|
 |
Research Interests
My research interests include two closely related areas: remote sensing of vegetation and ecosystem modeling. Remote sensing of vegetation provides critical input information to the modeling effort. My research goal is to improve our understanding of the roles of forest ecosystems in global carbon cycle using the state-of-the-art digital technology, remote sensing and geographic information systems, and ecological models.
Forest ecosystems offer the potential to offset substantially anthropogenic emissions of carbon dioxide into the atmosphere. A forest landscape is usually a mosaic of stands in a wide spectrum of successional stages (see the picture on the left), which are among the crittical factors that determine the capacity of carbon sequestration by forest ecosystems from the atmosphere. Remote sensing is perhaps the only viable option to monitor forest conditions in a timely and cost efficient manner over large areas. Much success has been achieved in monitoring deforestation in the past due to the dramatic change in surface reflectance, while monitoring the subtle yet important changes of forest growth and succession remains a challenge. Currently I am working on developing algorithms to derive forest successional stages on a stand basis from multitemporal remotely sensed data with the help of forest canopy BRDF models, and forest canopy structures from multiresolution remote sensing with the help of image spatial models.
Though remotely sensed data from space provide timely information on vegetation conditions, it is not enough to assess carbon dynamics of forest ecosystems. I integrate vegetation information from remote sensing with with other ecosystem models through geographic information systems to study forest ecosystem carbon cycle over regional scales. I am working with a hierarchical ecosystem model that emphasizes the roles of successional stages in carbon cycle of forest ecosystems over large areas. The biogeochemical process of carbon cycle at stand scale is simulated by a process-based individual model, and the regional carbon budgets are estimated based on stand age structure and other environmental conditions with the help of geographic information systems.
Selected Recent Publications
See Personal Website @ http://www.unc.edu/~csong
Teaching
Currently, I am teaching an undergraduate course of Introduction to Geographic Information (Geog070) and a graduate cause of Advanced Remote Sensing (Geog178). Geog070 provides a survey of sources of geographic information, including maps, air photos, satellite images, GPS, census information and others. Emphasis is on appropriate use, limitations, quantitative spatial analysis and interpretation of geographic data in physical and human geography applications. Geog178 is focused on advanced image processing techniques and quantitative analysis of remotely sensed images for information extraction of land surface characteristics. The course covers information extraction from spectral, temporal, spatial and directional domains. |
