Page Banner

United States Department of Agriculture

Agricultural Research Service

Research Project: USING REMOTE SENSING & MODELING FOR EVALUATING HYDROLOGIC FLUXES, STATES, & CONSTITUENT TRANSPORT PROCESSES WITHIN AGRICULTURAL LANDSCAPES Title: A simple method for retrieving leaf area index from landsat using MODIS LAI products as reference

Authors
item Gao, Feng
item Anderson, Martha
item Kustas, William
item Wang, Yuejie -

Submitted to: Journal of Applied Remote Sensing (JARS)
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: June 4, 2012
Publication Date: July 18, 2012
Citation: Gao, F.N., Anderson, M.C., Kustas, W.P., Wang, Y. 2012. A simple method for retrieving leaf area index from landsat using MODIS LAI products as reference. Journal of Applied Remote Sensing (JARS). 6(1):063554.

Interpretive Summary: Leaf area index (LAI) is a key biophysical parameter in most land surface models. Models that operate at multiple spatial scales require consistent LAI inputs at different spatial resolutions. To make a MODIS-consistent LAI product from Landsat imagery, a simple reference-based regression tree approach was developed. This approach uses pure, high quality LAI retrievals from MODIS as references to develop a regression tree relating these MODIS LAI samples to Landsat surface reflectances. Results show that the approach can produce accurate estimates of LAI from Landsat, as evaluated using field measurements collected in the Soil Moisture Experiment of 2002 (SMEX02), conducted in central Iowa during a period of rapid vegetation growth. LAI maps retrieved from Landsat are consistent with the MODIS estimates when aggregated to coarser scales. This approach demonstrates a simple automated framework for producing MODIS-consistent LAI from Landsat data for modeling the land-surface at different spatial scales.

Technical Abstract: Leaf Area Index (LAI) is a key parameter in most land surface models. Models that operate at multiple spatial scales may require consistent LAI inputs at different spatial resolutions or from different sensors. For example, the Atmosphere-Land Exchange Inverse (ALEXI) model and associated disaggregation algorithm (DisALEXI) uses the MODIS LAI product to model fluxes at regional scales (1-10 km grid resolution), and Landsat-based LAI to disaggregate to field scale (30-m grid). In order to make a MODIS-consistent LAI product from Landsat imagery for this combined scheme, a simple reference-based regression tree approach was developed. This approach uses pure and high quality LAI retrievals from MODIS as references to develop a regression tree relating these MODIS LAI samples to Landsat surface reflectances. Results show that the approach can produce accurate estimates of LAI from Landsat, as evaluated using field measurements collected during the Soil Moisture Experiment of 2002 (SMEX02), conducted in central Iowa during a period of rapid vegetation growth. The coefficient of determination (r2) computed between Landsat retrievals and field measurements is 0.92 at the field scale, with an overall mean bias error (MBE) of -0.12 and mean absolute difference (MAD) of 0.27. The MADs of 0.19 and 0.38 are obtained for low to moderate LAI (0-3) and high LAI (>3), respectively, with some underestimation for the high LAI (MBE=-0.37). The LAI maps retrieved from Landsat are consistent with the MODIS estimates when aggregated to coarser scales. The MADs computed between Landsat- and MODIS-derived LAI ranged from 0.07 to 0.83 for different Landsat dates, with no significant bias comparing to MODIS high quality retrievals. This approach demonstrates a simple framework for producing MODIS-consistent LAI from Landsat data for modeling the land-surface at different spatial scales.

Last Modified: 10/25/2014
Footer Content Back to Top of Page