In the Spencer, South Dakota tornado (1998), Doppler measurements at intervals of approximately 25-30 m are compared to a block by block F-scale estimation. This provides the most detailed comparison of Doppler measurements and a 2D area of damage. In the Spencer tornado, and in others, including the Mulhall, Oklahoma tornado (1999), and the Bridge Creek-Moore-Oklahoma City, Oklahoma (1999) tornado, Doppler measurements of winds ranged from 100-130 m/s, suggesting maximum speeds slightly higher, ranging from 110-135 m/s. However, interpretation of these measurements is problematical. High windspeeds in the Mulhall and Bridge Creek tornadoes were observed over very limited areas and would have translated across stationary observers in fractions of a second. In Mulhall, these peak winds were present in small (200 m scale) multiple vortices that propogated about the central tornado at 50 m/s. The peak measured winds would have occured over a stationary observer for only about 1 s. Therefore comparison with "standard" ground based measurements normalized to 1/4 mile of wind (Fujita) or 3 second gusts is difficult. Equally important is the level at which radar winds are typically measured, which is far above the height of standard wind measurements (10 m) or Fujita-scale comparisons (3-5 m).