Integrating Landsat, ICESat and ALOS PALSAR for Regional Scale Vegetation Structure Assessment

Through statewide, national and international programs (i.e., the Statewide Land cover And Trees Study (SLATS), the Terrestrial Ecosystem Research Network (TERN) Auscover project and the Japanese Space Exploration Agency (JAXA) Kyoto and Carbon (K&C) Initiative), statewide to national scale moderate spatial resolution products relating to fractional cover and persistent green cover, vegetation height and structure, and above ground biomass have been generated separately using time-series of Landsat sensor data, ICESat GLAS and ALOS PALSAR data. However, the combination of these data to enhance retrieval of vegetation biophysical properties has been limited (e.g., to mapping of regrowth stage through integration of ALOS PALSAR and Landsat-derived cover estimates. This research therefore investigated the potential benefits of integrating these data through object-based analysis. Focusing on the Injune Landscape Collaborative Project (ILCP) research area in Queensland, co-registered Landsat Foliage Projected Cover (FPC) and ALOS PALSAR L-band HH and HV mosaics were segmented to generate objects with similar radar backscatter and cover characteristics. Within these, height, cover, age class and L-band backscatter characteristics were summarised based on the ICESat and Landsat time-series and ALOS PALSAR datasets. The study established the complimentarity of the datasets (e.g., in terms of retrieving biophysical properties) and, through validation against ground truth data (obtained through field and airborne lidar data collection), potential for significant improvement in the retrieval of extent, height and cover. Future research is focusing on expanding the use of these combined datasets at a regional level with a view to developing operational methods to support policy and decision-making across government and non-government organisations. Source data sets to support this work are available here.