You are here

Shape selection in Landsat time series

Date: September 01, 2016

Presenting a new methodology for fitting ecologically feasible “shapes” to time series of Landsat imagery for modeling, mapping, and monitoring annual forest disturbance dynamics


Forest impacted by the mountain pine beetle.
Forest impacted by the mountain pine beetle.

Background

Understanding trends in forest disturbance and their effects on forest parameters such as tree canopy cover and biomass is important for carbon assessments, as well as for forest management decisions and scientific investigations across the globe. Data from the Landsat suite of remote sensing satellites offer a historically robust collection of earth observations, which can be used to understand forest dynamics at a variety of spatial and temporal scales.

An automated statistical technique is needed that is not band, index, or even application dependent, which can be run consistently for a wide variety of land cover changes, while maintaining the capability of detecting more subtle changes, which occur in response to periods of prolonged drought and/or insect and disease outbreaks.

Spruce beetle kill in mixed canopy forest.
Spruce beetle kill in mixed canopy forest.

    Research

    We present a new methodology for fitting nonparametric shape-restricted regression splines to time series of Landsat imagery for the purpose of modeling, mapping, and monitoring annual forest disturbance dynamics over nearly three decades. For each pixel and spectral band or index of choice in temporal Landsat data, our method delivers a smoothed rendition of the trajectory constrained to behave in an ecologically sensible manner, reflecting one of seven possible “shapes.” It also provides parameters summarizing the temporal pattern including year(s) of change, magnitude of change, and pre- and post- change rates of growth or recovery.

    Through a case study featuring fire, harvest and bark beetle outbreak, we illustrate how resultant fitted values and parameters can be fed into empirical models to map disturbance causal agent and tree canopy cover changes coincident with disturbance events through time.

    Key Findings

    • Output from the shapes algorithm is providing the driving variables in empirical models of forest disturbance nationally.

    • This output is also showing great promise in models of continuous forest attributes like tree canopy cover and biomass through time.

    • The shapes algorithm has been applied to the conterminous U.S. using annual Landsat data 1984 – 2011, and is available to collaborators

    • Code is provided in the R package ShapeSelectForest on the Comprehensive R Archival Network.

    Featured Publications

    Moisen, Gretchen ; Meyer, Mary C. ; Schroeder, Todd A. ; Liao, Xiyue ; Schleeweis, Karen ; Freeman, Elizabeth ; Toney, Chris , 2016


    Principal Investigators: 
    Forest Service Partners: 
    Karen Schleeweis, RMRS
    Elizabeth Freeman, RMRS
    Chris Toney, RMRS
    External Partners: 
    Mary Meyer, Colorado State University
    Xiyue Lio, Colorado State University
    Todd Schroeder, USGS
    Research Location: 
    Conterminous United States