Abstract

Studies during the Metropolitan Meteorological Experiment (METROMEX) sought to define influences of St. Louis on the summer atmosphere that led to alterations in rainfall. These studies defined how city influences caused an afternoon maximum of rainfall cast of the city. Rain data indicated a second rain maximum northeast of the city during the 2000–2400 CDT period. Study of this nocturnal maximum revealed a 58% localized rain increase, relative to the mean rainfall in the 5200 km2 network. The anomaly was present in all summers from 1971–1975. The northeast rain maximum is preceded by a local increase beginning 2 h earlier and 30 km west over the urban-industrial area. Most northeast anomaly-related storms were found to move either from the southwest (from over the urban area) or from the west-northwest (from a major industrial area), and to produce heavy rainfall rates; 19 storms moved from St. Louis between 2100–2400 and these produced 69% of the rainfall in the maximum rainfall area. The afternoon and nocturnal maximums both occurred when the entire area was receiving relatively heavy rainfall indicating that urban influences are most effective during relatively heavy rainfall conditions. All of the nocturnal anomaly rainfall occurred with well-organized convective systems. The individual convective raincells which led to heavy rainfall in the anomaly typically began over the urban industrial area and ended in the anomaly area. The raincell areas, volumes, and intensifies were much greater than rural raincells. Collectively, the results strongly suggest that the nocturnal anomaly is a result of urban influence that affect a few of the heavier rain events.

Abstract

Experiments designed to evaluate the potential of infrared (IR) color and standard color aerial photographs to measure crop-hail losses were performed. Detailed post-storm field measurements of loss (by standard adjusting techniques) and actual harvested yields (control data) were used for comparisons. Studies were made from a damaging hailstorm and from simulated hail damage on corn test plots.

Stereoscopic analysis of standard color film provided estimates of average field loss for corn and soybeans that were better than those derived from the “best” field adjusting that involved detailed sampling of 1 point per 5 acres. Badly damaged areas had unique “signature” on the photographs consisting of semi-circular areas of loss that suggest a hail-wind related series of vortices of 100–500 ft in diameter.

Measurements of IR film density for the simulated hail-damaged corn at the 14-leaf and later growth stages also showed a good relationship with the amount of harvested corn loss, but a weaker relationship existed at the 10-leaf stage (at hailstorm time and in the simulated fields), and none was apparent at the 6-leaf stage. The IR film density values for storm-damaged soybeans did not relate well to actual harvested losses. Corn-hail losses ≤30% per field at the 9- to 11-leaf stage were poorly ascertained by standard point assessment, stereoscopic analysis, and the IR film density technique. In this loss range, the IR photographs may have been taken too late after the storm or the physiological damage insufficient to be sensed in the infrared range. On standard color film, the stand reduction was too slight to be detected stereoscopically. Since the standard assessment method provided poor estimates of corn and bean losses when they were in the 1–30% range, either the causes for such low loss are not understood or succeeding weather conditions greatly alter the apparent loss at assessment time. If the latter is true, early season hail losses should be assessed at harvest. The very dense measurements of loss revealed amazing spatial variability in loss per field. Also, these dense measurements predicted final harvested field losses better and provided generally lower estimated losses than those obtained by insurance adjustors with their sampling methods.

Abstract

As a result of a 1980–81 drought, statistically derived outlooks of monthly and seasonal precipitation began to be issued to Illinois officials who were making management decisions relating to water supplies and agricultural activities. Outlooks of above, near or below normal precipitation have subsequently been issued operationally over a 3-year period for four areas of Illinois. They are assessed here as to their skill and major uses. This assessment shows that 56% of the seasonal outlooks were correct as opposed to 33% expected by chance, and 30% were correct when only persistence is used to forecast the coming season. The seasonal outlooks were correct most often in fall (67%) and least often in winter (42%). Monthly operational outlooks were correct 52% of the time. The skill levels in the monthly outlooks during the operational period were very similar to those in earlier experimental tests, being 53% correct in monthly tests of 1940–79. However, the seasonal tests using 1970–80 showed 41% accuracy compared with 56% in the 3-year operational period with the high value a result of sampling vagaries. The monthly outlooks were correct in detecting the occurrence of ending of extreme conditions, including the wet month that ended the 1980-81 drought, the lack of above normal rain during potential flooding conditions in the spring of 1982, and the deficient rainfall in summer 1983. The magnitude of the conditional probability value associated with the most likely monthly outlook value was well related to its correctness. When the maximum probability value obtained was 50% or more, 64% of the monthly outlooks (which specified that category) were correct, but when it was 35 to 45%, only 39% of the outlooks were correct.

Abstract

A methodology to estimate the potential value of proposed weather modification projects is described. An illustration of the technique is given to evaluate a hypothetical hail suppression project. This methodology requires that three crucial sets of data be developed: 1) the benefits attributable to altered weather, 2) the probabilities that such alterations can be accomplished, and 3) the costs associated with this technology. Given these data, a net benefit variable is determined and present value techniques are used to discount that quantity to current dollars.

Abstract

Historical (1901–79) temperature and precipitation data for four Illinois stations were used to determine the frequency with which summer and winter averages for periods of various length (i.e., different climatic normals) are closest to the value for the next year, and hence its best predictor. The normal achieving the highest frequency in this regard is considered the best for characterizing the recent climate for a given point in time and assessing the abnormality of the following year.

Normals for 5, 10, 15, 20 and 25 years were investigated, along with the 30-year ones generally used. Five-year normals most frequently provided the closest estimate of the next year's value for both parameters in both seasons. Ten-year normals also have a high probability of being the best predictors, whereas 20-year normals have a particularly low probability of such success. The standard 30-year normals also perform poorly in this regard. These results contrast strongly with earlier suggestions that 15–25 year normals are “optimum” for prediction because they possess the minimum extrapolation variance when normals are employed as predictors. This difference between the two sets of results indicated that 5-year normals tend to possess larger prediction errors when they are not the best predictors, than do other normals on the greater number of occasions they are not the best predictors. The present findings were used by the Illinois Commerce Commission in evaluating weather normalization rate adjustments proposed by utility companies in 1979–80.

An investigation also is made into the nature of the climatic variation occurring when each normal is the best predictor. Five-year normals tend to attain this position for precipitation when the difference from the preceding year and the departures from longer-term averages are all moderate-to-small. When 5-year normals are the best temperature predictors, in contrast, the departures from this normal (and hence prediction errors) are very large. The frequency with which various normals were the best predictors shows no marked temporal variation during the study period.

Abstract

Weather records for 49 locations and detailed agronomic data from 60 farms in a 400-mi² area in central Illinois were used to study corn-weather relations exhibited by actual farming operations in a typical agricultural area of the American Corn Belt. Weekly, monthly and seasonal rainfall and temperature values plus agronomic data were correlated with corn yields during 1955–1963. The results are of added importance because the data sample was from a period when new agronomic practices, which could alter corn-weather relationships established in previous studies, were being widely employed.

July and August mean temperatures and cumulative degrees above 90F during July and August had the strongest correlations with corn yield, −0.50, −0.69 and −0.51, respectively. July rainfall had a stronger association with yields than rainfall in any other month, but it was considerably less than the association for July temperature.

Weekly rainfalls and temperatures in early June and early August correlated moderately well with yields, a finding that does not agree with those from certain other corn-weather studies. Several weather variables were more highly correlated with corn yield than were any of the newer agronomic practices generally considered to be important factors relating to corn yields. A trend of increasing yields with time during the 9-yr period was related to steady improvement in technological practices and also to a trend for better (cooler) corn weather in July and August.

In general, the results indicate that corn-weather relations defined by data from cash-grain farms during a recent 9-yr period of technological change are somewhat different from those defined in earlier studies using either experimental farm data or regional average yield data.

Abstract

Information on hail intensity is not readily available on a national or local scale. As a result, a study was made to obtain general estimates of the mean areal patterns of hail intensity in the central and northwestern continental United States. Although the results of the study are based on indirect measures of hail intensity developed from crop-insurance data, the results provide useful estimates of a phenomenon heretofore unmeasured. Intensity measures derived from crop-insurance data available for 19 states revealed that summer hailfalls at points in the lee of the Rocky Mountains were 5 to 15 times more intense than those in the Middle West. A good correlation was found for the mean intensity indices of states and the mean hailday frequencies, suggesting that the widely available climatological data on hail frequency could provide useful estimates of hail intensity in other states.

Abstract

Historical hail-day records of U.S. Weather Bureau first-order stations and cooperative substations are the only long, objective records of hail occurrence available throughout the United States. Although hail-day data are limited in areal density and are not necessarily the most desired measure of seeding effects, they are the only data available to obtain a measure of the areal-temporal variability of hail for most areas of the United States. Consequently, hail-day data from Illinois have been employed in a pilot project to determine the time required to obtain statistically significant changes in hail-day frequencies over various sized areas. Four statistical designs were investigated using the historical hail-day data for five areas in Illinois. The results show that the optimum design for hail-day data is the continuous seeding (seeding on all days likely to have hail) over an area. The optimum test is the sequential test involving the Poisson and Negative Binomial distributions. Detection of a 20-percent reduction in summer hail days would require, on the average, a continuous seeding program ranging from 13 to 37 yr, depending on the level of precision desired, and the size and location of the seeded area. Major reductions, those in excess of 60 percent, would require experiments of only 1- to 3-yr length.

Abstract

Data from 103 hail echoes on 24 days in 1967 and 50 no-hail echoes from the same days were analyzed to describe hailstorm characteristics and to provide information useful in operational detection and forecasting of hail-producing echoes. Echo characteristics investigated included locations of echo formation and dissipation, echo reflectivities, echo-top heights, echo duration, direction of motion, speed, time of occurrence, and associated synoptic weather conditions. A single hail-echo model could not be derived because of the extreme variability found in all characteristics. However, distinctive echo models could be developed for the three predominant hail-producing synoptic weather conditions, cold fronts, stationary fronts, and low-pressure centers. The frontal hailstorms were faster moving, longer lived, and had taller echoes than those with low-pressure systems. Hail production after echo inception varied from an average of 32 min for low conditions to 59 min for cold frontal echoes. The average hail-echo top exhibited a 5,000-ft growth in the 15-min period prior to the average time of hail, suggesting that a major updraft surge was the prime producer of hail. The no-hail echoes occurring on hail days had characteristics of speed, direction of motion, reflectivity, and location that were very similar to the hail-producing echoes. The only distinct consistent difference between the hail and no-hail echoes in all synoptic situations was that the hail-echo tops averaged between 2,000 and 4,000 ft higher throughout their entire durations.

Abstract

A statistical methodology involving the analysis of three basic types of historical hail data on an areal approach is presented for the planning and evaluation of hail suppression experiments in Illinois. The methodology was used to generate nomograms relating the number of years required to detect significant results to 1) type I error, 2) type II error, and 3) power of the test for various statistical tests and experimental designs. These nomograms were constructed for various area sizes and geographical locations within the State.

Results indicate that, for an Illinois experiment, insurance crop-loss data are the optimum hail measurement if the study area has more than 60 percent insurance coverage. The optimum experimental design is the random-historical design in which all potential storms are seeded on a particular day, and 80 percent of the forecasted hail days are chosen at random to be “seeded days.” The recommended statistical analysis is the sequential analytical approach. If, however, conditions for the sequential analytical approach are not fulfilled by the data sample, the nonsequential approach utilizing a one-sample test with the historical record as the control (random-historical design) should be employed.

For a significance level of 0.05 and a beta error of 0.3, the average detection time in an area of approximately 1,500 sq mi would be 11 yr for a 20 percent reduction in the number of acres damaged, 2 yr for a 40 percent reduction, and 1 yr for a 60 and 80 percent reduction. If the nonsequential analyses were required, the number of years would be 25, 5, and 1, respectively.