Observations of the spatial structure and motion of auroral absorption regions obtained with the 38.2-MHz imaging riometer for ionospheric studies (IRIS) at South Pole Station, Antarctica, have been used to simulate the absorption that should be measured by the 20.5-, 30-, 38.2-, and 51.4-MHz broad-beam riometers also operating at the station. The simulation assumes that the absorption measured by each of the 49 narrow IRIS beams obeys the absorption-frequency relationship, A(ƒ) ∼ ƒ−2, predicted by generalized magnetoionic theory to apply at altitudes above ∼70 km. Excellent agreement to quite fine detail is obtained between the simulated and measured broad-beam absorption profiles. The power law exponents derived from absorption ratios, or from least squares fits to the absorption as a function of frequency, range between 1 and 2 and are essentially the same for simulated and measured broad-beam data, to within the experimental and computational uncertainties. These results indicate that spatial gradients in the absorption regions, corresponding to similar features in the electron precipitation source, are the cause of deviations from the inverse-square frequency dependence noted previously in broad-beam riometer measurements and erroneously interpreted as signifying very high energy electron precipitation and enhanced ionization at much lower altitudes.