Characterizing Bark Beetle Early Attack Using Landsat Analysis Ready Data (ARD) in the Pine Forest of Northern Colorado, USA

Authors: Su Ye*, Graduate School of Geography, Clark University, John Rogan, Graduate School of Geography, Clark University
Topics: Remote Sensing, Land Use and Land Cover Change
Keywords: Time series analysis, Continuous Change Detection and Classification, Landsat Analysis Ready Data, insect disturbance
Session Type: Paper
Day: 4/7/2019
Start / End Time: 8:00 AM / 9:40 AM
Room: Buchanan, Marriott, Mezzanine Level
Presentation File: No File Uploaded

In the last two decades, bark beetle outbreaks, linked to the recent enhanced temperature trends and prolonged drought, have caused an unprecedented level of tree mortality in conifer forest across the Western United States. Multiple synchronous bark beetle outbreaks have impacted the ability of these forests to provide ecosystem services such as carbon storage and timber production. The efficacy of management practices are contingent upon an early detection of beetle attack, where infested trees containing the next generation of beetles can be removed before brood emergence. However, systematic monitoring and assessment of gradual forest change such as forest health decline has been rarely pursued at landscape scales using satellite sensors. This paper addresses detection of beetle-related early-stage needle change for the recent bark beetle outbreaks in the Northern Colorado (1999 – 2018). The work proposes a new time-series approach called "Stochastic Continuous Change Detection and Classification (S-CCDC) " for Landsat Analysis Ready Data (ARD) time series. The new methodology is built upon a state-space framework, which assumes that observations are made of an unobserved series of states vectors on trends, annual and semi-annual components. Kalman filter is applied to update the system and predict the change each time the new observation is brought in. This study addresses a methodological goal pertinent to the utility of dense time-series of remotely sensed images for characterizing gradual forest change concerning beetle-induced forest degradation, which can be potentially applied to the global forest pest epidemic happening elsewhere

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login