Investigators:
Ram M. Narayanan
Grants:
"Airborne Remote Sensing for Forest Monitoring Applications", DOE,
Description:
We propose to develop and refine airborne remote sensing techniques for forest monitoring applications. This proposal combines some unique strengths and resources resident at the partner institutions, and fosters new research partnerships.We are proposing to develop algorithms for inversion of forest parameters, such as leaf area index (LAI), above ground biomass, vegetation height profile, species classification and identification, and vegetation stress using combined laser reflectance data at green (532 nm) and near-infrared (1064 nm) wavelengths, and radar reflectance at 10 GHz frequency. Prior research has shown that while laser reflectance at near-IR can be used to estimate leaf cover characteristics, green reflectance can be used to detect stressed vegetation. On the other hand, radar has good sensitivity to LAI and biomass if the incidence angle is appropriately chosen. Thus, these two sensors provide complementary data. The inversion models will be developed based upon a comprehensive set of concurrent airborne remote sensing and ground data measurements. Both theoretical and empirical models will be developed over a wide range of parameter dynamic range, and the inversion accuracy assessed. As part of the data analysis, we also propose to develop calibration techniques since calibration error will directly impact inversion accuracy. Since our sensors have polarimetric as well as ranging capability, we believe we will be able to develop techniques for better species classification (based on the depolarization characteristics) and for vegetation height profiling.
The initial airborne measurements will be performed at the University of Nebraska test site during the first year, and the algorithms will be developed and refined based on collaboration with scientists at the Brookhaven National Laboratory and Duke University. During the second and third years, the measurements are planned at the Duke Forest, which will enable us to establish the relationship between relevant forest parameters and multisensor reflectance data over a wide range and climatic conditions. In addition, the effects of spatial variability, so often neglected, will be investigated, as well as the confounding effects of other parameters, such as underlying soil. Duke University will provide us with concurrent ground measurements which we will use to refine our models.
At the conclusion of the 3-year research project, we hope to have developed an operational tool for routine monitoring of forests and forested areas.