In a study designed to determine the temporal development of ionospheric irregularities in the auroral region in magnetic storm periods, different types of simultaneous observations were compared for the storms of January 10, April 10--11, and May 15, 1997. The data sets consisted of ultraviolet images (UVI) from the Polar satellite, phase fluctuations and total electron content (TEC) from Global Positioning Systems (GPS) recordings at a large number of sites, magnetometer observations and hemispheric power precipitation. The large-scale global or macroscale picture of the magnetic storm showed the importance of universal time in the development of irregularities. The hemispheric total power picture and the global indices such as Kp show this macroscale picture of the entire storm period. However, individual sites show differences in localized magnetic field variations and the development of irregularities; this we term the microscale. The storms of January 10 and May 15 show the importance of local magnetic time and local magnetic variations at the sites, while the storm of April 10--11 was dominated by the UT storm development. During the intense activity of the storms, total electron content shows minute-by-minute increases with the satellite moving to positions 6 km apart in the minute. The structured precipitation either directly produces irregularities or indirectly sets instability conditions for irregularity development in the auroral region. Good correlation was established for the three storms between UVI intensity and phase fluctuation development. The UVI Lyman-Birge-Hopfield-long (170 nm) emission is sensitive to 100--200 km precipitation. Phase fluctuation development undoubtedly arises from perturbations in the F region as well. Topside and bottomside soundings have shown the high occurrence of spread-F in the auroral region. The comparison of the data sets from these storms and from other studies indicates that the creation of irregularities develops from conditions existing over a large range of bottomside heights, i.e., at altitudes from 100 km to the F layer. ¿ 2000 American Geophysical Union