Article in press (Icarus) is available in pdf format (icarus.pdf 475 Kb).
 

Summary:

We observe Io, volcanic satellite of Jupiter, in the near-infrared (1-2.5 microns) using the 10-m Keck II AO system on February 19,20 and 22, 2001. We used the NIRSPEC camera with a pixel size of 16.8 milliarcsec (mas). Applying the MISTRAL (Conan et al., 1998), a deconvolution method developed by ONERA and especially aimed at AO observations of planetary objects, we enhanced the sharpness of the image and achieve a final resolution of 35 mas in H band (1.62 microns), equivalent to the diffraction limit of the this 10-m telescope. Figure. 1 displays the 20 Feb. JHK images, after and before deconvolutions. The spatial resolution is 105 km on Io's surface, roughly three times better than the most of the global images of Galileo/NIMS (Doute et al., 2001).
 
 

Fig. 1: (A) Jupiter-facing hemisphere observed with the Keck AO system. the basic-processed images of the 20 February 2001 (first row) is displayed. The second row corresponds to the same images after applying the MISTRAL deconvolution process. Albedo feature, similars to the 20-km resolution reconstructed image (right column) are easily detected. The last row shows the 22 February 2001 images which are dominated by the presence of the Surt outburst. (B) Observations of the February 2001. Two hot spots, corresponding to Tvashtar (North) and Amirani are clearly detected in the H and K bands. [click on the figures to magnify them]

We compare our images with a Minnaert law reconstructed visible images based on the 23 km-resolution Galileo/SSI map (McEwen et al., 1998). Same albedo feature are visible on the 20 Feb. image. One can see easily the Loki, Daedalus, Svarog, Pele caldera and also the plume deposit around Pele. On the other two nights (Fig. 1b), several bright hot spots stood out with a much higher contrast than this albedo feature. These correspond to thermal emission from active hot spots, surprinsingly detected directly, against Io's sunlit hemisphere.
 

Fig. 2:  Localisation of the thermal sources detected on the AO images and comparison with a Galileo/Voyager map. The black square on the images indicates the position estimated with our AO observation (see Tab. 1). We compare the position of Surt with two Voyager images and a Galileo image. Note the surface change feature in the second Voyager image. The appearance of this caldera may have triggered the bright eruption seen from the ground by Sinton (1980). Its position is compatible to the bright outburst obtained from our fata.
  Table 1: Intensity, Temperature and Emissive Area of the Hot spots estimated from the black-body fits.[click on the table to magnify]

Locations of the hot spots correspond precisely to the known active area Amirani (010219A), Tvashtar (010219B) and Surt (010220A and 010222A) as displaid in the Fig. 2. Multi-wavelength detection of the thermal flux allows us to estimate temperature and areal coverage of the emission sources (Table 1). In the case of Surt we also applied a Silicate lava cooling model (Davies, 1996) which calculates the distribution of temperatures and areas seen during an eruption as a function of time. Assuming a basaltic composition for the eruption magma, the model fit to the Surt data produces a range of areas at temperatures ranging from a fraction of a square kilometer at 1475 K down to 8.9 km2 at 1030 K. The total area of the entire thermal anomaly is approximately 800 km2. The youghness of the oldest area (less than 100s) suggests that the eruption style is extremely dynamic, indicative of a highly-vigorous style of lave emplacement, such as a large fire fountains.

Comparing the thermal signature of the Surt eruption observed with other seen by Galileo NIMS, one can see that the thermal output  (7.4E13 GW) of this outburst is considerably larger than any other eruptions yet seen by the Galileo spacecraft (Fig. 3). This is in fact the most energetic outburst ever observed on Io.
Adding the three eruption intensities, we derive a mean total output for the February 2001 period of 1.86 W.m-2. It nearly matches the average global heat flow of 2.5 W.m-2 of Veeder et al. (1994).

Fig. 3: Best-fit spectra to the observation. The blue and purple lines correspond to Amirani and Tvashtar. The bold red line is the model spectrum of Surt derived from the silicate-cooling flow model. The classical 2-temperature model is overlaid in red normal line with its two components in dash-dash and dot-dot. For comparison, we displaid in green the spectrum deduced from the galileo/NIMS observations. The eruption detected by the AO observations clearly correspond to a much stronger eruption. The total output of each eruption is indicated (in GW).
 
 

References
Conan et al., Applied Optics, 37, 21, pp. 4614, 1998.
Doute et al., Icarus, 149, 1, pp. 107-132, 2001.
Sinton, W.M., ApJ, 235, p. L49-L51, 1980
Davies, A., Icarus, 124, 1, pp. 45-61, 1996
McEwen, A. et al., Icarus, 135, pp. 181-219, 1998