Comparative analysis of EO-1 ALI and Hyperion, and Landsat ETM+ data for mapping forest crown closure and leaf area index

Abstract

In this study, a comparative analysis of capabilities of three sensors for mapping forest crown closure (CC) and leaf area index (LAI) was conducted. The three sensors are Hyperspectral Imager (Hyperion) and Advanced Land Imager (ALI) onboard EO-1 satellite and Landsat-7 Enhanced Thematic Mapper Plus (ETM+). A total of 38 mixed coniferous forest CC and 38 LAI measurements were collected at Blodgett Forest Research Station, University of California at Berkeley, USA. The analysis method consists of (1) extracting spectral vegetation indices (VIs), spectral texture information and maximum noise fractions (MNFs), (2) establishing multivariate prediction models, (3) predicting and mapping pixel-based CC and LAI values, and (4) validating the mapped CC and LAI results with field validated photo-interpreted CC and LAI values. The experimental results indicate that the Hyperion data are the most effective for mapping forest CC and LAI (CC mapped accuracy (MA) = 76.0%, LAI MA = 74.7%), followed by ALI data (CC MA = 74.5%, LAI MA = 70.7%), with ETM+ data results being least effective (CC MA = 71.1%, LAI MA = 63.4%). This analysis demonstrates that the Hyperion sensor outperforms the other two sensors: ALI and ETM+. This is because of its high spectral resolution with rich subtle spectral information, of its short-wave infrared data for constructing optimal VIs that are slightly affected by the atmosphere, and of its more available MNFs than the other two sensors to be selected for establishing prediction models. Compared to ETM+ data, ALI data are better for mapping forest CC and LAI due to ALI data with more bands and higher signal-to-noise ratios than those of ETM+ data.