Relative Greenness Index for assessing curing of grassland fuel

• Published in: Remote sensing of environment 2011-06, Vol.115 (6), p.1456-1463
• Main Authors: Newnham, Glenn J., Verbesselt, Jan, Grant, Ian F., Anderson, Stuart A.J.
• Format: Article
• Language: English
• Subjects: Animal, plant and microbial ecology; Applied geophysics; australia; avhrr data; bidirectional reflectance; Biological and medical sciences; Cover fractions; Curing; derivation; Earth sciences; Earth, ocean, space; Exact sciences and technology; fire; Fire danger; Fuel moisture; Fuels; Fundamental and applied biological sciences. Psychology; General aspects. Techniques; Grassland; Grasslands; Imaging; Internal geophysics; Intervals; MODIS; modis data; moisture; NDVI; ndvi time-series; Regression; Spectroradiometers; Teledetection and vegetation maps; Time series; Vegetation; vegetation index
• ISSN: 0034-4257; 1879-0704
• DOI: 10.1016/j.rse.2011.02.005

Knowledge of the proportion of live and dead herbaceous fuel in grasslands is important in determining fire danger. This paper examines the Relative Greenness approach for quantifying these live and dead proportions. Relative Greenness places the Normalized Difference Vegetation Index ( NDVI) in the context of a time series of measurements. The parameters used to describe the temporal distribution of NDVI and the time interval over which this distribution is assessed impact Relative Greenness and the inferred characteristics of the vegetation. In this paper, the Relative Greenness approach was investigated using different NDVI distribution parameters derived from eight-day composites of surface reflectance from the Moderate Resolution Imaging Spectroradiometer (MODIS). We assessed the accuracy of Relative Greenness for predicting the degree of curing (equivalent to the dead proportion of herbaceous fuel) measured at 25 grassland sites in Australia from 2005 to 2009. Results showed that Relative Greenness explained a greater proportion of the variance and provided a more accurate estimate of the degree of curing than linear regression against NDVI. Relative Greenness was further improved using alternative parameters of the NDVI distribution and by selecting an appropriate time interval over which this distribution was assessed. ► Uses Relative Greenness to predict curing at 25 Australian grassland sites. ► The index was shown to be more accurate than NDVI regression. ► Sensitivity to the length of the time series was investigated. ► Greater than 1.7 years of data was required to improve on NDVI regression. ► The lowest error was achieved using a time series of 6.5 year.