Abstract:

Fractional Vegetation Cover (FVC) is an important ecology parameter, which is essential in the studies of hydrology, ecology and global variation. The results of the estimation methods used at present for FVC, including eyeballing method and digital camera imagery interpretation method, are obviously subjective and uncertain. Furthermore, it is rather difficult for the statistical relationship between FVC and leaf area index (LAI) to establish by measurement of millions of samples that have similar vegetation structure parameters. So it is an effective way for the estimation of FVC to develop a statistical model between FVC and LAI. In the paper, we took soybean for example, simulated the soybean leaves by using ellipses, and determined the position, orientation and size with random distribution function by choosing the key parameters in the soybean structure to obtain the FVC per area under different vegetation structure parameters. The model was validated by data measured in situ and the theoretical conclusion. The parameter sensitivity of the simulation model was analyzed, the result showed that (1) the length of stem is not a sensitive parameter when it is longer than foliage interval while the angle of stem is not a sensitive parameter until it reaches a threshold; (2) leaf tilt angle and foliage interval is in inverse portion to the fractional vegetation cover while the semimajor and semiminor axis of leaf is in direct portion to the fractional vegetation cover; (3) semiminor axis of leaf is a much more sensitive parameter than semimajor axis of leaf. It suggested that it was a novel and feasible way for FVC.

Key words: Fractional vegetation cover, Computer simulation model, Parameters sensitivity, Soybean