Steven E. Franklin holds a joint appointment as a Professor in the Department of Environmental Resource Science and the Department of Geography at Trent University in Peterborough, Ontario. His expertise in remote sensing and terrain analysis has been widely recognized, with recent awards by the Canadian Remote Sensing Society (Gold Medal, 2007) and the University of Waterloo Faculty of Environment (Distinguished Alumni Award, 2011).
Multi-platform and multi-sensor remote sensing of vegetation disturbance and recovery in Alberta’s Boreal Forest
Remote sensing provides an ideal data source to detect vegetation disturbance and recovery, and is currently the most feasible and cost-effective option for monitoringdisturbance in larger areas. To gain a better understanding of the dynamics and relationships at various scales, the integration of multiple platform and multiple sensor remote sensing have become a practical and promising approach. In this project we examine the use of satellite, airborne and UAV based data to assess vegetation disturbance and recovery in Alberta’s Boreal Forest. Landsat time series data are now freely available and a standard protocol/product is available for forest stand-replacing change. This project will push the envelope on the sensitivity of Landsat time series data to monitor non-stand-replacing and subtle features, such as seismic lines and partial harvest/defoliation. In this project, we will also use industrial UAV capability for multi-temporal and multi-spectral data acquisitions, with hypothesized advantages in operational deployment (e.g., platform and sensor stability), will be reviewed and tested for vegetation structure and change. Photogrammetric methods to generate point clouds from overlapping high spatial detail UAV imagery are now widely available and could potentially be used to produce point clouds from historical and current metric aerial photographs (typically, at a coarser resolution). This project will test the ability of such historical and current aerial photograph point cloud data to support analysis of disturbance features, such as seismic lines, and the pattern of vegetation recovery over time.
Oumer’s project was completed in fall 2016 and resulted in journal publications on “Classification of annual non-stand replacing boreal forest change using Landsat time series” and on “Hierarchical land cover and vegetation classification from an unmanned aerial vehicle”. Oumer’s contribution to other BERA publications can be found here.
My research is concerned with what accuracy disturbances related to mining and associated activities can be detected and attributed in the area around Conklin, Alberta. Using data from a Landsat time series of Alberta, this will be accomplished by testing the sensitivity of change detection methods and data set training methods to recognize and classify these disturbances. My particular area of interest is multiple disturbances occurring contemporaneously and consecutively in time. The goal of this work is to be able to support and make recommendations to existing protocols that will minimize confusion of these types of disturbances.
Rachel’s project was completed in spring 2017 with an honours thesis and a journal publication on “Detection accuracy of new well sites using Landsat time series data: a case study in the Alberta Oil Sands Region”.
Unmanned Aerial Vehicles (UAVs) are becoming increasingly popular for the remote sensing of vegetation, as these systems present a cost effective, efficient and repeatable method through which to collect high resolution spatial data. My research is focused on using multiple UAVs with different sensor arrays to determine several important forest structural characteristics such as Leaf Area Index, canopy cover and height. In addition, my research will examine the ability of the various UAVs to classify various species based on an analysis of their reflectance patterns. This will provide future researchers with a better understanding of what types of scientific data they can collect from a specific UAV platform.
Griffin’s project was completed in spring 2017 with an honours thesis on “Comparison of UAV–based camera data to multispectral sensor imagery for determining structural characteristics of a coniferous forest”.