Tree tree crown mortality associated with roads in the Lake Tahoe Basin: a remote sensing approach

YUANCHAO FAN, University of Nevada - Reno, Department of Natural Resources and Environmental Science

DR. PETER WEISBERG, Assistant Professor of Landscape Ecology, University of Nevada - Reno, Department of Natural Resources and Environmental Science

Abstract:

Forest degradation is an increasingly severe environmental problem throughout the world. The main causes include diseases, insects, drought, wildfire, climate change, and human disturbances. As for the last factor, the expanding road network and traffic within many forest areas is of significant impact on the forest ecosystem. Issues of road effect on forest health have been drawing attention since de-icing compounds such as sodium chloride (NaCl) have been observed to damage plants through both soil-mediated and airborne effects. For government agencies to assess the potential effects of de-icing salts and take corresponding mitigation strategies, it is necessary to quantify de-icing salt damage and separate it from damages caused by other factors, and to monitor forest changes before and after mitigation efforts.

Remote sensing technologies provide a means for understanding potential road-related effects on tree crown mortality in a large-extent and long-term context. Tree crown mortality here refers to loss of photosynthetic material in tree crowns, and does not necessarily imply tree death. Remote sensing has the ability to generalize limited findings from field-based forest mortality surveys to a much larger scale. Archival satellite imagery (e.g. Landsat TM, IKONOS) allows historical analysis of crown mortality due to road-related and non-road-related factors, and various sources of future imagery provide an inexpensive, efficient and repeatable protocol for future monitoring.

I have developed a remote sensing and GIS database containing Landsat TM5 images of each year from 1985 to 2008, IKONOS images of 2001, 2002, 2003, 2005, 2006 and 2009 for late summer or early fall, and other required GIS data such as DEM, DRG, vegetation map, road map, snowfall data, and de-icing salt application data. Major remote sensing work has been finished, including image preprocessing (geometric correction, atmospheric correction and topographic correction) and change detection (NDVI/MSAVI Differencing, Multi-temporal Kauth-Thomas transformation (MKT), and Post-Classification Comparison). Ground reference data for calibration and validation were collected in summer 2008, 2009 and fall 2010. In addition, two related projects from Dr. Robert Nowak's lab: one ongoing project “Tree health within the Lake Tahoe Basin: interactions with fuel management treatments”; another accomplished project “Effects of de-icing salts on vegetation in the Lake Tahoe Basin” (2006 and 2007) provide additional field data for my remote sensing study.

This project is at its final stage of statistical analyses in order to isolate road-related effects on tree crown mortality from non-road-related effects and to analyze the mechanisms of spatial and temporal variations of tree crown mortality associated with roads at different scales. Final results will come out and be summarized in an NDOT report and my thesis by this summer.

Funding provided by: Nevada Division of Transportation

Project duration: 2009-2012