At long wavelengths the technique is more difficult because Jupiter is much brighter -- but the long wavelength measurements are essential if one is to obtain the total energy output from the hot spots.
Shown below is an image of Jupiter and Io shortly before the occultation disappearance begins. Io is already in eclipse and can just barely be seen in this image adjusted to show Jupiter. However if the display is adjusted to emphasize Io it is clearly evident just to the right of Jupiter. It is also evident when one subtract from each frame an image of Jupiter obtained after Io has disappeared.
When one processes the series of images obtained as Io disappears, then measures the brightness of Io, one obtains the following curve. It clearly shows the disappearance of one major hotspot (Loki) which contributes 2/3 of the total flux. The disappearance of the remaining flux can be modeled by a smooth decrease which would imply many faint hotspots distributed across the disk. However there is some slight evidence for a two-step character for that remaining flux, which would be consistent with the shorter-wavelength occultation lightcurve obtained from the Wyoming Infrared Observatory (WIRO) two days earlier by Robert Howell and Justin Gregg.
The location of the three possible hotspots on Io can be obtained by drawing on a map of Io the location of the Jupiter limb at the times of the three steps. When that is done the location of the large step clearly corresponds to Loki, which was undergoing a major eruption during the summer of 1998. The first step may correspond to Kanehekili a persistent hotspot marked by the small square.
The noise in the 10.3 micron occultation is caused by several factors. The primary sources are imperfect registration and subtraction of the Jupiter image and missing flux caused by "bad pixels" in the camera. One of the 16 readout channels was not functioning, causing the pattern of bad pixels seen in the images below. Because Io occupies several pixels it is possible to interpolate across the missing ones to obtain a reasonable estimate of the flux, which was done in the above images and for the above lightcurve. However the interpolation is imperfect, especially if the peak Io pixel falls on or adjacent to one of the bad pixels. That effect is probably the cause of several discrepant points in the above lightcurve.
In the following movie the first frame is shown without Jupiter subtraction, and with the bad pixels "fixed" by interpolation. In subsequent frames a Jupiter image has been centered and subtracted, and the bad pixels are shown as black to enable one to judge possible interpolation problems. The Loki disappearance occurs at frame 71, 184 seconds after the occultation disappearance began. The frames are numbered and also labeled with time after event start.
25: -45 sec 26: -40 sec 27: -35 sec 28: -30 sec 29: -25 sec
30: -20 sec 31: -15 sec 32: -10 sec 33: - 5 sec 34: 0 sec
35: 5 sec 36: 10 sec 37: 15 sec 39: 21 sec 39: 26 sec
40: 30 sec 41: 36 sec 42: 41 sec 43: 46 sec 44: 51 sec
45: 56 sec 46: 61 sec 47: 66 sec 48: 71 sec 49: 76 sec
50: 81 sec 51: 86 sec 52: 91 sec 53: 96 sec 54: 102 sec
55: 107 sec 56: 112 sec 57: 116 sec 58: 122 sec 59: 127 sec
60: 132 sec 61: 137 sec 62: 142 sec 63: 147 sec 64: 152 sec
65: 157 sec 66: 162 sec 67: 168 sec 68: 172 sec 69: 177 sec
70: 182 sec 71: 188 sec 72: 193 sec 73: 198 sec 74: 203 sec
75: 208 sec 76: 213 sec 77: 218 sec 78: 223 sec 79: 228 sec
80: 233 sec 81: 238 sec 82: 243 sec 83: 248 sec 84: 253 sec
Sept. 28, 1998 Occultation results