Mapping seagrass coverage and spatial patterns with high spatial resolution IKONOS imagery

Abstract

The capability of IKONOS (IKO, 4 m resolution) data for mapping seagrass (%SAV) cover and quantifying spatial patterns along the mid-western coast of Florida, USA, was explored. Both IKO and Landsat TM data were preprocessed to calculate water depth invariant bands and the textural information was extracted from IKO data. Our results demonstrate that the IKO sensor produced a higher accuracy than the TM sensor in a three-class %SAV cover classification, and the overall accuracy of %SAV cover maps created with IKO data was 5–20% higher than these from other studies published. We also examined the spatial patterns of seagrass over a spatial range of 4–240 m using the Ripley's K function and IKO data that represented four different grain sizes (4 m, 8 m, 12 m, and 16 m) from seagrass cover. The Ripley's K metrics repeatedly show a dispersed (or slightly dispersed) pattern across 4 – 32 m and a random or slightly clustered pattern across 33 – 240 m. Therefore, use of the seagrass classification scheme introduced in the paper, along with data from the IKO sensor, offers an opportunity to synoptically record seagrass cover dynamics at both small and large spatial scales.