Retreat of glaciers due to melting of ice often leads to formation of new glacial lakes and/or expansion of existing ones. this effect is more pronounced and severe in alpine conditions/regions of himalaya. due to increased rate of snow melt, water accumulation in these lakes has been rising rapidly. burst or sudden discharge of large volumes of water along with debris from these lakes is the reason behind glacial lake outburst floods (glofs) in downstream valley resulting in massive damage to infrastructure (dams, hydropower plants, roads, residential areas, farms etc), natural resources (forests, fauna) and human life. therefore monitoring of these kinds of lakes in terms of their formation, expansion and vulnerability is very important and towards this the earth observation (eo) based techniques can be very useful. the present study is focused on accurate estimation of the volume of glacial lakes in sikkim, india as an aid to simulation of dam breach conditions to generate a flood risk map. traditional approaches in this regard in himalayan regions comprise of use of empirical models to derive lake volume. this technique has some inherent flaw in terms of the achievable accuracy. in the present study an in-house designed, compact and handy robotic-sonar system is developed to obtain accurate depth of lakes in an automated manner. high altitude (~16,500ft), difficult terrain and extreme environmental conditions make the lakes completely inaccessible for on board bathymetry survey. at the same time robotic rover eliminates the chances of human injuries and fatalities which cannot be ignored in case of on board survey. bathymetry data of lake from robotic-sonar system is used with remotely sensed multispectral datasets to derive glacial lake depth. as a result, the output is more accurate compared with traditional empirical models. once the bathymetry data have been recorded for the given lake and the volume is computed. dam breach conditions can then be simulated for disaster risk management, integrating volume information, high resolution dem and moraine constituent data form field survey.