Have you ever flown into an Alaskan airport and seen one of these curious buildings alongside the runway? They were known by several names: balloon inflation shelter, rawinsonde building, Upper Air Facility, or Upper Air Inflation Building (UAIB), among others. Although small and utilitarian, the UAIB had an oversized role in Alaskan history. As the buildings where the balloons used for the rawinsonde program were stored, inflated, and tracked, the UAIBs contributed to climate-related research and the development of improved and more comprehensive weather forecasting in the United States. 

In Alaska, where aviation became an essential form of transportation given the nature of the terrain and road system, reliable and local weather forecasting was particularly important. The UAIBs, with their distinctive fiberglass radomes, represent both a technological approach to obtaining upper atmospheric conditions and the collaboration between federal agencies such as the Civilian Aeronautics Administration (CAA, later Federal Aviation Administration) and Weather Bureau (later National Weather Service), the military, and civic forces in promoting Alaskan aviation. 

Design Chronology

Design of the UAIBs originated in the Weather Bureau’s recommendation that “whenever possible,” weather balloons should be inflated in a shelter because “inflation in the open is usually unsatisfactory because of the difficulty in estimating total lift. Since much of the success of a sounding is dependent upon the ascensional rate, and hence the lift, it is important that the inflation be accomplished under controlled conditions.”(1) As the Weather Bureau began to establish radiosonde stations across the United States in the late 1930s, development of a standard inflation shelter followed. Over a period of about five years, the Weather Bureau devised a standard inflation shelter that embodied “features found desirable at many stations.” The wood structures measured 11 feet wide by 13 feet long with “garage-type doors” to allow an exit with a fully inflated balloon and a flat roof for observations. The Weather Bureau built one of these standardized UAIBs at Juneau, one of the early Alaskan radiosonde stations established in 1938. 

First generation UAIB right) at “an isolated Alaskan weather station at a landing field” in 1943.

First generation UAIB (left) at Juneau. Control Building with the Weather Bureau office on the right, 1950.

World War II accelerated development of Alaska’s communications, weather service, and aviation networks to counter the threat from Japan. When Germany’s invasion of Poland triggered war in 1939, Alaska was relatively undefended. Alongside military efforts to improve Alaska’s defenses with a vastly expanded communications system and construction of air and naval bases, the CAA proposed building 200 airfields in Alaska at an estimated cost of $20,550,000.(2) Although the full Alaskan program was scaled back, many locations received funding for new airfields or improvement of existing facilities. The Weather Bureau expanded its Alaskan operations to match the growth in aviation infrastructure, adding radiosonde stations at Bethel, Nome, Point Barrow, Anchorage, and Ketchikan by 1941. Others at McGrath, Kotzebue, and Gambell became operational the following year.(3)

Map of Radiosonde and Rawinsonde Stations in Alaska, 1948.

 These early stations used the Weather Bureau’s first generation standardized UAIB design. Some had an attached shed, which may have housed a hydrogen generator. The Weather Bureau preferred using helium as an inflation gas for the weather balloons, but hydrogen was cheaper and easier to generate with onsite equipment. Hydrogen, however, had an unfortunate propensity to catch fire and explode if not properly handled. Some photographs show metal gas storage tanks, which may have contained helium, piled near the UAIB. UAIBs at more remote locations, which would have incurred high shipping costs for helium, typically had an accompanying shed, probably for a hydrogen generator.  

UAIB (right) at Bethel with attached shed, 1953.

UAIB (bottom left) at McGrath with attached shed, 1950.

Alaskan UAIBs were located at airfields and airports, usually near the “control building” or more officially, the “air traffic control services” building, where the Weather Bureau maintained an office. The close proximity of the weather personnel to airfield operations facilitated an exchange of information. Small complexes shared by the CAA and Weather Bureau developed at these sites. Typically consisting of a control building, various quarters, power plant, communications equipment, warehouses and storage sheds, the complexes reified the cooperative relationship between aviation and the Weather Bureau.

Because of the importance of weather forecasting to military operations, the Army and Navy invested in improving upper atmospheric observations during the war. At its Eatontown Signal Laboratory, the Army Signal Corps worked on hydrogen gas generators, balloon materials, and a radio direction finder. The Signal Corps’ meteorological direction finder “could track up to 60,000 feet through the darkness or cloud the signals of a radiosonde dangling beneath a soaring weather balloon.”(4) Later automated, the radio-theodolites allowed the determination of winds aloft data (radio-wind or “rawin”) from the radiosonde while in flight. 

Photo of Kotzebue station in 1954 shows both new UAIB type (left) and original type (right).

Adding rawin capabilities to the weather stations led to the development of a new UAIB in the mid-1950s. Slightly larger, the new UAIB design had a similar shape and materials except for a protective radar dome or “radome” on the roof to shelter the rawin tracking equipment. Whereas the earlier UAIBs might have been mistaken for a tall garage, the distinctive radomes — constructed of “radio-transparent” fiberglass — pointed to something else. The Weather Bureau began replacing older UAIBs with the new version in the mid to late 1950s.(5) 

By the late 1960s, 66 radiosonde stations in the contiguous United States, 14 in Alaska, and two in Hawaii were sending up two balloons a day, either a 600-gram neoprene balloon inflated with 119 cubic feet of helium, or a larger 1,200-gram balloon with 136 cubic feet. Annually this equated to about 17 million cubic feet of helium. With natural reserves of helium declining and its cost soaring, the Weather Bureau noted that “the outlook for the future use of helium in radiosonde soundings is rather bleak.”(6) As a result, in the late 1960s the Weather Bureau began shifting the gas used in its high-altitude balloon program to hydrogen, despite its higher potential for explosions, because of the high cost and relative scarcity of helium. UAIBs with lean-to additions built off the main massing to house hydrogen generators reflect this conversion. As part of this transition, typically a metal hydrogen storage tank was located outside the building in a fenced enclosure. UAIBs following this pattern of construction include those located in Cold Bay, King Salmon, Kotzebue, Nome, and Yakutat.(7) 

As the expense of maintaining personnel at the more remote Alaskan locations became an increasing burden for the Weather Bureau (renamed the National Weather Service in 1970), staffing was reduced and housing eliminated. Most of the quarters, offices, and structures associated with the former CAA and Weather Bureau complexes were torn down by the 1990s. Later, the NWS pursued another automation project to reduce costs: the replacement of human launched rawinsonde balloons with autolaunchers. After a successful demonstration project with an autolauncher at Kodiak in 2017, the NWS moved forward with installing autolaunchers at 11 other UAIB sites. A NOAA representative explained that the shift to the autolaunchers was in part due to the difficulty of recruiting for the remote locations, but also to save money and free up staff to work on more pressing matters.(8) Although loss of the UAIBs will remove the last vestiges of the CAA and Weather Bureau stations in Alaska, the autolauncher represents the continued importance of the balloon-based upper atmospheric rawinsonde program, if in new form.

“The importance of the radiosonde as a tool for gathering meteorological data is enormous. The radiosonde allows, in a matter of minutes and at relatively low cost, measurement and transmission of data from regions mostly inaccessible to humans.” (9)

“Understanding and accurately predicting changes in the atmosphere requires adequate observations of the upper atmosphere. Radiosondes  provide a primary source of upper-air data and will remain so into the foreseeable future.”(10)

Alaskan UAIB Gallery

Bethel

Original UAIB at Bethel (right), CAA/Weather Bureau offices (left), 1953.

Flooding inundates Bethel airport and CAA/Weather Bureau complex in 1955. Original UAIB indicated with arrow.

Recurrent flooding at Bethel prompted construction of a new airport at a less vulnerable location in the late 1950s. New UAIB shown staged for relocation, 1958.

UAIB in a new location near air traffic control service building at recently relocated Bethel airfield, 1959.

Aerial photograph showing UAIB location adjacent to ATCS building at new Bethel Airport, 1960. CAA/Weather Bureau complex in upper right corner.

Aerial photograph showing relocation of Bethel UAIB closer to CAA/Weather Bureau complex, 1963.

After relocation of Bethel UAIB, CAA/Weather Bureau complex in the background, 1963.

Bethel UAIB showing lean-to addition, 2022.

Bethel UAIB (left), new autolauncher (right), 2022.

McGrath

Aerial photograph showing relationship between CAA/Weather Bureau complex and UAIB at McGrath, 1953.

CAA/Weather Bureau housing at McGrath, 1956.

Original UAIB at McGrath, circa 1956.

New UAIB building at McGrath in the same location as original, 1960.

Second generation UAIB at McGrath adjacent to ATCS building, circa 1960.

As-built drawing of lean-to addition for hydrogen generator, 1966.

As-built west elevation of McGrath UAIB showing lean-to addition, 1966.

McGrath UAIB showing lean-to addition, 2022.

Additional Reading and Endnotes

• Out of Thin Air: The History and Evolution of Upper-Air Observations

• The Trusty Weather Balloon

1. United States Department of Commerce, Instructions for Modulated Audio Frequency Radiosonde Observations (Washington DC: Government Printing Office, 1943), 47.

2. “Reveal Huge Alaska Airfield Program,” Anchorage Daily Times, September 30, 1940, 1.

3. United States Department of Commerce, The Weather Bureau Record of War Administration, Part 10 (Washington DC: Government Printing Office, 1948, 58. 

4. George Raynor Thompson and Dixie R. Harris, The Signal Corps: The Outcome (Mid-1943 Through 1945) (Washington DC: Office of the Chief of Military History, 1966), 465.

5. US Department of Commerce, Instructions for Modulated Audio Frequency Radiosonde Observations, 48.

6. James K. Angeli, “Meteorological Soundings by Balloon” Helium Symposia Proceedings in 1968—A Hundred Years of Helium (1969): 22-23.

7. Hart Crowser, Inc., “Cultural Resources Inventory and Evaluation, National Weather Service Facilities, Alaska,” prepared for SRI International (February 26, 2003): 15–16, 31, 33, 37, 42, 49.

8. Jina Rosen, “Robot-Launched Weather Balloons in Alaska Hasten Demise of Remote Stations,” Science (25 April 2018).

9. Federico Flores, et. al., “The Life Cycle of a Radiosonde,” Bulletin of the American Meteorological Society 94 (February 2013): 187.

10. National Weather Service, “Radiosonde Observation,” https://www.weather.gov/upperair/factsheet, accessed December 12, 2025.