Introduction

In this letter, we examine how the Earth's magnetosphere responded to an unusually tenuous solar wind using magnetic field data from the Polar spacecraft. On May 10-12, 1999, a prolonged interval of extremely low solar wind density was observed by Wind and ACE spacecraft. Figure 1 shows the magnetic field data [Lepping et al., 1995] and solar wind data [Ogilvie et al., 1995] from the WIND spacecraft as well as the Dst index covering this prolonged period. The solar wind density started to decrease on May 10 and dropped to below 1 cm^-3 for over a day during the interval from $\sim$ 21 UT, May 10 to $\sim$ 02 UT, May 12. At the minimum the solar wind dynamic pressure was less than 0.1 nPa. Meanwhile, the solar wind flow velocity maintained its normal value of $\sim$ 350 km/s and the interplanetary magnetic field was $\sim$ 6 nT and weakly northward on average. The Earth's bow shock was crossed by the Wind spacecraft as far as $\sim$ 53 R_E in radial distance from the Earth, as evident by the brief period of enhanced magnetic field strength from 1730 to 1936 UT on May 11 when Wind was in the magnetosheath. The subsolar bow shock location is estimated from current empirical models to have moved out to at least 58 R_E.

We compare the Polar magnetic field observations on May 11 to a control day, May 14 in this study. The solar wind and IMF conditions for May 14 are also shown in Figure 1. This control day was selected based on several factors that enable us to concentrate on the investigation of the effect of the solar wind dynamic pressure on the magnetosphere. First of all, Polar probes the same region of the magnetosphere at the same time of the day every three days because its orbital period is close to 18 hours. Thus the effect of differing dipole tilt angle, if any, is minimal when we compare observations 3 days apart. This procedure allows a direct comparison of various regions of the magnetosphere seen on these two days. Secondly, on the control day of May 14 the solar wind had a typical plasma condition with a density of $\sim$ 5 cm^-3, a velocity of $\sim$ 440 km/s and a dynamic pressure of $\sim$ 2 nPa. Thus, observations on May 14 provide a "baseline" for comparing the magnetospheric behavior under typical solar wind conditions. Lastly, the IMF conditions were similar for these two days. The IMF Bz for May 14 was also weakly northward on average. Moreover, the IMF magnitude and cone angle were similar on May 11 and 14, providing suitable conditions for comparing Pc 3-4 pulsations in the magnetosphere, a topic of future research.

Figure 2 shows one complete Polar orbit from 04 to 22 UT on May 11 in GSM coordinates. The Polar orbit on May 14 is essentially the same as that on May 11. The Polar orbit plane was very close to the noon-might meridian for these two days. The magnetopause drawn is that expected for May 14, i.e., for a nominal solar wind dynamic pressure of 2 nPa and northward IMF. To demonstrate the inflation of the magnetosphere due to the extremely low solar wind dynamic pressure on May 11, we scale the size of the Polar orbit by the sixth root of the time-varying solar wind dynamic pressure seen at Wind, shown as the irregularly shaped trace in Figure 2. Polar is clearly confined to the inner magnetosphere on May 11, 1999.