NOAA History Banner
gold bar divider
home - takes you to index page
about the site
noaa - takes you to the noaa home page
search this site
white divider
arrow A Nation at War
arrow WWII
arrow Technology Tales

the weather bureau record of war administration part 10

Page: left arrow 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 click for next page

Not the least important function of the unit was the training of weather officers from the allied military services in the techniques of extended forecasting. Ten weather officers of the Army Air Forces were given a six-week course of intensive training in 1944 and subsequently assigned to active duty in various areas. Instructors from the AAF Weather School at Chanute Field visited the Section from time to time for special instruction, as did also meteorologists from the Navy and from the Canadian Meteorological Office. A weather officer from the British Admiralty was assigned to the Section for an eight-week period for special instruction. Training was provided also in techniques of Northern Hemisphere map analysis and upper-air analysis for officers of the Army Weather Central which was located for a time in the Weather Bureau building.

Members of the staff of the Extended Forecast Section were called upon frequently for special consultation on extended and long-range forecast problems. Mr. Namias was assigned to the Army Air Forces Weather Division for a month asa consultant to work with a meteorological mission from Russia on problems of long-range forecasting. He served also as chairman of an AAF committee on Far Eastern Weather Planning.

A list of titles of major research projects carried on by the Extended Forecast Section during the war is filed in the Archives.

2. The "Extrapolation" Method of Forecasting.-In contrast to the method of forecasting under development by the Extended Forecast Section, which proceeds by consideration of trends and other relationships shown by five-day mean maps, is the method developed by the Special Forecast Section under the guidance of Mr. C. L. Mitchell. The basic difference in the approach to the forecasting problem lies in the use by the Special Forecast Section of weather elements on daily maps and their interrelationship to forecast subsequent weather changes, rather than mean maps as used by the Extended Forecast Section. These two Sections were operated side by side early in the war in order to obtain information on the best features of each method and to provide an interchange of ideas out of which it was hoped would come a combined unit having higher skill than could be attained by either unit working alone. To a considerable extent the interchange of ideas was worthwhile, and some improvement was noted in the accuracy of both methods. Later on the tremendous demand for extended forecast service to the Army and Navy made it necessary to assign the Special Forecast Section almost exclusively to serving the needs of the Navy for forecasts and research.

The specialized needs of the Navy were met by placing the Special Forecast Section in rooms adjacent to the navy Long-Range Forecast Office in the Weather Bureau building. This had the further advantage of making it possible for the Weather Bureau Extended Forecast units to see and use observations data from areas under military control which for security reasons could not be given general distribution.

Mr. Mitchell of this Section also received a citation from Navy Secretary Knox for his work in forecasting for the invasion of North Africa.

Research problems arising in the course of the work were handled by a unit financed largely by funds transferred from the navy. A list of the major projects is contained in the Archives.

3. Forecasting From World-Wide Weather Relationships.-One of the most extensive studies of world-wide atmospheric relations was made in the period of 1906 to 1936 by Sir Gilbert Walker. This work consisted in large part of a search for relationships between the weather of widely different areas of the earth, with the hope that a sufficiently large time lag would be found between the nature of the atmospheric circulation in one area and the subsequent weather in another area that useful forecasts could be made for months and seasons in advance. The work had received considerable attention, particularly the concept of a "Southern Oscillation" which represented an interrelation between seasonal pressures and other elements at selected stations in the Southern Hemisphere. A new survey of this work seemed in order, particularly as forecasting possibilities seemed to include areas in which military operations were taking place, and ten to twenty years of data were available for an independent test of Walker's work.

Accordingly, a small staff began work in January 1943, on the following problems, selected because of their obvious importance in military planning of future campaigns:

  • Forecasting the summer (monsoon) rainfall of Burma.
  • Forecasting the winter temperature and precipitation of Northwest Europe.
  • Forecasting the summer rainfall of Japan.
  • Forecasting the August to November rainfall of selected points in the East Indies and Formosa
  • Forecasting the July and August rainfall of southern Italy and the Mediterranean coast of France.

The program called for a review of literature to find all work done previously on the particular problem, and a search for variable features of the general circulation of the atmosphere which might be expected to control the variations in weather for the selected locations and seasons. In general, an additional ten years of data were available since the earlier work, and these data were used for a test of the continued validity of such relationships as had been suggested in the published literature. If this test looked promising, a forecasting formula was developed, and experimental forecasts prepared.

This work was carried on in part in cooperation with Blue Hill Observatory. Although it was possible to develop a forecasting formula for each of these problems, their reliability was not great enough to justify a high degree of confidence in the forecast. Since it was not possible to ascribe a clear-cut physical process to the relationships which were used, little confidence could be placed in the formula as originally derived and the period of record was in no case long enough to permit a conclusive statistical evaluation of the formula on independent data. It is believed that this method of approach to the long-range forecast problem would be of considerably more value if the development and testing of relationships could proceed from physical reasoning to statistical testing in small units such that each step in the physical process envisioned could be placed on the firm foundation of fact, rather than left as a hypothesis.

4. Forecasting From Solar Phenomena.-Two investigators, Dr. C. G. Abbot, former Secretary of the Smithsonian Institution, and Mr. H.H. Clayton, have from time to time reported on studies of the relation of sunspot, solar radiation, and solar faculae variations to subsequent weather variations. So little is known about the effects which variations in solar radiation might have on the earth's atmosphere that these studies have been almost wholly empirical.

Nevertheless, it is generally recognized that variations in the solar output of energy would in some manner be reflected in the atmospheric processes on the earth, and it is entirely conceivable that variations in the electrical or magnetic fields of the earth due to solar variations other than in radiation might have important influences on weather.

Although research funds available to the Weather Bureau did not permit thorough investigation of these possibilities, some studies were made in an attempt to check hypotheses suggested by Dr. Abbot and Mr. Clayton. A breakdown was made according to years, seasons, latitudes and ocean or land locations, but a detailed analysis of variance disclosed no statistically significant relationship between the sunspot activity as measured and the variations in the surface pressure. Even those effects which approached statistical significance were too small to be considered of practical importance in the immediate development of long-range foreecasts.

5. Forecasting From Ocean Currents and Polar Ice.-An extensive investigation was made of the possibility of estimating the variations in transport of the Gulf Stream as a basis for explaining part of the variation in temperature at stations in Great Britain, France, Germany and other countries of Northeast Europe. Earlier work had been done on this problem but meteorologists were not in agreement, nor were oceanographers, as to the extent to which variations in Gulf Stream transport would influence surface water temperatures in the northeast Atlantic, nor the extent to which such variations of water temperature would influence air temperature over adjacent land areas.

The plan in this study was to estimate changes in Gulf Stream transport by using wind observations along the course of the Stream, on the assumption that the greater the component of wind speed in the direction of flow, the greater would be the volume transport. It is known that this assumption is not entirely correct, but direct observations of Gulf Stream transport were not available. The statistical analysis gave a slight suggestion of a real relationship between the wind and subsequent temperatures at lags of five and six months, but even those effects were too small to be of practical importance for forecasting.

In cooperation with Blue Hill Observatory, a study was made of the relation of the extent of ice in the polar seas to subsequent temperature and precipitation in Europe. Although this study was based on what seemed to be reasonable physical theory, many cases were found in the record which did not support the hypothesis. Nevertheless, an experimental forecast was made for a number of European stations for October to December 1944. In general the verification was poor. Work on this project was discontinued in July 1945, because of the poor prospect of obtaining useful results in the immediate future.

6. Persistence as a Factor in Forecasting.-An attack on the basic problem of persistence in the general circulation was made by computing correlations between successive five-day pressure patterns for the Northern Hemisphere stratified into areas representing the branches of the general circulation. The object was to discover under what conditions pressure patterns tend to be stable (persistent) and under what conditions they tend to change rapidly. The period for which five-day mean pressures were available was not long enough for any conclusions to be drawn, although it appeared from the preliminary study that there were results of potential value in extended and long-range forecasting. This study should be continued with the more adequate data now available.

Another project was carried through to examine the persistence in monthly mean temperature and precipitation. Auto-correlations with lags up to 12 months were computed for 100 years of London temperature, Rome temperature and precipitation, and New Haven and St. Louis temperature observations. Significant correlations were found and in the case of London were used in preparing experimental monthly temperature forecasts for summer months. Because these forecasts did not have a high degree of reliability, and because personnel were not available to make the necessary meteorological studies to combine this information with other factors, the investigation was discontinued.

One further project was completed which was significant in the development of the extended and long-range forecast research program, but which was not aimed specifically at solving a forecast problems. This was a project to test the potential usefulness of prognostic pressure patterns in extended forecasting. The project was designed to help indicate the direction in which research efforts might be most effectively used to quickly increase the accuracy of extended forecasts. The results, by indicating the relative magnitude of errors in two phases of the forecasting procedure, led to a useful change in emphasis in the extended forecast research program. Whenever in the future it is required that research facilities be used in the most effective manner possible, such experiments should be very useful in pointing the direction for research and development projects.

Flight Advisory Weather Service.-The number of military flights (scheduled, training, and transient) increased steadily throughout the period of the war. Protection of these flights, as well as protection of civilian scheduled and transient operations required the development of a meteorological service which would concentrate its attention upon their specific needs.

To meet these requirements the Weather Bureau issued to this filed stations in June 1943 instructions concerning the provision of special services to pilots and to traffic controllers. At the same time an experimental unit was established at la Guardia Field to provide highly specialized service to the Airway Traffic Controllers, and through them to the pilots in flights. The success of this installation resulted in establishment of another unit at Washington National Airport in September 1943, this one planned along somewhat different lines. At the request of the Army Air Forces this program was expanded as rapidly as possible, and 22 additional units were established during the late months of 1943 and the early months of 1944.

At about the same time the Army Air Forces inaugurated their Flight Service Centers, located adjacent to the Flight Advisory Weather Service units in the Air Traffic Control centers. Army Flight Service personnel acted as clearance officers for many classes of military flights and also checked every military flight plan to be sure that weather was satisfactory and that navigational facilities and field conditions were suitable.

The volume of traffic handled increased steadily to a peak in July and August of 1945 when more than 26,000 clearances per month were issued and flight plans involving more than 650,000 aircraft per month were handled. The meteorological information for this task was provided by the Flight Advisory Weather Service. The Army has expressed full approval of the service rendered by the Flight Advisory Weather Service units and has indicated that only a very small number of advisories were found to be unsound.

Personnel were provided to maintain a continuos watch over weather along the airways and at terminals. Teletype installations in the Flight Advisory Weather Service offices insured receipt of airway forecasts and weather reports as early as possible, and close contact with communications operators provided prompt receipt of in-flight weather reports transmitted by radio. With this information available the Flight Advisory Weather Service forecaster was able to keep Air Traffic Control and Army Flight Service personnel advised promptly of weather developments and anticipated changes which were significant in the dispatching, control, and protection of both civil and military aircraft.

- Top of Page -

Publication of the National Oceanic & Atmospheric Administration (NOAA), NOAA Central Library.

Last Updated: June 8, 2006 9:24 AM

Privacy Policy | Disclaimer