Los Alamos National Laboratory  
GIS Colloquium

Home
About GISLab
Research
Publications
Presentations
Projects
GIS Services
Spatial Data Warehouse
GIS FAQs
Policies
Guides & Stds
Links
Team Page
What's New
 

Tuesday 11/22/2005 at 2 - 3 PM
EES-DO/ISR-3 Conference Room (TA3, SM40, Rm 275)

Multispectral Thermal Imagery and its Application to the Geologic Mapping of the Koobi Fora Formation, Northwestern Kenya

Mary K. Greene
LANL GISLab, EES-9

Host: Paul Rich, Environmental Geology and Spatial Analysis (EES-9). Contact Paul Rich (pmr@lanl.gov) if you wish to meet with the speaker during the day to discuss ideas for collaboration.

NOTES: Biography: Mary Greene recently joined GISLab (EES-9), where she works with GIS services as a cartographer and GIS technician. Prior to her employment with EES-9, she worked in ISR-2 as a student (mentor: Paul Pope) to complete the research detailed in this talk. During her studies she was employed as a computer technician with CCN, specializing in large data set visualization and Windows security. Mary earned her B.S. in Geology and B.S. in Physical Anthropology with honors from Beloit College (1995), where she focused her senior thesis on geophysics applied to archaeology in Greece. She will receive her M.S. with honors in Geology at the University of New Mexico in December 2005. Mary's interests include geologic applications of remote sensing, epidemiological applications of GIS, cartographic methods, and advanced visualization.


Abstract:
The Koobi Fora Formation in northwestern Kenya has yielded more hominin fossils dated between 2.1 and 1.2 Ma than any other location on Earth. This research was undertaken to discover the spectral signatures of a portion of the Koobi Fora Formation using imagery from the DOE's Multispectral Thermal Imager (MTI) satellite, and to create a digital geologic map of the region based on those signatures. MTI is unique amongst multispectral satellites in that it co-collects data from 15 spectral bands ranging from the visible to the thermal infrared with a ground sample distance of 5 meters per pixel in the visible and 20 meters in the infrared. A total of four unique combinations of geologic classes were analyzed using a linear spectral mixing algorithm. The tuffs within the Koobi Fora Formation were defined with 100% accuracy using a combination of pure spectra from the basalt, vegetation, and tuff geologic classes.


 


Contact Us | Webmaster | EES-9 | EES Division | SR |
  Lanl logo  

Operated by the University of California for the National Nuclear Security Administration, of the US Department of Energy. | Copyright © 2003 UC | Disclaimer/Privacy |

Last Revised:  18-Nov-2003

spacer spacer spacer spacer