May 23, 2018
Congratulations to UW/PFC/RSGAL 2 newly minted MSc — defense videos available here!
May 21st, 2018 Caileigh Shoot successfully defended her MSc work entitled:
Classifying FIA Forest Type from a Fusion of Hyperspectral and LiDAR Data
In this study we develop a methodology for classifying United States Forest Service (USFS) Forest Inventory and Analysis (FIA) defined forest type across the Tanana Inventory Unit (TIU) using a fusion of hyperspectral and LiDAR data. The hyperspectral and LiDAR data used in this study were collected as part of the 2014 acquisition with the NASA Goddard’s LiDAR, Hyperspectral & Thermal Imager (G-LiHT). In order to determine the best classification method, we tested 5 classification algorithms: Naive Bayes Classifier, K-Nearest Neighbor, Multinomial Logistic Regression, Support Vector Machine, and Random Forests.
Each model was trained and validated using the forest type corresponding to each FIA subplot, alongside raw hyperspectral band data (114 in total), hyperspectral vegetation indices, and selected LiDAR-derived canopy height and topography metrics. Six different combinations of this input data were tested to determine the most accurate classification algorithm and model inputs. A 3-fold cross validation was performed in order to ensure that all data was included in both training and validation, but never within the same model. Of the five models and six model input combinations tested, we found random forest with hyperspectral vegetation indices as well as topography and canopy height metrics as model inputs had the highest accuracy at 77.53% overall. With the completion of this work, we hope to use this “best” model to classify forest types across the Tanana Inventory Unit in central inland Alaska where there is GLiHT coverage.
Video of the Defense & Power Point of the Defense Talk
Note: Caileigh is doing an internship with NASA in Washington DC this summer
_________________________________________________________
May 22nd, 2018 Travis Axe successfully defended his MSc work entitled:
Leaf Area Index in Riparian Forests: Estimation with Airborne Lidar Vs. Airborne Structure-from-Motion and the Societal Value of Remotely Sensed Ecological Information
Remote Sensing technology has expanded tremendously over the past few decades and has created value when integrated into environmental concepts and practices. But there is unmet potential for bolstering ecosystem services and creating additional value for society. Impediments such as the cost and complexity of the technology, and the difficulty of readily assimilating it into a decision-making process, must be overcome to facilitate broader use.
This study demonstrates the capacity for an emerging and inexpensive remote sensing technology to estimate an important ecological indicator and then discusses the broader implications for societal value. First, we compare the estimation of effective leaf area index (LAIE) between two remote sensing methodologies: discrete-return airborne laser scanning (ALS) and airborne structure-from-motion (SfM). LAIE is an indispensable component of process-based ecological research and can be associated with a variety of ecosystem services. SfM data acquisition is inexpensive and more frequent compared to ALS, but its capabilities less explored. Two point-cloud data files for each technology were evaluated using respective field-measured reference data. SfM shows promise: a combinational linear regression revealed that the distribution elevation values of upper-canopy point returns and the elevation values representing mid and max stand-level, when paired grey-level co-occurrence matrix (GLCM), can estimate LAIE (r2 = 0.62). Although it did not perform as well as ALS, which has more data representing light attenuation behavior (r2 = 0.66), SfM as an alternative methodology for remotely sensing ecological data has demonstrated potential and warrants further investigation.
Next, we discuss how remotely sensed ecological information like LAIE can create value for society. We provide a primer on the ways in which society values the environment and how these values may be perceived and quantified, and the dynamic behavior that exists between them. Next, we introduce a major policy tool used in quantifying these values, benefit cost analysis, and why it is useful for framing environmental issues and how remote sensing can contribute to its outcomes. Finally, we review remote sensing applications used in increasing our understanding of the interaction between society and existing opportunities for value addition.
Video and Power Point of the defense talk.
Note: Travis is doing an internship with Weyerhaeuser Company this summer
