Search Results

You are looking at 31 - 36 of 36 items for

  • Author or Editor: Peter S. Ray x
  • Refine by Access: All Content x
Clear All Modify Search
Peter S. Ray
,
R. J. Doviak
,
G. B. Walker
,
D. Sirmans
,
J. Carter
, and
B. Bumgarner

Abstract

On 20 April 1974 a tornadic storm passed between the two NSSL Doppler radars spaced about 42 km apart. Both radars simultaneously collected Doppler data throughout the storm. Air motions synthesized from these data provide the first three-dimensional display of Doppler-derived wind fields in a tornadic storm. Cyclonic circulation, associated with the tornado, and regions of intense up- and down-drafts are clearly evident.

Full access
Richard J. Doviak
,
Peter S. Ray
,
Richard G. Strauch
, and
L. Jay Miller

Abstract

Variance in horizontal and vertical winds are predicted when these components are computed from dual-Doppler velocity measurements combined with terminal velocity estimates and the continuity equation. Errors in horizontal wind magnitude and direction are shown to be functions of wind direction and speed as well as spatial location. Vertical wind could be estimated with errors less than a few meters per second up to altitudes near 14 km over a region 4d × 4d, where 2d is the radar separation. Vertical wind variance at high altitudes is related to accumulation of errors due to the integration of the continuity equation. The cause of wind variance is assumed to be uncertainty in mean Doppler velocity estimates produced by spectrum broadening mechanisms (e.g., shear, turbulence). Two interpolation methods, used to estimate Doppler velocity at common grid locations, are compared and their contribution to Doppler velocity variance reduction is calculated. Terminal velocity variance has been related to uncertainties in drop-size distributions and reflectivity estimate variance. The methods derived herein are applied to determine the errors in wind speeds calculated from dual-Doppler data.

Full access
Kenneth W. Johnson
,
Peter S. Ray
,
Brenda C. Johnson
, and
Robert P. Davies-Jones

Abstract

Observations of the 20 May 1977 tornadic storms are used to evaluate recent theories on the initiation of rotation at mid-and low levels and to verify recent thermodynamic retrieval results. Using the lengthy data record from a variety of sensors available for this day, it appears that the mechanism that initiates low-level rotation is different from that at midlevels. Attempts to identify the source of the low-level rotation as vertical tilting baroclinically generated horizontal vorticity were inconclusive.

The recent thermodynamic retrieval results of Hane and Ray and of Brandes for these storms are in good agreement with independent measurements where available. However, verification is hindered by the sparseness of these measurements. Noticeable differences in the region of the rear-flank downdraft suggest that there is room for improvement in the retrieval methods.

Investigation of the cyclic generation of rotation along gust fronts indicates that the source of low-level rotation is not derived from baroclinically generated horizontal vorticity as seems to be the case with the initial mesocyclone core. Instead, vertical vorticity amplification along the gust front leading to successive generation of mesocyclone cores and discrete mesocyclone propagation is the result of the concentration of low-level preexisting vertical vorticity through convergence.

Full access
Steven V. Vasiloff
,
Edward A. Brandes
,
Robert P. Davies-Jones
, and
Peter S. Ray

Abstract

Nearly 2½ hours of dual-Doppler radar data with high temporal and spatial resolution are used to examine the evolution and morphology of a thunderstorm that evolved from a complex of small cells into a supercell storm. Individual storm cells and updrafts moved east-northeastward, nearly with the mean wind, while the storm complex, which encompassed the individual cells, propagated toward the south–southeast. Cells were first detected at middle levels (5–10 km) on the storm's right flank and dissipated on the left flank. Generally, the storm contained three cells—a forming cell, a mature cell, and a dissipating cell; life stages were apparently dictated by the source of updraft air. During the growth stage, cell inflow had a southerly component. As the cell moved through the storm complex, it started ingesting stable air from the north and soon dissipated.

A storm-environment feedback mechanism of updraft–downdraft interactions, in conjunction with increasing environmental vertical wind shear and buoyancy, is deemed responsible for an increase in the size and intensity of successive cells and updrafts. With time, a large region of background updraft, containing the updrafts of individual cells, formed on the storm's right flank. Unlike the individual cells, which moved nearly parallel to the mean wind and low-level shear vector, the region of background updraft moved to the right of the mean wind and low-level shear vector. It is believed that the formation and rightward motion of the background updraft region led to strong rotation on the storm's right flank. The larger cell and updraft size, with the same center-to-center spacing as at earlier times, made individual cell identification difficult, resulting in a nearly steady-state reflectivity structure.

The data support a growing consensus that a continuum of storm types, rather than a dichotomy, exists.

Full access
Peter S. Ray
,
Conrad L. Ziegler
,
William Bumgarner
, and
Robert J. Serafin

Abstract

The use of one, two, three or more Doppler radars has become increasingly common in research programs. The advantage in increasing the number of radars is in the increased area covered and the accuracy with which wind estimates may be obtained. Although multiple-radar systems can yield special quantitative insight, a great deal of information can still be determined in real time from a single radar. It should be noted that the interpretation of radial velocity estimates from a single radar are not always unambiguous. Color displays of single-Doppler radial velocity patterns aid in the real-time interpretation of the associated reflectivity fields and can reveal important features not evident in the reflectivity structures alone. Such a capability is of particular interest in the identification and study of severe storms. A display utilizing a 5 cm Doppler radar is used to illustrate the patterns seen from several tornadic storms that occurred in central Oklahoma on 20 May 1977. Interpretation of some complicated or ambiguous features is aided by including data from additional radars. Further explanations on such structure are given from an analysis based on a new dual-Doppler analysis technique for one of 16 tornadic storms that occurred on 20 May 1977.

Several alternative analysis schemes for two to four Doppler radars are also demonstrated and compared. These illustrate the major differences found in error propagation, use of information, and in difference quantities, such as divergence. It is shown that an analysis that specifies boundary values for w is not strongly dependent on the number of radars.

Full access
Peter S. Ray
,
K. K. Wagner
,
K. W. Johnson
,
J. J. Stephens
,
W. C. Bumgarner
, and
E. A. Mueller

Abstract

The advantages of three or more radars is illustrated in terms of extended areal coverage, error variance reduction and mitigation of poor sampling effects of divergence fields near the ground. Errors in the horizontal and vertical particle motion fields are illustrated for a three-, four- and five-radar network, and the importance of experimental design is illustrated. A variationally formulated multiple-Doppler radar analysis, which includes data from a surface network, and an illustrative analysis of storm data are presented.

Full access