Several studies to date have provided an extensive knowledge base for estimating forest aboveground biomass (AGB) and recent advances in space-based modeling of the 3-D canopy structure, combined with canopy reflectance measured by passive optical sensors and radar backscatter, are providing improved satellite-derived AGB density mapping for large scale carbon monitoring applications. A key limitation in forest AGB estimation from remote sensing, however, is the large uncertainty in forest cover estimates from the coarse-to-medium resolution satellite-derived land cover maps (present resolution is limited to 30-m of the USGS NLCD Program). As part of our CMS Phase II activities, we have demonstrated the use of Landsat-based estimates of Leaf Area Index and ICESat Geoscience Laser Altimeter System (GLAS) derived canopy heights for estimating AGB at a 30-m spatial resolution, which compare relatively well with inventory based plot level estimates. However, uncertainties in forest cover estimates at the Landsat scale result in high uncertainties for AGB estimation, predominantly in heterogeneous forest and urban landscapes. We have successfully tested an approach using a machine learning algorithm and High-Performance-Computing with NAIP air-borne imagery data for mapping tree cover at 1-m over California and Maryland. In a comparison with high resolution LiDAR data available over selected regions in the two states, we found our results to be promising both in terms of accuracy as well as our ability to scale nationally. In this project, we propose to estimate forest cover for the continental US at spatial resolution of 1-m in support of reducing uncertainties in the AGB estimation. The generated 1-m forest cover map will be aggregated to the Landsat spatial grid to demonstrate differences in AGB estimates (pixel-level AGB density, total AGB at aggregated scales like ecoregions and counties) when using a native 30-m forest cover map versus a 30-m map derived from a higher resolution dataset. The process will also be complemented with a LiDAR derived AGB estimate at the 30-m scale to aid in true validation. The proposed work will substantially contribute to filling the gaps in ongoing NASA CMS research and help quantifying the errors and uncertainties in NASA CMS products.
Contact Support to request an email list of project participants.
Project URL(s):
None provided.
Data Products:
Product Title: Aboveground biomass at Landsat scale and Lidar-derived biomass maps.
Time Period: 2000-2012
Description: - Provide aboveground biomass estimates.; - Compare differences between pixel-level AGB density and total AGB at aggregated scales like ecoregions and counties.
Status: Planned
CMS Science Theme(s): Land Biomass
Keywords: Carbon Stocks (; terrestrial)
Spatial Extent: CONUS
Spatial Resolution: 30 m
Temporal Frequency: Yearly
Input Data Products: Landsat Leaf Area Index, ICESat GLAS, NAIP airborne imagery data, G-LiHT lidar
Intercomparison Efforts/Gaps: - Compare between a native 30-m forest cover map versus a 30-m map derived from a higher resolution dataset. Compare with Lidar derived tree cover.
Uncertainty Estimates: AGB uncertainty propagation using monte carlo error propagation model.
Uncertainty Categories: ensemble, model-data
Application Areas: - Forest inventory; - Land management, Fire Management
Relevant Policies/Programs: FIA, National Climate Assessment, IPCC
Potential Users: CMS land biomass product developers, USFS
Stakeholders:
Current Application Readiness Level: 2
Start Application Readiness Level: 1
Target Application Readiness Level: 5
Future Developments: State-wide AGB maps to be ready by 2015 Spring
Limitations: None
Date When Product Available: California AGB map available. Rest of the states: 2015 Spring
Assigned Data Center: ORNL DAAC
Metadata URL(s):
Data Server URL(s):
Archived Data Citation:
Product Title: Tree cover maps.
Time Period: 2010-2012
Description: - Provide tree cover estimate for the continental U.S.; - Reduce uncertainties in the aboveground (AGB) biomass estimation.
Intercomparison Efforts/Gaps: - Compare between a native 30-m forest cover map versus a 30-m map derived from the 1-m tree cover data. Compare with Lidar derived tree cover data.
Uncertainty Estimates: Detailed propagation of error analysis for both field and remote sensing steps. Conditional Random Fields for estimating pixel specific uncertainty in tree cover.
Uncertainty Categories: ensemble, model- data
Application Areas: - Forest inventory; - Land management, Fire Management, Land Cover Change
Relevant Policies/Programs: FIA, National Climate Assessment, IPCC
Potential Users: CMS land biomass product developers, USFS
Stakeholders:
Current Application Readiness Level: 2
Start Application Readiness Level: 1
Target Application Readiness Level: 5
Future Developments: Tree cover map ready by 2015 spring
Limitations: Continuous Yearly Sampling (dependent on aerial NAIP imagery) - epoch level possible
Date When Product Available: California Data already available. Rest of the states: 2015 Spring
Assigned Data Center: ORNL DAAC
Metadata URL(s):
Data Server URL(s):
Archived Data Citation:
Publications:
Basu, S., Ganguly, S., Mukhopadhyay, S., DiBiano, R., Karki, M., Nemani, R. 2015. DeepSat. Proceedings of the 23rd SIGSPATIAL International Conference on Advances in Geographic Information Systems. DOI: 10.1145/2820783.2820816
Basu, S., Karki, M., Ganguly, S., DiBiano, R., Mukhopadhyay, S., Gayaka, S., Kannan, R., Nemani, R. 2016. Learning Sparse Feature Representations Using Probabilistic Quadtrees and Deep Belief Nets. Neural Processing Letters. 45(3), 855-867. DOI: 10.1007/s11063-016-9556-4
Boyda, E., Basu, S., Ganguly, S., Michaelis, A., Mukhopadhyay, S., Nemani, R. R. 2017. Deploying a quantum annealing processor to detect tree cover in aerial imagery of California. PLOS ONE. 12(2), e0172505. DOI: 10.1371/journal.pone.0172505
Choi, S., Kempes, C. P., Park, T., Ganguly, S., Wang, W., Xu, L., Basu, S., Dungan, J. L., Simard, M., Saatchi, S. S., Piao, S., Ni, X., Shi, Y., Cao, C., Nemani, R. R., Knyazikhin, Y., Myneni, R. B. 2016. Application of the metabolic scaling theory and water-energy balance equation to model large-scale patterns of maximum forest canopy height. Global Ecology and Biogeography. 25(12), 1428-1442. DOI: 10.1111/geb.12503
Basu, s., M. Karki, S. Ganguly, R. DiBiano, S. Mukhopadhyay, R. Nemani.2015. Learning Sparse Feature Representations using Probabilistic Quadtrees and Deep Belief Nets, European Symposium on Artificial Neural Networks, ESANN 2015 https://www.elen.ucl.ac.be/esann/proceedings/papers.php?ann=2015
Zhang, G., Ganguly, S., Nemani, R. R., White, M. A., Milesi, C., Hashimoto, H., Wang, W., Saatchi, S., Yu, Y., Myneni, R. B. 2014. Estimation of forest aboveground biomass in California using canopy height and leaf area index estimated from satellite data. Remote Sensing of Environment. 151, 44-56. DOI: 10.1016/j.rse.2014.01.025
2015 NASA Carbon Cycle & Ecosystems Joint Science Workshop Poster(s)
G-LiHT: Multi-Sensor Airborne Image Data from Denali to the Yucatan
-- (Bruce Cook, Lawrence A Corp, Douglas Morton, Joel McCorkel)
[abstract]
[poster]