mm 
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008 
2. DATA PROCESSING 
The image data returned from the spacecraft are distributed to 
the Cassini imaging team in VICAR (Video Image 
Communication and Retrieval) format [http://www- 
mipl.jpl.nasa.gov/extemal/vicar.html]. The first step of the 
image processing is the radiometric calibration of the images 
using the ISS Team’s CISSCAL computer program (Porco et al., 
2004). The next step of the processing chain is to map project 
the images to the proper scale and map format - a process that 
requires detailed information about the global shape of the 
satellite. The inner Saturnian satellites are best described by tri- 
axial ellipsoids as derived from ISS images by Thomas et al. 
(2006). However, to facilitate comparison and interpretation of 
the maps, ellipsoids were used only for the calculation of the 
ray intersection points, while the map projection itself was done 
onto a sphere with the mean radius. The Cassini orbit and 
attitude data for the calculation of the surface intersection 
points are provided as SPICE kernels [http://naif.jpl.nasa.gov] 
and were improved using a limb-fitting technique (Roatsch et 
al., 2006). The coordinate system adopted by the Cassini 
mission for satellite mapping is the IAU “planetographic” 
system, consisting of planetographic latitude and positive West 
longitude (Seidelmann et al., 2006). Because a spherical 
reference surface is used for map projections of the satellites, 
planetographic and planetocentric latitudes are numerically 
equal. The Hapke photometric model (Hapke, 1993) was 
applied to adjust the brightness of each map pixel so that it 
represents the reflectance that would be observed at the nadir at 
30-deg phase angle. Imaging data viewed at incidence and 
emission angles greater than 80-deg were omitted from the map. 
After photometric correction, mosaicking is the final step of the 
image processing (Roatsch et al., 2006). 
3-D control nets are common to correct errors in the nominal 
camera pointing data. Due to the lack of appropriate stereo data, 
only in the case of the Enceladus mosaic we got acceptable 
results of the least- squares adjustment. 
3. MOSAICS AND BASEMAPS 
Imaging of the medium-sized icy satellites is ongoing and will 
continue until the end of the Cassini mission, making it possible 
to improve the image mosaics during the tour. 
The basemaps (global mosaics) are usually produced using 
images of very different resolutions. Effort was made to choose 
a map resolution in pixel per degree expressible as 2 n , where n 
is an integer. It was not practicable to follow this rule for all 
basemaps (see Table 2) avoiding a loss of high-resolution image 
information. The map projection for the basemaps is the 
equidistant projection centered at 0 degrees longitude. These 
basemaps cover the whole satellite from -180° till 180° West 
longitude and from -90° till 90° latitude. However, some areas 
of the satellites are imaged at very high resolution. These higher 
resolution images were processed to separate mosaics (see 
Figure 1 for an example). 
At the time of writing, thousands of images from the icy 
satellites are available (e.g. 3073 Enceladus images, 2553 
Dione images, and 2000 Tethys images). This total data set 
contains images obtained through a variety of different ISS 
color filters and at spatial resolutions ranging from 3 m/pixel up 
to 14 km/pixel. For our mosaic, we selected only those images 
taken with the filters CL1, CL2 or GRN, as these images show 
comparable albedo contrasts among different surface terrains. 
For example, 83 Cassini NAC images and two Voyager image 
were used to produce a 64 pixel/deg global mosaic of Dione 
(Figure 2). The resolution of the selected Cassini images varies 
between 0.16 and 1.72 km/pixel whereas the resolution of the 
Voyager images are 1.45 km/pixel and 6 km/pixel. 
4. ICY SATLLITE ATLASES 
We produced a global printed map of Phoebe in a scale of 
1:1,000,000 and atlases of Enceladus (1: 500,000), Dione 
(1:1,000,000), and Tethys (1:1,000,000) (Table2). 
Satellite 
Mean radius 
used for the 
map 
projection 
[km] 
Resolution 
of the 
digital map 
[pixel/ 
degree] 
Map scale of the 
atlas 
Mimas 
198.8 
8 
- 
Enceladus 
252.1 
40 
1:500.000 
Tethys 
536.3 
32 
1:1,000,000 
Dione 
562.53 
64 
1:1,000,000 
Rhea 
764.0 
20 
- 
Iapetus 
736.0 
16 
1:3,000,000 
. (in preparation) 
Phoebe 
106.8 
8 
1:1,000,000 
Table 2. Resolution and scale of the maps 
The atlases consist of 15 tiles each of them conform to the 
quadrangle scheme proposed by Greeley and Batson (1990) and 
Kirk (1997, 2002, 2003) for large satellites (Figure 3). The 
scales guarantee a mapping at the highest available Cassini 
resolution and results in an acceptable printing scale for the 
hardcopy map of 4.5 pixel/mm. The individual tiles were 
extracted from the global mosaic and reprojected, coordinate 
grids were superposed as graphic vectors and the resulting 
composites were converted to the common PDF-format using 
the software package Planetary Image Mapper (PIMap) of the 
Technical University Berlin (Gehrke et al., 2006). The 
equatorial part of the map (-22° to 22° latitude) is in Mercator 
projection onto a secant cylinder using standard parallels at -13° 
and 13° latitude. The regions between the equator region and 
the poles (-66° to -21° and 21° to 66° latitude) are projected in 
Lambert conic projection with two standard parallels at -30° 
and -58° (or 30° and 58°, respectively). The poles are projected 
in stereographic projection (-90° to -65° latitude and 65° to 90° 
latitude). The Mercator maps are 72° in longitude dimension, 
the Lambert maps 90°, and the poles 360°. The individual tiles 
overlap in the North-South direction by one degree, however no 
overlapping region is present in the East-West direction (see 
Figure 3). We have produced the maps using the same scaling 
factors in overlapping regions at the matching parallels ±21.34° 
and ±65.19° latitude, 1.0461 and 1.0484 respectively (Snyder, 
1987). We also added a resolution map and an index map for 
every individual tile, showing the image resolution, the image 
numbers and the location of the images within the individual 
quadrangle (Roatsch et al., 2008a and b). 
The Cassini imaging team proposed new names for geological 
features, in addition to the features already named by the 
1012