AGB

Aboveground biomass (AGB) mapping and monitoring has been CAFRI’s flagship project to-date, with the goal of eventually developing spatially explicit carbon monitoring, reporting, and verification information for New York State DEC. AGB has been our primary target, despite carbon information being our eventual goal, since AGB is observable from aerial/spaceborne remote sensing platforms and is directly related to aboveground carbon and belowground carbon. We have been tasked with historical mapping (annually, back to 1990), continuous annual monitoring, and forecasting for the entirety of New York State.

LiDAR-AGB

This sub-project aimed to use the best available remotely sensed data (LiDAR) to provide an accurate snapshot of current AGB conditions in New York State. Our efforts to-date are encapsulated in a preprint, now published at the International Journal of Applied Earth Observation and Geoinformation. We plan to update our models maps with new LiDAR data as it becomes available.

Landsat-AGB

Landsat offers complete statewide coverage at 30m resolution, with annual timesteps, and historical records dating back to 1984, making it a prime candidate for our historical mapping and ongoing monitoring efforts. Despite all of these benefits, optical imagery is provides less information about forest structure than LiDAR or other remotely sensed datasets derived from active sensors, and as a result the accuracy of Landsat-derived AGB predictions suffers. We have attempted two major modeling approaches (described below), with one set of models trained directly on Forest Inventory and Analysis (FIA) measurements, and the latter trained on predictions from our LiDAR-AGB models (see above). Finally, we have ensembled the winning models from these two component approaches.

Forecasting

Starting from the 30-year time series produced by Landsat-AGB, we’re fitting phenomenological time-series models (namely, damped trend ETS and simple linear regressions against historical trends) to every single 30m pixel in New York State and forecasting the period from 2020-2050.

People

• Lucas Johnson
• Mike Mahoney
• Sam Gordon
• Colin Beier

Key Papers

• Huang, Wenli, et al. “High-resolution mapping of aboveground biomass for forest carbon monitoring system in the Tri-State region of Maryland, Pennsylvania and Delaware, USA.” Environmental Research Letters 14.9 (2019): 095002. doi: 10.1088/1748-9326/ab2917
• Hudak, Andrew T., et al. “A carbon monitoring system for mapping regional, annual aboveground biomass across the northwestern USA.” Environmental Research Letters 15.9 (2020): 095003. doi: 10.1088/1748-9326/ab93f9
• Hawbaker, Todd J., et al. “Light detection and ranging-based measures of mixed hardwood forest structure.” Forest science 56.3 (2010): 313-326. doi: 10.1093/forestscience/56.3.313
• Matasci, Giona, et al. “Large-area mapping of Canadian boreal forest cover, height, biomass and other structural attributes using Landsat composites and lidar plots.” Remote sensing of environment 209 (2018): 90-106. doi: 10.1016/j.rse.2017.12.020

Links

• LiDAR-AGB manuscript repo (paper and analysis)
• LiDAR-AGB repo
• Landsat-FIA AGB repo
• Landsat-LiDAR AGB repo
• Landsat-FIAxLiDAR AGB ensemble repo
• Forecasting repo

Outputs

• LiDAR-AGB maps can be downloaded from labrador, with product name "lidar_agb".

• Landsat-AGB maps can be downloaded from labrador with the folliwng product names:

  • "landsat_fia_agb_<year>"
  • "landsat_lidar_agb_<year>"
  • "landsat_ensemble_agb_<year>"

Blog Posts