Using Object Based Image Analysis to Track Changes in Russian Olive Distributions Along the Missouri River in Response to Flooding— 70p — Zachary Schild, Mark Dixon, Venkatesh Kolluru, Sakshi Saraf

Russian olive (Elaeagnus angustifolia; RO) is an invasive nitrogen-fixing tree that has become prevalent in the riparian ecosystems of the western United States. The Missouri National Recreational River (MNRR) has witnessed the expansion of RO over the past few decades, potentially a consequence of modified flow patterns from upstream dams. RO has the potential to compete with native tree species, many of which have already seen declines in recruitment post dam construction, which could result in reduced tree species richness with potential impacts on populations of birds and other wildlife. Our objective is to develop a model employing object-based image analysis (OBIA) to accurately identify RO using high resolution National Agriculture Imagery Program (NAIP) Imagery. Image segmentation will be performed within eCognition Developer using 4-band (RGB-NIR) imagery and two indices, Enhanced Bloom Index (EBI) and NDVI. To accomplish this, 376 present day locations of RO were identified in the field to train and validate the model. This model will serve as a valuable tool for mapping current distributions of RO and contrasting it with historical distributions from 2010, 2012, and 2018. Through analysis of historical distributions, we seek to quantify the impacts of flood events in 2011 and 2019 on RO distributions within the MNRR. This research will contribute to our understanding of the invasive potential of RO in the MNRR and can provide insight for land managers on how RO is spreading and affected by floods. This model may also serve as a prototype for mapping other riparian invasive species using high-resolution imagery.

University of South Dakota
Ranjeet John