Authored By: G. Moisen, R. L. Czaplewski, K. Brewer, S. Healey
Gretchen Moisen, Raymond L. Czaplewski, Ken Brewer, Sean Healey
USDA Forest Service Rocky Mountain Research Station (1,2,4) and Remote Sensing Applications Center (3)
Continuous improvement in risk assessment requires monitoring to directly detect and assess the extent and severity of realized disturbances relative to the predicted risk of those disturbances. The Forest Inventory and Analysis (FIA) program, conducted by the Research and Development branch of the USDA Forest Service, is well suited for monitoring all US forestlands. FIA measures over 160,000 forested field plots that are systematically distributed approximately every 5-km (1 plot per 2,400-ha). Plots uniformly cover all public and private lands. This sampling intensity often permits statistically reliable assessments for analysis-areas as small as 1-million acres. Tree- and plot-level conditions are carefully measured in the field within each of four 0.017-ha sub-plots that cover a 0.4-ha field plot. Seven to 10 years are required to completely re-measure all 160,000 field plots in the USA. These spatial and temporal scales are sufficient to monitor slowly-evolving disturbances that are finely-distributed over large areas (e.g., climate change). However, this standard FIA protocol is not well suited for monitoring rapid and coarse-grained changes, such as wildfires, hurricanes and ice storms. Using a combination of remote sensing technologies and an augmented sampling frame, FIA is developing the capability to rapidly detect and assess changes in US forests. Within a few days after a disturbance, areas impacted by a disturbance are classified into different severity levels with coarse-scale remote sensing (e.g., MODIS, Doppler radar) or risk assessments. Estimates of forest resources potentially affected within each severity class are produced using existing FIA plot data. Within a few months, improved estimates of forest resources likely affected are obtained by applying disturbance-specific models (built with historic disturbance data) to finer resolution imagery. Ultimately, direct measurements of forest conditions using multi-stage and/or multi-phase sampling with large-scale aerial photography and/or field crews will yield the best estimates of forest resources actually affected, but will require a new post-disturbance data collection effort. This rapid response system might be relevant for monitoring invasive species, pests, diseases and conversion in land uses.
Monitoring Methods Session - Tuesday Afternoon
corresonding author:
Raymond L. Czaplewski
USDA Forest Service
Rocky Mountain Research Station
2150 Centre Avenue, Bldg. A
Fort Collins, CO 80526
970-295-5973
rczaplewski@fs.fed.us
note: oral presentation only