5.3.2 The Ionia “1km” Multi Platform CD-Browser software 
The browse software is based on the Ionia CD-browser philosophy, overall it will work on Multi Platform 
including, Unix machine. Mac and PC and will be included in the Multi Mission Browsing System actually 
under development at ESRIN and will allow: to address the overall data set, to emit a query, to display the Quick 
Look products, to select/deselect and finally to generate a product order. 
5.3.3 The Ionia “1km” Net-Browser software 
The Tnriia “1km” Net-Browser is based on the same software environment as its predecessor described above. 
When linked, a normal session begins with the query parameters to be entered by the user (geographic 
coordinates and time). The geographic selection can also be done using the mouse by clicking on a world map. 
The retrieved items are presented as time dependent lists of products coming from the overall accessible Quick 
Look archive. When rlirking in this list the IEF information and the previews can be viewed after transmission 
through the network. Further detailed inspection can be done by retrieving the full size Quick Look. The Ionia 
‘ 1km” Net-Browser is in demonstration during the symposium at the ESA stand, (Mungo et al„ 1994). 
6 Improving and extending SHARP Level 2 production 
Recognised and consolidated scientific software have been interfaced with the SHARK processing system in 
order to extend the AVHRR geophysical products available to the users. The project is actually in a prototyping 
phase and addresses the following points: cloud detection and classification, atmospheric corrections and Land 
Surface Temperature (LSTj generation. A set of specifications for the operational phase has been generated from 
the tests of the prototypes. Some of these software prototypes are in demonstration during the symposium and 
includes: 
• APOLLO (Saunders and Kriebel, 1988, Gesell et al., 1993) modified to cope with tropical conditions. 
• 5S, (Tanre et al., 1990) for the atmospheric corrections of the visible channels and LOWTRAN 7 (Kneizys et 
al., 1988) for the atmospheric correction of the infrared channels. 
• Land Surface Temperature (LST). (Ottle and Vidal Madjar, 1992) generation given the knowledge of surface 
emissivity. 
Data output described in Figure 2 will be distributed under the SHARP level 2 format Maintaining the coherence 
between formats allow the users to re-use the same software to address the new data products. In any SHARP 
level 2 generation, the classification derived from the APOLLO software will replace the old classification, 
(Muirhead and Malkavi, 1989), in the spare bits of the imagery file. The atmospherically corrected radiometry of 
the five channel will form a product and will be called the SHARP level 2S. LST will be put in band 5 of the 
actual SHARP level 2B for the cloud-free pixels over land where you can actually find the SST over sea. 
6.1 Cloud detection and classification 
Cloud detection, together with classification into more than 21 meaningful classes including land, water, ice, not 
ice, snow, not snow, totally cloudy, cloud free and partially cloudy is performed using an extended version of 
APOLLO (Saunders and Kriebel, 1988, Gesell et al., 1993) modified to cope with tropical conditions. 
The APOLLO software makes use of a battery of 5 threshold algorithms to detect cloud free pixels and 2 
threshold algorithms for fully cloudy pixels. APOLLO was developed for AVHRR over temperate regions. It has 
been adapted to cope with tropical regions (i.e. warm surfaces) over land by adding a test on the brightness 
temperature in channel 4 to recover the pixels considered as cloudy by the visible test whereas this is due to high 
reflective and warm surfaces. The thresholds are automatically set as a function of location (latitude) and time (in 
the year). The implementation of APOLLO within SHARK is fully described by Gesell et al.. (1993). The 
classification into more than 21 meaningful groups of pixels (land/sea, cloud, snow, ice) will be reported in the