Genus-Level Mapping of Invasive Floating Aquatic Vegetation Using Sentinel-2 Satellite Remote Sensing

Abstract

Invasive floating aquatic vegetation negatively impacts wetland ecosystems and mapping this vegetation through space and time can aid in designing and assessing effective control strategies. Current remote sensing methods for mapping floating aquatic vegetation at the genus level relies on airborne imaging spectroscopy, resulting in temporal gaps because routine hyperspectral satellite coverage is not yet available. Here we achieved genus level and species level discrimination between two invasive aquatic vegetation species using Sentinel 2 multispectral satellite data and machine-learning classifiers in summer and fall. The species of concern were water hyacinth (Eichornia crassipes) and water primrose (Ludwigia spp.). Our classifiers also identified submerged and emergent aquatic vegetation at the community level. Random forest models using Sentinel-2 data achieved an average overall accuracy of 90%, and class accuracies of 79–91% and 85–95% for water hyacinth and water primrose, respectively. To our knowledge, this is the first study that has mapped water primrose to the genus level using satellite remote sensing. Sentinel-2 derived maps compared well to those derived from airborne imaging spectroscopy and we also identified misclassifications that can be attributed to the coarser Sentinel-2 spectral and spatial resolutions. Our results demonstrate that the intra-annual temporal gaps between airborne imaging spectroscopy observations can be supplemented with Sentinel-2 satellite data and thus, rapidly growing/expanding vegetation can be tracked in real time. Such improvements have potential management benefits by improving the understanding of the phenology, spread, competitive advantages, and vulnerabilities of these aquatic plants.

Monitoring the Spatiotemporal Distribution of Invasive Aquatic Plants in the Guadiana River, Spain

“…In [51], the Random Forest (RF) classifier was applied to Sentinel-2 images allowing for the detection of water hyacinth (among other land covers). In [52], linear discriminant analysis (LDA) was used on Sentinel-2 images acquired in the Greater Letaba river system, South Africa.…”

Invasive Aquatic Vegetation in the Sacramento–San Joaquin Delta and Suisun Marsh: The History and Science of Control Efforts and Recommendations for the Path Forward

“…In this area, significant progress has occurred in the use of remote-sensing technology to provide synoptic maps of IAV coverage for the full region (Hestir et al 2008; reviewed in Hestir and Dronova, this issue), something that is not possible with land-based methods. Recent monitoring advancements include the ability to map FAV species at the genus level using satellite data that has a high frequency of collection which establishes the potential to create maps intraannually to inform management (Ade et al 2022).…”

“…The same approach is and Sutula 2015;Ta et al 2017;Khanna, Conrad, et al 2018) that uses a remote sensing approach because it is the only way to produce synoptic coverage maps. A promising approach includes use of satellite-acquired data, which is cheaper and collected more frequently than piloted aircraft (Bubenheim et al 2021;Ade et al 2022).…”

Remote Sensing of Primary Producers in the Bay-Delta

“…In cases where species-and genus-level mapping are not needed, pixel sizes of 30 m or less are sufficient for many wetland and salt marsh mapping applications (Turpie et al 2015;Byrd et al 2018;Muller-Karger et al 2018). In the Delta, recent work has shown that 30-x-30-m pixels from Landsat and 20-x-20-m pixels from Sentinel-2 can be used to map wetlands, FAV and SAV, and rice paddies (Baldocchi et al 2016;Ta et al 2017;Ade et al 2022), and to track changes in spectral indicators of primary production (Anderson et al 2016).…”

“…While less spectral information is contained in multispectral data, these data sets form the bulk of currently available, free, openaccess remote-sensing data sets, although this is changing (see "Looking Ahead: The Very Near Future of Remote Sensing of Primary Producers in the Estuary"). Multispectral data provide valuable information about landcover (Clark 2017), the extent and distribution of wetlands (Quinn and Epshtein 2014), different vegetation communities and plant functional types (Villa et al 2015;Ade et al 2022), and even wetland biodiversity . Multispectral information can also be used to assess the physiological status of plants, primarily using spectral indexes, such as the NDVI.…”