Authors: Xianfeng Chen*, Slippery Rock University, Jack Livingston, Slippery Rock University, Nickolas Orbask, Slippery Rock University
Topics: Remote Sensing, China, Biogeography
Keywords: spectral unmixing, Landsat 8 OLI, desert vegetation
Session Type: Paper
Start / End Time: 8:00 AM / 9:40 AM
Room: Lafayette, Marriott, River Tower Elevators, 41st Floor
Presentation File: No File Uploaded
Arid and semi-arid regions account for over one third of global land surface. Although vegetation is very it sparse and less diverse than its counterpart in tropic climate region, it plays an essential role in the physiological and biogeochemical function of arid and semi-arid ecosystem. Monitoring of both photosynthetic vegetation (PV) and non-photosynthetic vegetation (NPV) is needed to gain insight into disturbance events such as overgrazing and desertification, and climate change. The study area, Manashi river watershed located at Xinjiang, China, is extended from the northern slopes of Tian Shan Mountain to Junger basin. Spectral mixture analysis is applied to Landsite 8 OLI (Operational Land Imager) data to estimate PV, NPV, and bare soil covers. The endmember spectra were collected from both image and field. The study systematically compared these two different approaches to collect endmembers for quantifying PV, NPV and bare soil covers using spectral mixture models. The accuracy of spectral mixture analysis in desert scrubland and dry grassland of the study area was tested by ground truth data. The field sampling strategy using GPS and high resolution digital camera was adopted for this study. Two scenes of high spatial resolution satellite images, QuickBird and Worldview-2, were also used for evaluation of spectral mixture analysis. The results indicate that spectral mixture analysis provides a promise tool to quantify fractional covers of PV, NPV, and bare soil.