We compare the large‐scale goemagnetic response to the three magnetic clouds observed by Wind in October 1995 (OCT95), May 1996 (MAY96), and January 1997 (JAN97), studying specifically storm and substorm activity, and other global effects due to untypically large and variable solar wind dynamic pressures. Since the temporal profiles of the interplanetary parameters of the three clouds resemble one another closely, the comparison is meaningful. Using the integrated Poynting flux into the magnetosphere as a rough measure of energy input into the magnetosphere, we find relative energy inputs to be OCT95: JAN97:MAY96 = 22:11:4, with most of the accumulation in the 3‐day periods occurring during passage of the Bz < 0 cloud phase. The peak Dst ring current indices, corrected for magnetopause currents, were in the ratio −138:−87:−38, and hence OCT95 caused a major, JAN97 a moderate, and MAY96 a weak storm. The empirical criterion derived from studies near solar maximum that a solar wind dawn‐dusk electric field ≥ 5mV m−1 lasting for at least 3 hours is necessary and sufficient to generate major storms does not hold for JAN97. Storm main phase onset coincides with cloud arrival in all three cases. The number of substorm onsets during the cloud periods were OCT95: JAN97:MAY96 = 5:3:2, with peak AL values in the ratio −1180:−1750:−570. The dayside magnetosphere was variably compressed, the largest amplitude of variation being on JAN97, where the dynamic pressure change spanned 2 orders of magnitude. MAY96 showed the least variation. The interaction of the individual clouds with the faster trailing flows had two major effects on the magnetosphere: (1) a compression of the cavity during passage of the Bz > 0 cloud phase and the leading edge of the fast stream; and (2) a weakening of the control of the cloud field on magnetosheath flow during the Bz > 0 cloud phase. In summary we find that under most of the aspects considered, OCT95 is the most geoeffective. The buffetting of the magnetospheric cavity by dynamic pressure changes was, however, strongest on JAN97. The profound differences in the magnetospheric response elicited by the clouds is found to be due to the amplitude, duration and rapidity of change of the relevant interplanetary parameters. At present, interplanetary monitors are indispensable for understanding the geomagnetic response to interplanetary structures.
