B42B-01
A Preliminary Study of Biomass Estimation of Boreal Forest in Alaska Using ALOS PALSAR Polarimetric Images
A reliable forest biomass determination is essential for understanding and modeling ecosystem dynamics, regional and global Carbon-fluxes, and their implications in global climate and environmental changes. The full- (HH+HV+VV+VH) and 2-polarization (HH+HV) information available from the Phased Array type L-band Synthetic Aperture Radar (PALSAR) on the newly launched Japanese satellite, Advanced Land Observation Satellite (ALOS), offers a great opportunity for estimating forest biomass from Space. We investigate ALOS PALSAR application in forest biomass estimation through two approaches. The first involves SAR polarimetry. Multiple scattering within forest, especially within forest canopy causes change of vibration direction in return radar signals, resulting in increase of the proportion of cross-polarization (i.e., HV and VH) components. Consequently, the degree of polarization (the ratio of co-polarized received power-HH+VV-over the total received power) thus decreases with increasing biomass. The second approach involves with polarization phase difference (PPD) and its standard deviation. These two parameters increase with increase in multiple scattering. Therefore, they are also indicators of higher biomass. ALOS PALSAR data sets at test sites at Bonanza Creek Experimental Forest, Alaska, were acquired through the Americas ALOS Data Node (AADN) at Alaska Satellite Facility. Interferometric tools are used in data processing to relate complex images in different polarizations. The resulting interferometric phase images are used for PPD and coherence images for deviations in PPD. Analysis is made through comparison of the interferometric-based patterns of PPD and its standard deviation with ground truth gathered through previous field campaigns.
B42B-02
Mapping Mangrove Canopy Structure Using Lidar, inSAR and Field Data
This paper describes a remote sensing method to map mangrove canopy structure at the landscape scale using synergy between Lidar and inSAR. While the inSAR data allows landscape scale mapping of canopy height, the Lidar data is used with an inversion model to retrieve tree size distribution within the footprint. The tree size distribution is then used as an input to a mangrove ecological model to estimate ecosystem productivity throughout the landscape. The inversion uses mangrove tree architecture and assumes the Lidar measures the canopy surface. This assumption was verified using coincident airborne Lidar and Spaceborne ICEsat/GLAS data in Everglades National Park. A user-friendly web interface is presented to examine inSAR derived canopy height and examine ICEsat/GLAS Lidar waveforms over several mangrove sites around the Caribbean and Gulf of Mexico. The interface also allows the user to interactively analyze the waveforms and estimate local ecosystem productivity with user-defined input parameters on nutrient availability and stressors
B42B-03 INVITED
Theory and Practice in Determining the Long-Term Spatial Productivity of Drylands: A California Blue Oak Case Study
Herbivory, fire, and climatic events such as El Niño-Southern Oscillation (ENSO) and La Niña have been shown to have proximal and evolutionary effects on the dynamics of Dryland fauna, flora, and soils. However, spatially-explicit historical impacts of these climatic events on Dryland ecosystems is not known. Consequently, this paper has the purpose of presenting the theory and practical application for estimating the historical spatial impacts of these climatic events. We hypothesize that if remotely-sensed vegetation indices (VI) are correlated to historical tree ring data and also to functional ecosystem processes, specifically gross primary productivity (GPP) and net ecosystem production (NEP) as measured by eddy covariance flux towers, then VIs can be used to spatially and temporally distribute GPP and NEP within the species- or community-specific land cover extent over the length of the tree ring record of selected Dryland ecosystems. Secondly, the Shuttle Radar Topography Mission (SRTM) digital terrain model (DTM) data has been used to estimate tree height and in conjuction with plant allometric equations: biomass and standing carbon in various forest ecosystems. Tree height data in relation to tree ring age data and fire history can be used to reconstruct the spatial distribution of savanna demographic age structure, predict standing carbon and thus provide a complementary and independent dataset for comparison to DTMs from Multiangle Imaging Spectroradiometer (MISR), Interferometric Synthetic Aperture Radar (IFSAR), and Moderate Resolution Imaging Spectroradiometer (MODIS) derived GPP spatial maps. We developed a database consisting of a dendrochronology record, SRTM data, globa fre history data, Long term Data Record Advanced Very High Resolution Radiometer Normalized Difference Vegetation Index (LTDR AVHRR NDVI, 1981 – 2003), contemporary gridded climate data, National Land Cover Data (NLCD), and short term eddy covariance flux tower data for the California Blue Oak woodland ecosystem to estimate both regional aboveground productivity and past disturbance history relative climate, particularly droughts, for the last 500 years.
B42B-04
Forest Canopy Characterization on the Eastern Shore of Virginia Using SRTM Data.
Accurate land cover classifications are necessary for many environmental studies and models. On complex mosaic landscapes, like that of the eastern shore of Virginia, it is particularly important to be able to accurately represent land cover classes and structures. As part of the study looking at land use change on the eastern shore of Virginia, forest canopy heights and structure were derived for the area using a combination of SRTM, remotely sensed and modeled data. Canopy height was calculated using SRTM and the USGS 2-Arc-Second DEM. Further classification and characterization of the forest canopy was made using a decision tree with additional remotely sensed and modeled data. This simple and quick technique gave a relatively accurate first order characterization of the forest canopy of the eastern shore at a suitable resolution and scale for a regional study of this kind. The results of this forest canopy characterization compare favorably to what is seen on the ground, as well as other datasets for the area. This technique appears to be a simple and effective method for estimating canopy structure with sufficient accuracy to populate models and datasets for regional and large area studies of this kind.
B42B-05
Automatic Tree Crown Delineation Using Discrete Return Lidar and its Application in ICEsat Vegetation Product Validation
The Geoscience Laser Altimeter System (GLAS) has acquired over 250 million individual lidar observations over forest regions globally; an unprecedented dataset of vegetation heights. The vertical extent of waveforms collected by GLAS increases as a function of terrain slope and footprint size (the area on the ground that is illuminated by the laser), but we have demonstrated the ability to retrieve accurate vegetation heights from them. It is infeasible to characterize the elevation of terrain and the crown geometry of trees within the 50-70m diameter GLAS footprint at the level of precision required for our current analyses. In addition, it would be prohibitively expensive to collect the amount of data required to develop a consistent field relationship for forests globally. Airborne discrete return lidar (DRL) has the potential to characterize the vegetation and terrain surfaces within the footprint with a level of detail and precision not available from fieldwork. Furthermore, high density DRL datasets are becoming generally available for forested sites worldwide. Numerous studies have used DRL point cloud data to estimate individual tree parameters successfully. In the GLAS waveform, tree crowns are represented in proportion to their crown area, and therefore estimated forest height is weighted towards those trees with the largest crowns. This assumption is validated by a three dimensional lidar waveform model. To estimate average heights of trees (rather than the average heights of DRL points), we developed a crown delineation algorithm to automatically identify dominant and co-dominant tree crowns from DRL data. We then calculated a crown size weighted average height for 511 plots located at five sites along the west coast of North America from California to Alaska. Parameterizing the existing height estimation equation with the new plot dataset results in very consistent relationships between the DRL forest height and forest height estimated from GLAS. The R2 is 0.64 and RMSE is 6.5m for 511 plots. Keywords: discrete-return lidar, crown delineation, ICEsat, GLAS, forest height, validation
B42B-06
Beyond Potential Vegetation II: Using Repeat Lidar Data on Changes in Vegetation Height to Test Model Predictions of Ecosystem Dynamics
Carbon estimates from terrestrial ecosystem models are limited by large uncertainties in the current state of the land surface, as previous disturbances have important and lasting influences on ecosystem structure and fluxes and can be difficult to detect or assess. Previous studies have illustrated how data on the vertical structure of vegetation from lidar can help to provide needed information on successional status for model initialization and constrain estimates of both carbon stock and fluxes. Here, we illustrate how repeat lidar data on vegetation structure can be used to test model predictions of ecosystem dynamics at a tropical forest site at La Selva, Costa Rica (108259 N, 848009 W). Airborne lidar remote sensing was used to measure spatial heterogeneity in the vertical structure of vegetation in 1998 and 2005. The ecosystem demography model (ED) was used to estimate corresponding patterns of carbon stocks, fluxes, and ecosystem dynamics during the interval. Lidar-initialized ED estimates of changes in maximum canopy height) were comparable to but significantly lower than observed (0.85 +/- 0.9 m observed vs. 0.53 +/- 0.4 m modeled) over the whole domain. Most of the model-data difference was due to growth of primary forest trees that exceeded model estimates (0.44 +/-0.9 m observed vs. 0.04 +/-0.1 m modeled), while the model-data comparison was significantly better over secondary forest areas (1.84 +/- 0.18 m observed vs. 1.71 +/-0.9 m modeled). The results of this study provide a promising illustration of the power of using repeat lidar data on changes in vegetation height to test estimates of ecosystem dynamics from height-structured ecosystem models. Extending these capabilities to regional and global scales will require repeat lidar data sets from space, and the continued development of height-structured ecosystem models.
B42B-07
Optimization of Geoscience Laser Altimeter System Waveform Metrics to Support Vegetation Measurements
The Geoscience Laser Altimeter System (GLAS) has acquired over 250 million individual lidar waveforms over forest regions globally; an unprecedented dataset of vegetation heights. We have demonstrated the ability to retrieve accurate vegetation heights using waveform metrics including vertical extent and transformations of the depth of the waveform's trailing and leading edges. All three indices are highly dependant upon the signal strength and signal-to-noise ratio of the waveform, as the background noise contribution to the waveforms has to be removed before calculation of these waveform metrics. Over the last five years, GLAS has collected data during 12 observation periods using illumination from three different lasers. Power levels of these lasers have varied over time, resulting in variable signal characteristics. To minimize this effect, we optimized a noise effect parameter which varies for each observation period. This parameter is used with the mean and standard deviation of the background noise to create a noise level that is removed from the waveform. The optimization analysis uses a global dataset of waveforms that are near- coincident with waveforms from other time periods; the goal of the optimization is to minimize the difference in vertical extent between these spatially overlapping GLAS shots. Waveforms within 10 m of each other are considered near-coincident; 10 m is small relative to the 70 m diameter of the footprint. Initial optimization results are promising; the relative consistency of the waveform noise characteristics is known and the root mean square error of the difference in vertical extent between overlaps has decreased from 7.6 m to 6.6 m. Further optimization of the noise effect parameter will result in more accurate and precise estimates of canopy height from GLAS.