SECT. 1] GBAVITY AT SEA 149 



entire earth on a single equipotential surface, thus eliminating the theoretical 

 difficulty raised by masses between a measurement on land and the sea-level 

 reference surface. Further, it would be possible to carry out the mapping in a 

 matter of decades rather than generations. 



The first airborne tests of the LaCoste-Romberg meter were made by 

 Thompson and LaCoste on board a United States Air Force KC-135 (a military 

 version of the Boeing 707) over Edwards Air Force Base, Cahfornia (Thompson, 

 1959; Thompson and LaCoste, 1960). Precise navigational data were provided 

 from the airplane by a Doppler radar system with barometric measurements 

 for elevation changes which were also obtained from the ground by photo- 

 theodolites. An overall accuracy of about 10 mgal was obtained. 



Further tests were made in a joint venture by the Gravity Meter Exploration 

 Company, Fairchild Aerial Surveys, LaCoste and Romberg, and the Institute 

 of Geophysics of the University of California (Nettleton, LaCoste and Harrison, 

 1960). A B-17 airplane was used; aerial photographs were used for horizontal 

 positioning and altitude, and vertical velocities were determined by a combina- 

 tion of data from a hypsometer (pressure measuring device) and from a precision 

 radar altimeter. Again it was found possible to take airborne gravity measure- 

 ments and, despite some Eotvos corrections in excess of 1000 mgal, the corrected 

 gravity values at the intersection of flight lines usually agreed to within 10 mgal. 

 Repeated flights along the same line gave agreeing results and the contour map, 

 constructed at 12,000 ft elevation from the airborne results, was similar to the 

 map of free-air anomahes on the ground. The actual gravity values at 12,000 ft 

 agreed to within 10 mgal with the ground values projected upward, except 

 where the ground stations were situated on very high terrain and there were 

 insufficient data to project the ground values upward with any confidence. A 

 3-min averaging time was used for these measurements, during which the air- 

 plane covered about 10 miles over the ground. 



These tests show that it is possible to measure gravity with useful accuracy 

 from an airplane and open up the possibility of making a regional survey of the 

 oceans in a few years. However, it is necessary to develop a navigational 

 system, or combination of systems, to enable the airplane to obtain its position, 

 elevation, true course and speed with the required accuracy when over the sea 

 and beyond the range of visual aids to navigation. The error in speed must not 

 greatly exceed one knot, and the airplane's true course must be known with 

 an accuracy of ± |°. A change in vertical velocity during a measuring interval 

 is equivalent to a net vertical acceleration of the meter and can only be dis- 

 tinguished from a change in gravity by accurate measurement of variations in 

 the airplane's height. The hypsometer and radar altimeter used in the B-17 

 tests were accurate to about 2 ft. A sinusoidal height variation of 1-ft amplitude 

 (2-ft range) and 6-min period produces a corresponding acceleration of nearly 

 10 mgal. 



These requirements are stringent but can probably be met. Certainly the 

 method is sufficiently promising to warrant further tests and a careful in- 

 vestigation into the navigational problems. Probably all the relevant data 



