They use this as ref:
2.2 Historical Examination of the Wake- Turbulence Hazard
Wake turbulence is a natural by-product of powered flight, but was not generally regarded as a serious flight hazard until the late 1960s. Upsets or turbulence encounters associated with other aircraft were usually accredited to “propwash” and later on, with “jet wash.” Interest in this phenomenon greatly increased with the introduction of large, wide-body tur- bojet aircraft during the late 1960s and a con- cern about the impact of greater wake turbulence. This was the impetus to conduct research to gain additional information and determine what safety considerations were necessary as more and more large aircraft entered the industry fleets.
An investigation of the wake-turbulence phe- nomenon, conducted by Boeing in mid 1969 as part of the FAA test program, included both analysis and limited flight test and pro- duced more detailed information on wake vortices. The flight tests provided a direct comparison between the B-747 and a repre- sentative from the then current jet fleet, a B- 707-320C. The smallest Boeing jet transport, the B-737-100, was used as the primary wake- turbulence probing aircraft along with an F- 86 and the NASA CV-990. Smoke generating
towers were also used to observe the wake turbulence generated by aircraft as they flew by. Several observations were made.
• The strength of the wake turbulence is governed by the weight, speed and wing- span of the generating aircraft.
• The greatest strength occurs when the generating aircraft is heavy, at slow speed with a clean wing configuration.
Initial flight tests produced sufficient infor- mation about the strength, duration and move- ment of wake turbulence to come to conclusions and recommendations on how to avoid it. The wake was observed to move down initially and then level off. It was never encountered at the same flight level as the generating aircraft or more than 900 feet be- low the generating aircraft. Therefore, a fol- lowing aircraft could avoid the wake turbulence by flying above the flightpath of the leading aircraft. While this can be accom- plished in visual conditions, an alternative was developed for instrument meteorologi- cal conditions. Aircraft were placed into cat- egories determined by their gross weight. It was noted that a division based on the wing- span of the following aircraft was a more technically correct way to establish catego- ries; however, it did not appear to be an easily workable method. Since there is a correlation between aircraft gross weight and wingspan, gross weight was selected as a means of cat- egorizing aircraft and wake-turbulence strength. Minimum radar-controlled wake- turbulence separation distances were estab- lished for following aircraft. The separation distances depend on the weight of both the leading and following aircraft. Adjustments in separation distances were made as more information on the wake-turbulence phenom- enon was gained during the 1960s, 1980s and 1990s, but the basic concept of using aircraft weights remained constant.