Healey, Sean: USDA Forest Service (Project Lead)
Project Funding:
2011 - 2012
NRA: 2010 NASA: Science Definition Team for Carbon Monitoring System
Funded by NASA
Abstract:
Sean Healey is proposed as a member of the Science Definition Team (SDT) of NASA's Carbon Monitoring System (CMS). He and the listed co-investigators are all affiliated with the Forest Service's FIA (Forest Inventory and Analysis) program, and as a group, they have extensive experience using FIA data both to develop official forest statistics and to calibrate and validate maps created with remotely sensed data. This proposal will add to the SDT a practical element related to the limits of FIA data and the data needs of the forest management community. In addition, independent research is proposed to leverage the properties of FIA's sample to better understand how CMS might perform both beyond the borders of the United States and in the DESDynI era.
The optical satellite imagery which will form an important part of any near-term global CMS biomass product often does not offer good resolution of moderate and high levels of forest biomass. In the absence of predictors able to discriminate among levels of a particular target variable, many modeling approaches minimize prediction error by predicting toward the mean of the reference dataset. If that reference dataset is biased (not a representative sample), biomass predictions can be systematically skewed either up or down. While FIA in this country does comprise a representative sample, CMS in many parts of the globe will have to rely upon ad hoc collections of management inventory stand exams which are often skewed toward harvestable (high biomass) conditions. Over large areas, the potential for even small systematic prediction bias may create very large errors in carbon storage estimates.
We will use intentionally biased sub-samples of FIA data from the state of Oregon as reference data to test the effect of such bias upon state-level CMS estimates of biomass. In addition, using these same biased sub-samples, we will replace optical data in the CMS system with pseudo-data representing the higher correlation with biomass anticipated with DESDynI-based predictors. DESDynl's increased prediction precision may reduce the rate at which predictions default toward the mean and may therefore reduce propagation of ground sample bias in the CMS mapping process. These activities should shed light upon: 1) the likely effects of non-representative reference data on the global CMS biomass product, and 2) the degree to which DESDynI may diminish prediction error related to ground sample bias.
Publications:
Healey, S. P., Patterson, P. L., Saatchi, S., Lefsky, M. A., Lister, A. J., Freeman, E. A. 2012. A sample design for globally consistent biomass estimation using lidar data from the Geoscience Laser Altimeter System (GLAS). Carbon Balance and Management. 7(1). DOI: 10.1186/1750-0680-7-10
Healey SP, Patterson PL, Saatchi S, Lefsky MA, Lister AJ, Freeman EA, Moisen GG. (2012). Applying inventory methods to estimate aboveground biomass from satellite light detection and ranging (LiDAR) forest height data. In: RS Morin (ED) Moving from status to trends: Forest Inventory and Analysis (FIA) symposium 2012; Gen. Tech. Rep. NRS-P-105. Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. pp. 404-409. (http://www.nrs.fs.fed.us/pubs/gtr/gtr-nrs-p-105papers/67healey-p-105.pdf)
2013 NASA Terrestrial Ecology Science Team Meeting Poster(s)
- The Global Forest Biomass Inventory
-- (Sean P Healey, Erik Lindquist, Paul Patterson, Sassan Saatchi, Michael Lefsky, Michael Hernandez, Alicia Peduzzi)
[abstract]
2011 NASA Carbon Cycle & Ecosystems Joint Science Workshop Poster(s)
- Using Sample Theory to Develop Forest Biomass Estimates from ICESat/GLAS Data
-- (Sean P Healey, Paul Patterson, Sassan Saatchi, Michael Lefsky, Andrew Lister)
[abstract]
More details may be found in the following project profile(s):
