UCB > CNR > Center for Forestry > Forestry@Berkeley > April 2001 > Meet the Faculty

April 2001, Volume 2, Issue 1

Interview by John A. Helms
helms@nature.berkeley.edu

Meet the Faculty: Greg Biging and Peng Gong

Two CNR faculty are developing fundamental theory and new applications for forest research and management. Greg Biging is Professor of Biometrics and Peng Gong is Associate Professor of Geographic Information Systems and Remote Sensing. I tracked them down in their offices and am tremendously impressed with the breadth of their work, which is helping make Berkeley an internationally-recognized leader in forest science.

Greg Biging

Greg Biging

obtained his Ph.D. in forest biometrics from the University of Wisconsin in 1978 and has been a member of the Berkeley faculty since then. His early work on modeling the dynamics of growth and yield of forest stands, done with industry sponsorship, contributed to the development of the widely used CACTOS simulator that projects growth of California conifer stands. Greg then developed a simulator called STAG, which was important in enabling CACTOS users to generate realistic stand parameters despite having missing tree data. Based on his work helping to model tree competition within CACTOS, Greg has recently turned his attention to modeling crown shape and taper. He and his former doctoral student Samantha Gill (now at Cal. Poly, San Luis Obispo) have developed time-series, stochastic models that produce realistic measures of the crown architecture of individual trees. Greg and Samantha aim to improve crown models, which are fundamental to understanding stand dynamics, competition, and development of wildlife habitat and understory shading.

Greg's recent research included developing sampling systems for change detection accuracy assessment (with David Colby and Russ Congalton now at New Hampshire), modeling tree mortality (with former Ph.D. student Matthias Dobbertin now at the Swiss Federal Research Institute), comparing two kinds of satellite imagery for mapping cover types (with former Ph.D. student Jesús San Miguel-Ayanz now at the European Space Agency), modeling the effects of wind velocity and slope on flame properties (with former Ph.D. student David Weise now at the Riverside Fire Lab), and testing airborne multispectral videography for assessing the accuracy of wildlife habitat maps (with Ed Murphy of Sierra Pacific Industries and Matthias Dobbertin). Current collaborations include exploring techniques for improved analysis of spatial point patterns with John Battles (ESPM), modeling natural and fire induced tree mortality with David Weise and investigating ways of speeding up forest simulations by clustering input data with Eric Turnblom (now at the University of Washington).

Peng Gong

Peng Gong

another very interesting and extremely busy guy, obtained his Ph.D. in geography in 1990 at the University of Waterloo, Canada. We managed to lure him to the Berkeley faculty in 1995 and his research consists of two parts: First, the broad area of remote sensing image processing, analysis, and application, which is being done with three graduate students and two post-docs. A key theoretical component involves hyperspectral analysis (involving recognition of subtle changes in hundreds of spectral bands that cover the spectral range) and developing radiometric and geometric corrections for recognizing conifers by species, determining spatial structure and phenologicial characteristics of tree canopies, estimating conifer leaf area index, and resolving uncertainties in feature boundaries. Another important application is land use or cover mapping, which can be used for resource inventory and change detection in relation to potential global climate change. And all this leads into expert systems, artificial intelligence, and multi-objective land resource assessment.

The second, new area of Peng's research is in fire mapping. With strong NASA backing, Peng and three post-docs are using 15-years of NOAA AVHRR data to quantify the past occurrence and extent of wildfires in both the U.S. and Canada. This is no small task as it involves developing, calibrating, and validating algorithms capable of processing the enormous data base obtained from daily passes of satellites over North America. These techniques can also be applied for evaluating large-scale changes over time in land use or carbon sinks. On campus, Peng is collaborating with John Radke (Landscape Architecture) and Rick Standiford (ESPM) on fire monitoring, with Ye Qi (ESPM) on carbon balance of China, with Ron Amundsen (ESPM) on evaluating the diversity of soils in the US, with Maggi Kelly (ESPM) on sudden oak death in California, and with Bob Spear (Public Health) on the control of schistosomiasis in China.

Collaboration

The collaborative research being developed by Greg and Peng is especially impressive. They work together daily in a synergistic manner drawing on each other's special competence. They have founded a new field of expertise, which they call photo-ecometrics and define as the science and technology of using remote sensing for the precise measurement of ecological parameters. Using this approach they are developing new theory and analytical tools to support three projects. The first, supported by forest industry, is establishing a new, high resolution method of forest sampling, which is not plot-based but uses 3-D image analysis to provide, with ground truthing, highly accurate measures of tree and stand parameters. The second project, supported by NASA, is a collaborative effort with UC Davis and UC Santa Barbara. This project is evaluating the potential thermal loading of downstream water following flood irrigation of an extensive meadow in the eastern Sierra. Aerial photography, satellite imagery, and hyper-spectral data from a new Earth Observing Hyperion (EO-1) satellite are being used to map vegetative types, establish water table levels, and monitor diurnal surface temperature changes. The third project, using fuzzy set theory, is aimed at developing more accurate classification algorithms of all types of remotely sensed data to better characterize and monitor landscape change.

Internationally, Greg and Peng are collaborating with researchers in Italy, Germany, Spain, Canada, and China. In March they leave for a bilateral meeting with collaborators in Spain and, later in the year, they are off to Argentina as part of a NASA science team collaborating on resource inventory. More details of Greg and Peng's research can be found on their websites at http://nature.berkeley.edu/~gong and http://nature.berkeley.edu/~biging.

UCB > CNR > Center for Forestry > Forestry@Berkeley > April 2001 > Meet the Faculty