Joseph T. Schaefer

An Augmented Tornado Climatology

Abstract Careful screening of the National Severe Storms Forecast Centers tornado log eliminated almost 20% of the reports as doubtful, leaving 17 659 tornadoes during 27 years, 1950–76 (654 annually). Newspaper accounts and other local information provided intensifies (Fujita wale) for all but 2346 tornadoes and path lengths for all but 2011 tornadoes. There were 14 409 tornadoes for which both intensity and path length estimates were made. Of these, 61.7% were weak (<112 mph), 36% strong (113–206 mph), and only 2.3% violent (207–318 mph). However, the 340 violent tornadoes caused 68% of the 3070 fatalities attributed to tornadoes for which force estimates could he made (113.7 annually). Most violent tornadoes came in swarms except in southeastern United States, where no day had mart than one. Some 61% of the violent tornadoes had intermediate paths (3.2–31 mi), while 73% of weak and strong tornadoes had short paths. Violent tornadoes occurred at all times of day and night, while weak and strong tornado...

Climatology of Nontornadic Severe Thunderstorm Events in the United States

Abstract While the climatology of excessive rain and tornadoes is well-documented, little is known of storms that produce high winds or large hail. The characteristics of the approximately 75 000 severe thunderstorms which occurred in the United States from 1955 through 1983 are analyzed in an attempt to rectify this situation. The distribution of over 29 000 storms causing hail larger than 19 mm shows marked diurnal, seasonal, and geographic preferences. These storms occur most frequently during the midafternoon hours of May and June in a zone running from central Texas to Nebraska. Spring storms tend to occur south of the Kansas-Nebraska border and summer storms north of it. Thunderstorm winds which produce either “structural” damage or are reported as faster than 25.8 m s−1 generated about 46 000 reports. These storms typically occur during midafternoon in June and July. While the geographic distribution of violent windstorms is similar to that hailstorms, a zone of weaker severe thunderstorm gusts lie...

A Minimum Assumption Tornado-Hazard Probability Model

Abstract One of the principle applications of climatological tornado data is in tornado-hazard assessment. To perform such a hazard-potential determination, historical tornado characteristics in either a regional or tom area are complied. A model is then used to determine a site-specific point probability of a tornado greater than a specified intensity occurring. Various models require different climatological input. However, a knowledge of the mean values of tornado track width, tornado track width, tornado affected area and tornado occurrence rate as both a function of tornado intensity and geographic area, along with a violence frequency distribution, enable Mod of the models to be applied. The NSSFC-NRC tornado data base is used to supply input for the determination of these parameters over the United States. This climatic data base has undergone extensive updating and quality control since it was last reported. For track parameters, internally redundant data were used to cheek consistency. Further, r...

On the Interpolation of a Vector Field

Abstract By its very nature, interpolation in a vector field is ambiguous, owing to the somewhat arbitrary nature of the vector norm. Since a two-dimensional vector field cm be specified by two scalar quantities. which can be separately interpolated, the ambiguity can be resolved by forcing the interpolated wind field to preserve the vorticity and divergence fields associated with the raw data. A method to calculate divergence and vorticity directly from randomly spaced wind observations is developed and, using analytically generated data, shown to produce more accurate results than conventional computations. Two methods of retrieving the wind field from the analysed scalar fields are presented and also tested on the analytic field. Finally, total analysis, from wind observations to gridded wind fields, is demonstrated on real meteorological data.

The Correlation between U.S. Tornadoes and Pacific Sea Surface Temperatures

Abstract The correlation between tornadic activity in several regions of the United States and the monthly mean sea surface temperature over four zones in the tropical Pacific Ocean is examined. Tornadic activity is gauged with two mostly independent measures: the number of tornadoes per month, and the number of tornadic days per month. Within the assumptions set forth for the analysis, it is found that there appears to exist a statistically significant but very weak correlation between sea surface temperature in the Pacific Ocean and tornadic activity in the United States, with the strength and significance of the correlation depending on the coordinates at which the sea surface temperatures are assessed and the geographic region of the United States. The strongest evidence found is for the correlation between the number of days with strong and violent (F2 and greater) tornadoes in an area that runs from Illinois to the Atlantic Coast, and Kentucky to Canada and a cool sea surface temperature in the cent...

A Markov Chain Model of Tornadic Activity

Abstract Tornadic activity in four U.S. regions is stochastically modeled based on data on tornado counts over the years 1953–98. It is shown that tornadic activity on a given day is mostly affected by the activity on the previous day. Hence, the process can be modeled as a Markov chain. A parametric nonhomogenous Markov chain model is developed based on the well-known increase of tornadic activity in the spring and summer months. This model, with only eight parameters, describes tornadic activity quite well. The interpretability of the estimated parameters allows a diagnosis of the regional differences in tornadic activity. For instance, a comparison of the values of the parameters for the four regions suggests that in the South tornado persistence is specific mostly to the early part of the year. Finally, within the framework of probabilistic forecast verification, it is shown that the Markov chain model outperforms the climatological model, even though the former is far simpler in terms of the number o...

Tornado 1982: A Near-Record Year

Abstract The tornado events of 1982 are reviewed. Significant and interesting aspects of the 1047 reported storms are noted. The synoptic patterns associated with four major tornado days are examined.

The Inherent Position Errors in Double-Theodolite Pibal Measurements

Abstract A simple analysis of the position error inherent in double-theodolite pibal systems is presented. The quality of data collected by double theodolites is very sensitive to the geometric design of the system, and care must be taken in the interpretation of results.

On the Solution of the Generalized Ekman Equation

Abstract The generalized Ekman equation is often used for micrometeorological applications. Use of the noniterative numerical technique of superposition reveals that the vertical variation of both the eddy diffusivity and the thermal wind is important to the determination of the wind profile when this equation is applicable.

The Comparative Environment of Beijing and Topeka during the Thunderstorm Season

Abstract Beijing, Peoples Republic of China, and Topeka, Kansas, United States of America, are located at approximately the same latitude and are affected by similar synoptic weather patterns. However, their thunderstorm climatology differs significantly. Rowinsonde data from the two stations are compared. It is found that the typical distribution of the oceanic subtropical high pressure areas subtly modifies the synoptic scale environment accounting for the observed differences.