Monitoring U.S. forest carbon stocks is critical for natural resource management and national greenhouse gas reporting activities. The USFS Forest Inventory and Analysis (FIA) program 'the largest network of permanent forest inventory plots in the world' covers most U.S. forestlands. However, more than 450,000 km2 of forests in interior Alaska (15% of US forestland) are not included in the FIA program, as these remote regions are difficult and expensive to monitor with standard field methods. Recent and projected future impacts from climate change on forest carbon stocks, composition, and extent have elevated the need to develop new approaches for forest monitoring in Alaska. In particular, airborne remote sensing offers unique advantages for monitoring remote forested regions. In many respects, the methods, logistics, and timeliness of carbon monitoring in Alaska are analogous to ongoing efforts to develop carbon monitoring systems for remote tropical forest regions to Reduce Emissions from Deforestation and forest Degradation and enhancing forest carbon stocks (REDD+).
Here, we propose to develop the first regional estimates of forest carbon stocks for the Tanana Inventory Unit of interior Alaska (146,000 km2). The proposed research leverages a sizable investment ($800k) by the USFS FIA Program in 2014 for new forest inventory plots and airborne data collection with Goddard's LiDAR, Hyperspectral, and Thermal Airborne Imager (G-LiHT; http://gliht.gsfc.nasa.gov). G LiHT is a well-calibrated airborne remote sensing package that is assembled from commercial off-the-shelf (COTS) instruments and a proven track record of timely, free, and open access to both low-and high-level products. The USFS project, a pilot study for LiDAR-assisted forest inventory in interior Alaska, does not provide support for research collaboration between NASA and USFS scientists, data analysis, or methods development. In this project, we will expand the Tanana research activity to 1) collaborate on the experimental design for optimal integration of field and LiDAR data for forest carbon monitoring; 2) compare established model-based and model-assisted methods for estimating forest carbon stocks using both plot and LiDAR information; 3) enhance the inventory activity using individual tree, species composition, and vegetation cover information from the combination of G-LiHT hyperspectral, thermal, and LiDAR sensors; and 4) characterize the impacts of recent fires and risk of future fire-driven carbon losses using the systematic sample of G-LiHT flight lines over ~2.5% of the Tanana region (3800 km2); and 5) develop new, spatially explicit estimates of carbon stocks and uncertainties using Bayesian statistical methods. The main outcomes from this work will be estimates of forest carbon stocks and associated uncertainties for the Tanana Inventory Unit. These estimates provide critical and timely information for resource management, and baseline conditions for the spatial distribution of forest cover and carbon stocks in a region that is rapidly changing from climate warming.
Contact Support to request an email list of project participants.
Project URL(s):
None provided.
Data Products:
Product Title: Maps of carbon stocks with pixel-level carbon estimates and pixel-level uncertainties.
Time Period: July and August of 2014
Description: - Quantify Forest carbon stocks and uncertainties in a region with sparse Ground-based data for inventory and management purposes.
Status: Planned
CMS Science Theme(s): Land Biomass
Keywords: Carbon Stocks (; terrestrial); ; Uncertainties & Standard Errors
Spatial Extent: Tanana Forest Management District of Interior Alaska (Tetlin Wildlife Refuge, Bonanza Creek Experimental Forest, Caribou Poker Creeks Experimental Watersheds, Tanana Valley State Forest, and USFS Tanana Inventory Unit)
Spatial Resolution: 30 m
Temporal Frequency: 1 sampling snapshot
Input Data Products: Airborne Lidar (G-LiHT): ALS, hyperspectral/thermal/downwelling, during July & August 2014, 10,000 square km of G-LiHT flight lines (8% of total area); FIA-like plots (0.006 square km: NPS, DoD, university researchers); tree variables from G-LiHT multi-sensor data (stem density, size distribution, species composition); burn statistics from MODIS and Landsat; Landsat 5 & 7 (NLCD 2001); ASTER v.2 topography
Algorithm/Models Used: Bayesian hierarchical model
Evaluation: Evaluation against sparse ground-only data
Intercomparison Efforts/Gaps: Comparison among ground-only, Bayesian, 2-phase model-based, and 2-stage design-based estimates for the five study areas.
Uncertainty Estimates: Spatially explicit uncertainties at 30m resolution
Uncertainty Categories: ensemble
Application Areas: - MRV; - Forest inventory; - Land management
Relevant Policies/Programs: FIA, FLPMA
Potential Users: USFS in Alaska, NASA CMS and ABoVE science teams
Stakeholders: EPA (Point of Contact: Tom Wirth, Wirth.tom@epa.gov); USFS (Point of Contact: Hans Eric Andersen)
Current Application Readiness Level: 6
Start Application Readiness Level: 6
Target Application Readiness Level: 9
Future Developments: - Publish results by 2016.;
- Provide USFS in Alaska with maps and estimates by the end of 2016.;
- Possible deployment of best technique(s) to four remaining Alaskan USFS inventory units.;
- USFS briefing in April 2015;
- Fire analysis: AGB and composition
- Team meeting in August 2015
Product Title: Statistical estimates of carbon stocks at stratum level.
Time Period: July and August of 2014
Description: Provide statistical estimates of Forest carbon stocks with uncertainties for Comparison purposes.
Status: Planned
CMS Science Theme(s): Land Biomass
Keywords: Carbon Stocks (; terrestrial)
Spatial Extent: Tanana Forest Management District of Interior Alaska (Tetlin Wildlife Refuge, Bonanza Creek Experimental Forest, Caribou Poker Creeks Experimental Watersheds, Tanana Valley State Forest, and USFS Tanana Inventory Unit)
Spatial Resolution: stratum-level
Temporal Frequency: 1 sampling snapshot
Input Data Products: Airborne Lidar (G-LiHT): ALS, hyperspectral/thermal/downwelling, during July & August 2014, 10,000 square km of G-LiHT flight lines (8% of total area); FIA-like plots (0.006 square km: NPS, DoD, university researchers); tree variables from G-LiHT multi-sensor data (stem density, size distribution, species composition); burn statistics from MODIS and Landsat; Landsat 5 & 7 (NLCD 2001); ASTER v.2 topography
Algorithm/Models Used: Two-phase model-based approach (Ståhl et al. 2011) and two-stage design-based model-assisted approach (Gregoire et al. 2011)
Evaluation: Evaluation against sparse ground-only data
Intercomparison Efforts/Gaps: Comparison among ground-only, Bayesian, 2-phase model-based, and 2-stage design-based estimates for the five study areas.
Uncertainty Estimates: 2-Phase Model-Based: Function of sampling variability and model variability; ; 2-Stage Design-Based: Function of sampling variability between sample lines and sampling variability within sample lines, and variability between model estimates and ground estimates
Uncertainty Categories: model-data comparison
Application Areas: - MRV; - Forest inventory; - Land management
Relevant Policies/Programs: FIA, FLPMA
Potential Users: USFS in Alaska, NASA CMS and ABoVE science teams
Stakeholders: EPA (Point of Contact: Tom Wirth, Wirth.tom@epa.gov); USFS (Point of Contact: Hans Eric Andersen)
Current Application Readiness Level: 6
Start Application Readiness Level: 6
Target Application Readiness Level: 9
Future Developments: - Publish results by 2016.;
- Provide USFS in Alaska with maps and estimates by the end of 2016.;
- Possible deployment of best technique(s) to four remaining Alaskan USFS inventory units.;
- USFS briefing in April 2015;
- Fire analysis: AGB and composition
- Team meeting in August 2015
Limitations: - No spatial maps but tables of statistical estimates for two of the three methodological approaches: model-based and design-based model-assisted.
Alonzo, M., Morton, D. C., Cook, B. D., Andersen, H., Babcock, C., Pattison, R. 2017. Patterns of canopy and surface layer consumption in a boreal forest fire from repeat airborne lidar. Environmental Research Letters. 12(6), 065004. DOI: 10.1088/1748-9326/aa6ade
Cahoon, S. M. P., Sullivan, P. F., Brownlee, A. H., Pattison, R. R., Andersen, H., Legner, K., Hollingsworth, T. N. 2018. Contrasting drivers and trends of coniferous and deciduous tree growth in interior Alaska. Ecology. 99(6), 1284-1295. DOI: 10.1002/ecy.2223
Ene, L. T., Gobakken, T., Andersen, H., Naesset, E., Cook, B. D., Morton, D. C., Babcock, C., Nelson, R. 2018. Large-area hybrid estimation of aboveground biomass in interior Alaska using airborne laser scanning data. Remote Sensing of Environment. 204, 741-755. DOI: 10.1016/j.rse.2017.09.027
Heaton, M. J., Datta, A., Finley, A. O., Furrer, R., Guinness, J., Guhaniyogi, R., Gerber, F., Gramacy, R. B., Hammerling, D., Katzfuss, M., Lindgren, F., Nychka, D. W., Sun, F., Zammit-Mangion, A. 2018. A Case Study Competition Among Methods for Analyzing Large Spatial Data. Journal of Agricultural, Biological and Environmental Statistics. 24(3), 398-425. DOI: 10.1007/s13253-018-00348-w
Finley, A. O., Banerjee, S., Zhou, Y., Cook, B. D., Babcock, C. 2017. Joint hierarchical models for sparsely sampled high-dimensional LiDAR and forest variables. Remote Sensing of Environment. 190, 149-161. DOI: 10.1016/j.rse.2016.12.004
Datta, A., Banerjee, S., Finley, A. O., Gelfand, A. E. 2016. On nearest-neighbor Gaussian process models for massive spatial data. WIREs Computational Statistics. 8(5), 162-171. DOI: 10.1002/wics.1383
Salazar, E., Hammerling, D., Wang, X., Sanso, B., Finley, A. O., Mearns, L. O. 2016. Observation-based blended projections from ensembles of regional climate models. Climatic Change. 138(1-2), 55-69. DOI: 10.1007/s10584-016-1722-1
Junttila, V., Finley, A. O., Bradford, J. B., Kauranne, T. 2013. Strategies for minimizing sample size for use in airborne LiDAR-based forest inventory. Forest Ecology and Management. 292, 75-85. DOI: 10.1016/j.foreco.2012.12.019
Babcock, C., Matney, J., Finley, A. O., Weiskittel, A., Cook, B. D. 2013. Multivariate Spatial Regression Models for Predicting Individual Tree Structure Variables Using LiDAR Data. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing. 6(1), 6-14. DOI: 10.1109/JSTARS.2012.2215582
Finley, A. O., Banerjee, S., Cook, B. D., Bradford, J. B. 2013. Hierarchical Bayesian spatial models for predicting multiple forest variables using waveform LiDAR, hyperspectral imagery, and large inventory datasets. International Journal of Applied Earth Observation and Geoinformation. 22, 147-160. DOI: 10.1016/j.jag.2012.04.007
Guhaniyogi, R., Finley, A. O., Banerjee, S., Kobe, R. K. 2013. Modeling Complex Spatial Dependencies: Low-Rank Spatially Varying Cross-Covariances With Application to Soil Nutrient Data. Journal of Agricultural, Biological, and Environmental Statistics. 18(3), 274-298. DOI: 10.1007/s13253-013-0140-3
Babcock, C., Finley, A. O., Cook, B. D., Weiskittel, A., Woodall, C. W. 2016. Modeling forest biomass and growth: Coupling long-term inventory and LiDAR data. Remote Sensing of Environment. 182, 1-12. DOI: 10.1016/j.rse.2016.04.014
Datta, A., Banerjee, S., Finley, A. O., Gelfand, A. E. 2016. Hierarchical Nearest-Neighbor Gaussian Process Models for Large Geostatistical Datasets. Journal of the American Statistical Association. 111(514), 800-812. DOI: 10.1080/01621459.2015.1044091
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]
Large-area inventory of boreal forest carbon stocks in interior Alaska using G-LiHT data and forest inventory plots
-- (Douglas Morton, Bruce Cook, Hans Erik Andersen, Robert Pattison, Ross Nelson, Andrew Finley, Chad Babcock, Lawrence A Corp, Matthew E Fagan, Laura Duncanson)
[abstract]