Assessment of Vegetation Stress Using Reflectance or Fluorescence Measurements
- P. K. E. Campbell *ab,
- E. M. Middletonc,
- J. E. McMurtreyd,
- L. A. Corpe and
- E. W. Chappellec
- a Joint Center for Earth Systems Technology, Univ. of Maryland, Baltimore County (UMBC), Baltimore, MD 20771, USA
b (current address), Biospheric Sciences Branch, Code 614.4, NASA/Goddard Space Flight Center, Greenbelt, MD 20771 USA
c Biospheric Sciences Branch, Code 614.4, NASA/Goddard Space Flight Center, Greenbelt, MD 20771 USA
d Hydrology and Remote Sensing Lab., Agricultural Research Service, USDA, Beltsville, MD 20705 USA
e Science Systems and Applications Inc. (SSAI), Lanham, MD 20706 USA
Current methods for large-scale vegetation monitoring rely on multispectral remote sensing, which has serious limitation for the detection of vegetation stress. To contribute to the establishment of a generalized spectral approach for vegetation stress detection, this study compares the ability of high-spectral-resolution reflectance (R) and fluorescence (F) foliar measurements to detect vegetation changes associated with common environmental factors affecting plant growth and productivity. To obtain a spectral dataset from a broad range of species and stress conditions, plant material from three experiments was examined, including (i) corn, nitrogen (N) deficiency/excess; (ii) soybean, elevated carbon dioxide, and ozone levels; and (iii) red maple, augmented ultraviolet irradiation. Fluorescence and R spectra (400–800 nm) were measured on the same foliar samples in conjunction with photosynthetic pigments, carbon, and N content. For separation of a wide range of treatment levels, hyperspectral (5–10 nm) R indices were superior compared with F or broadband R indices, with the derivative parameters providing optimal results. For the detection of changes in vegetation physiology, hyperspectral indices can provide a significant improvement over broadband indices. The relationship of treatment levels to R was linear, whereas that to F was curvilinear. Using reflectance measurements, it was not possible to identify the unstressed vegetation condition, which was accomplished in all three experiments using F indices. Large-scale monitoring of vegetation condition and the detection of vegetation stress could be improved by using hyperspectral R and F information, a possible strategy for future remote sensing missions.Please view the pdf by using the Full Text (PDF) link under 'View' to the left.
Copyright © 2007.