Integration of ALS and TLS for calibration and validation of LAI profiles from large footprint lidar
Abstract
The Global Ecosystem Dynamics Investigation (GEDI) is designed to provide measurements of forest vertical structure and above-ground biomass density (AGBD) over tropical and temperate regions. The GEDI is a multi-beam waveform lidar that will acquire transects of forest canopy vertical profiles in conditions of up to 99% canopy cover. These are used to produce a number of canopy height and profile metrics to model habitat suitability and AGBD. These metrics include vertical leaf area index (LAI) profiles, which require some pre-launch refinement of large-footprint waveform processing methods for separating canopy and ground returns and estimation of their reflectance. Previous research developments in modelling canopy gap probability to derive canopy and ground reflectance from waveforms have primarily used data from small-footprint instruments, however development of a generalized spatial model with uncertainty will be useful for interpreting and modelling waveforms from large-footprint instruments such as the NASA Land Vegetation and Ice Sensor (LVIS) with a view to implementation for GEDI. Here we present an analysis of waveform lidar data from the NASA Land Vegetation and Ice Sensor (LVIS), which were acquired in Gabon in February 2016 to support the NASA/ESA AfriSAR campaign. AfriSAR presents a unique opportunity to test refined methods for retrieval of LAI profiles in high above-ground biomass rainforests (up to 600 Mg/ha) with dense canopies (>90% cover), where the greatest uncertainty exists. Airborne and Terrestrial Laser Scanning data (TLS) were also collected, enabling quantification of algorithm performance in plots of dense canopy cover. Refinement of canopy gap probability and LAI profile modelling from large-footprint lidar was based on solving for canopy and ground reflectance parameters spatially by penalized least-squares. The sensitivities of retrieved cover and LAI profiles to variation in canopy and ground reflectance showed improvement compared to assuming a constant ratio. We evaluated the use of spatially proximate simple waveforms to interpret more complex waveforms with poor separation of canopy and ground returns. This work has direct implications for GEDI algorithm refinement.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2016
- Bibcode:
- 2016AGUFM.B54B..04A
- Keywords:
-
- 0414 Biogeochemical cycles;
- processes;
- and modeling;
- BIOGEOSCIENCESDE: 0428 Carbon cycling;
- BIOGEOSCIENCESDE: 0439 Ecosystems;
- structure and dynamics;
- BIOGEOSCIENCESDE: 0480 Remote sensing;
- BIOGEOSCIENCES