“…However, it is difficult to clearly distinguish kelp canopies from open ocean water in satellite imagery because of factors like: sun glint, breaking waves, sediments, dissolved organic matter or phytoplankton blooms that also contribute to water reflectance [14]. Techniques and sensors used for kelp mapping have included aerial photogrammetry [11,15], Landsat Multispectral Scanner System (MSS) [16], Landsat Thematic Mapper (TM) [13,14], Landsat Operational Land Imager (OLI) [17], Sentinel-2 [18], Satellite Pour l'Observation de la Terre (SPOT) [19,20], and aircraft multispectral sensors with spatial resolutions~2m [17,21]). Methodologies employed to detect kelp canopies using such imagery have been varied: from using thresholds of band brightness value ratios to indices such as the Normalised Difference Vegetation Index (NDVI; [13]) to methods such as unsupervised classification [19,20], supervised maximum likelihood classification [16,17,22], Principal Component Analysis (PCA; [13]), Spectral Mixture Analysis (SMA; [18,23]), and particularly Multiple Endmember Spectral Mixture Analysis (MESMA; [14,24]), which has been effective in studying different characteristics of kelp ecology, such as genetic differentiation within geographic clusters [25], long-term biomass trends [26], fish, invertebrate and algae assemblages in kelp communities [27], and biogeochemical [28] and physical [14] dynamics.…”