Luo Jiancheng

Remotely sensed mapping of inland lake area changes in the Tibetan Plateau

Numerous inland lakes of the Tibetan Plateau play sensitive indicators to climate change,and mapping and monitoring these lakes would facilitate our understanding of regional climate and environmental changes.The paper is mainly focused on issues of lake mapping and detecting techniques based on multi-temporal remote sensing imagery,and the differences of lake seasonal and decadal changes were also analyzed.With a series of Landsat MSS/TM/ETM+ imagery in recent 30 years,Tibetan lakes in 1970s,1990s,2000s and 2009 are mapped and their changes are detected temporally and spatially.The lake mapping results showed that Landsat datasets have a good performance on lake area mapping whose lake areas are larger than 0.1km2.There are minor seasonal variations of lake areas from September to December,so that the season could be taken as the best time periods for lake area mapping.The maximum rate of lake area change is less than 2%.Temporally,lakes presented a shrinking trend during 1970s-1990s,while an expanding trend during 1990s-2009,and the total lake area change rate during 1970s-2009 is 27.3%.Spatially,lakes changing rates varied with their spatial locations: lakes in Qiangtang Plateau show dramatic shrinking changes during 1970s-1990s while expanding changes during 1990s-2009;lakes in Selin Co Basin and its surrounding areas show expanding changes in the past 30 years,while lakes in northern valley of the Kangdese and southern Kunlun Mountains remain relatively stable.

Study and implementation of sensor information computation environment

As the technical basis of theory of sensor information computation, the sensor information computation environment is the premise of leading remote sensing spatial information to the area of application and services. Under the direction of sensor information computation theory, this paper proposes the architecture of SINCE. Afterward, the technical policy of computation based on pixel and objects is analyzed, and then the feature of this environment is explained through the example of multi-resolution segmentation. Finally, the current research step and future directions are introduced and summarized.