GEOSCAN Search Results: Fastlink


TitleComparison and evaluation of Medium Resolution Imaging Spectrometer leaf area index products across a range of land use
AuthorCanisius, F; Fernandes, RORCID logo; Chen, JORCID logo
SourceRemote Sensing of Environment vol. 114, 2010 p. 950-960,
Alt SeriesEarth Sciences Sector, Contribution Series 20090005
PublisherElsevier BV
Mediapaper; on-line; digital
File formatpdf
AreaOttawa; Nepean
Lat/Long WENS-75.7667 -75.7500 45.3167 45.3000
Subjectsgeophysics; vegetation; remote sensing; satellite imagery; MERIS; LAI
Illustrationssatellite images; graphs; flow charts
ProgramGroundwater Mapping Program
AbstractLeaf area index (LAI) is a commonly required parameter when modelling land surface fluxes. Satellite based imagers, such as the 300 m full resolution (FR) Medium Spectral Resolution Imaging Spectrometer (MERIS), offer the potential for timely LAI mapping. The availability of multiple MERIS LAI algorithms prompts the need for an evaluation of their performance, especially over a range of land use conditions. Four current methods for deriving LAI from MERIS FR data were compared to estimates from in-situ measurements over a 3 km × 3 km region near Ottawa, Canada. The LAI of deciduous dominant forest stands and corn, soybean and pasture fields was measured in-situ using digital hemispherical photography and processed using the CANEYE software. MERIS LAI estimates were derived using the MERIS Top of Atmosphere (TOA) algorithm, MERIS Top of Canopy (TOC) algorithm, the Canada Centre for Remote Sensing (CCRS) Empirical algorithm and the University of Toronto (UofT) GLOBCARBON algorithm. Results show that TOA and TOC LAI estimates were nearly identical (R2 > 0.98) with underestimation of LAI when it is larger than 4 and overestimation when smaller than 2 over the study region. The UofT and CCRS LAI estimates had root mean square errors over 1.4 units with large (not, vert, similar 25%) relative residuals over forests and consistent underestimates over corn fields. Both algorithms were correlated (R2 > 0.8) possibly due to their use of the same spectral bands derived vegetation index for retrieving LAI. LAI time series from TOA, TOC and CCRS algorithms showed smooth growth trajectories however similar errors were found when the values were compared with the in-situ LAI. In summary, none of the MERIS LAI algorithms currently meet performance requirements from the Global Climate Observing System.

Date modified: