Assessing the accuracy of the MODIS LAI 1-km product in southeastern United States loblolly pine plantations: Accounting for measurement variance from ground to satellite

<p>Leaf area index (LAI), defined here as one-half of the total leaf area per unit ground surface area (Chen, 1996), has been estimated at a global scale from spectral data processed from the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor aboard two NASA EOS-AM spacecraft, Terra (launched in 1999) and Aqua (launched in 2002). The MOD15A2 LAI product is a 1 km global data product composited over an 8-day period and is derived from a three-dimensional radiative transfer model driven by an atmosphere corrected surface reflectance product (MOD09), a land cover product (MOD12) and ancillary information on surface characteristics. The United States Environmental Protection Agency (US EPA) initiated validation research (2002) in the evergreen needle leaf biome, as defined in the MOD12 classification, in a regional study located in the southeastern United States. The validation effort was prompted by the potential use of MODIS LAI inputs into atmospheric deposition and biogenic emission models developed within the US EPA Office of Research and Development. The MODIS LAI validation process involves the creation of a high spatial resolution LAI surface map, which when scaled to the MOD15A2 resolution (1 km) allowed for comparison and analysis with the 1 km MODIS LAI product. Creation of this LAI surface map involved: (1) the collection of in situ LAI measurements via indirect optical measurements, (2) the correlation of land cover specific LAI estimates with spectral values retrieved from high resolution imagery (20 m--30 m), and (3) the aggregation of these 30 m cells to 1 km spatial resolution, matching the resolution of the MODIS product and enabling a comparison of the two LAI values (Morisette et al. 2006). This research assessed the uncertainty associated with the creation of the high-resolution LAI reference map, specifically addressing uncertainty in the indirect in situ optical measurements of LAI and the uncertainty in the land cover classification process. Also addressed was the influence of vegetative understory on satellite-derived vegetation indices from the IKONOS sensor.</p>

Name thesis:degree_name

Doctor of Philosophy

Level thesis:degree_level

Dissertation

Year

2006

Author dc:creator

• Iiames, John Shepherd, Jr.

Contributors dc:contributor

• Russ Congalton

• Environmental Sciences
• Remote Sensing
• Agriculture
• Forestry and Wildlife

Repository record dc:identifier

https://scholars.unh.edu/dissertation/317

OAI identifier oai:identifier

oai:scholars.unh.edu:dissertation-1316