The global network FLUXNET supplies environmental scientists with valuable data on ecosystem exchange processes along with meteorological measurements. Ecosystem characteristics at these sites can be efficiently complemented by remote sensing data from the MODIS sensors on-board the NASA satellites Terra and Aqua. The ORNL DAAC makes resampled MODIS key products as so called 'Land Product Subsets' available. These subsets comprise selected MODIS products in a 7x7 km grid centered on FLUXNET sites. One of these products is the leaf area index (LAI). Despite the frequent application of MODIS LAI data in ecosystem models, there is still no consensus on its usage and the employment of the additionally provided quality criteria (QC). In this study, we analyze the effects of various QC filters, spatial aggregations and sensor choices on magnitude and temporal dynamics of LAI data at six FLUXNET sites. Additionally, we assess the sensitivity of a simple soil-vegetation-atmosphere-transfer (SVAT) model on differently post-processed LAI times series. It is found that it is advantageous to combine the products of both sensors. The consideration of quality assessments is essential, but the QC application is not straightforward for forest sites and the QC choice can have significant effects on the resulting LAI time series with considerable consequences on the outcome of subsequently applied SVAT models.
"Post-processing analysis of MODIS leaf area index subsets," Journal of Applied Remote Sensing 4(1), 043557 (1 November 2010). https://doi.org/10.1117/1.3524265