 942 are selected as training samples. These training samples are examined in a feature-selection algorithm to select an optimal subset of the characteristic features that could accurately estimate TC intensity on unknown samples in a K-nearest-neighbor (K-NN) algorithm. Using the 15 selected features as the representative vector and the best-track intensity as the ground truth, the 98 testing samples (taken from four TCs) are presented to the K-NN algorithm. A root-mean-square error (rmse) of 19.8 kt is produced. This “snapshot” approach is enhanced (rmse is 18.1 kt) when a TC intensity history feature is added to 71 of the 98 samples. Reconnaissance data are available for two recent (1999) Atlantic hurricanes, and a comparison is made in the rmse using those data as ground truth versus best track. For these two TCs (17 SSM/I images), an rmse of 15.6 kt is produced when best track is used and an rmse of 19.7 kt is produced when reconnaissance data are used as the ground truth.

Full access
Rolf H. Langland, Paul M. Tag, and Robert W. Fett

Abstract

Satellite imagery from 18 April 1978 suggests the presence of a semicircular zone of calm or new-calm seas in Monterey Bay, California. It is hypothesized that sea breeze circulations account for the calm zone in the bay, although a lack of in situ surface and upper air observations prevents direct verification of this theory. A three-dimensional numerical model of the marine atmospheric boundary layer is used to simulate the development of the low-level wind field on the day in question, under sea breeze conditions. The model produces a zone of winds speeds under 1.0 m s−1 over the center of the bay, near the time of the satellite image. These model results suggest that a sea breeze circulation may have accounted for the zone of very light winds and calm sea.

Full access
Francis W. Murray, L. Randall Koenig, and Paul M. Tag

Abstract

A two-dimensional field-of-flow numerical model of cloud development is used to study a cloud that formed over a refinery as a result of heat dissipated to the atmosphere. The observed vertical structure of the atmosphere provided initial conditions. The wind necessarily was simplified to a unidirectional flow. The cloud-initiating perturbation consisted of sensible and latent heat equal to the waste heat rejected to the atmosphere by the refinery.

When conditions were matched to those reported, the simulated cloud agreed in most particulars with the observations. Sensitivity tests showed that the simulated cloud depends too strongly on ambient wind speed and shear. This perhaps is a generic defect of two-dimensional formulations. The response of the simulated cloud to expected changes in heat flux density appears more realistic than its response to small changes in ambient wind.

The cloud evolution consists of bubbles forming and breaking away from the main cloud mass, then moving downwind and dissipating. This behavior characterizes real clouds associated with a stationary heat source. The simulations also predict that under appropriate conditions secondary clouds form far downwind.

Full access
Paul M. Tag, Richard L. Bankert, and L. Robin Brody

Abstract

Using imagery from NOAA’s Advanced Very High Resolution Radiometer (AVHRR) orbiting sensor, one of the authors (RLB) earlier developed a probabilistic neural network cloud classifier valid over the world’s maritime regions. Since then, the authors have created a database of nearly 8000 16 × 16 pixel cloud samples (from 13 Northern Hemispheric land regions) independently classified by three experts. From these samples, 1605 were of sufficient quality to represent 11 conventional cloud types (including clear). This database serves as the training and testing samples for developing a classifier valid over land. Approximately 200 features, calculated from a visible and an infrared channel, form the basis for the computer vision analysis. Using a 1–nearest neighbor classifier, meshed with a feature selection method using backward sequential selection, the authors select the fewest features that maximize classification accuracy. In a leave-one-out test, overall classification accuracies range from 86% to 78% for the water and land classifiers, with accuracies at 88% or greater for general height-dependent groupings. Details of the databases, feature selection method, and classifiers, as well as example simulations, are presented.

Full access
Paul M. Tag, Francis W. Murray, and L. Randall Koenig

Abstract

A two-dimensional model is used to test the effects of using several forms of eddy viscosity parameterization to simulate subgrid-scale turbulence. A well-documented observation of a quasi-steady cumulus cloud which formed over a refinery is used for simulation and comparison. The control parameterization of eddy viscosity is one based on both the deformation and buoyancy fields. When compared to observations, this control run overestimates somewhat the liquid water contents and slightly underpredicts the vertical velocities. Parameterizations based on deformation alone, two-dimensional turbulence theory, and several constant values of eddy viscosity result in cloud simulations that are deficient primarily in their significant overprediction of liquid water content. These experiments confirm that a buoyancy term in the prescription of eddy viscosity is necessary when thermal instability plays an active role in the subgrid forcing.

Full access
Edward E. Hindman II, Paul M. Tag, Bernard A. Silverman, and Peter V. Hobbs

Abstract

The paper mill at Port Townsend, Wash., is a source of large and giant condensation nuclei (CCN). These CCN cause the concentrations of droplets ≥30 μm in diameter to be higher in small, nonraining warm clouds located in the plume of the mill than in similar clouds unaffected by the plume. Calculations based on a model for nonsheared, warm cumulus clouds and a model for warm stratus clouds indicate that the higher concentrations of large droplets in the clouds in the plume should not cause any significant changes in the rainfall from these clouds. These results indicate that the large and giant CCN emitted by the mill are not by themselves responsible for the increased rainfall measured in the vicinity of the mill. The heat and moisture emitted by the mill, in combination with the CCN, may have been responsible for the increased rainfall.

Full access