The cost of Bangladesh is extremely exposed to tropical cyclones because of its long costal line containing shallow deltas and densely populated offshore islands. This entails that tropical cyclones will have a particularly severe impact causing high fatality mostly among the poor people on the costal region. In spite of regular attack of devastating cyclones in Bangladesh coast, it-s impact on human lives as well as infrastructure and natural resources have been reduced considerably. The damage reduction reached at a level which is now difficult to down further because of two reasons. First is the insufficient information for early preparedness and second is the ineffective after-cyclone damage restoration system. It is therefore especially important in this part of the world that cyclones can be predicted well in advance before landfall. Normally, cyclone tracking and forecasting is based on satellite images supplemented with data collected using floating buoys, naval ships, dropsondes, airplanes carrying a wide range of active and passive sensors. In Bangladesh however, there is a lack of supplementary data. Meteorologists in Bangladesh are therefore forced to rely on satellite images only (NOAA-AVHRR). The aim of this project is to assess the informational contents of satellite images and identify their limitations in forecasting tropical cyclones. A first step in the project is to assess to what extent current models used for forecasting around the world are dependent on supplementary data, and on the other hand to what extent they could be used to predict tropical cyclones using satellite images only.