Abstract
Diver assessments inform coral reef health, but global fieldwork is challenging. This thesis linked field data from >1,000 dive sites to a suit of remote sensing datasets to show the potential of upscaling traditional surveys. Chapter 1 modelled satellite-derived variables with reef health metrics (coral, fish, and algae), resulting in high predictive accuracy for a transect of sites from the South Pacific. The key satellite drivers varied by reef zone and scale, cautioning against a one-size-fits-all approach when predicting reef metrics in new areas. Chapter 2 tested the ability of two satellite-dervied benthic maps to proxy fish and coral diversity. Habitat map heterogeneity was positively correlated with biodiversity, especially when using KSLOF-GRE maps vs. Allen Coral Atlas maps. Chapter 3 examined the relationship between spectral variability from remote sensing imagery with fish biodiversity, with the best correlations realized when using cm-scale UAV sensors. Overall, remote sensing has potential for guiding reef conservation globally as it bridges the gap between diver assessments and the drivers of reef health and biodiversity.