Long term monitoring of seagrass distribution in Moreton Bay, Australia, from 1972–2010 using Landsat MSS, TM, ETM+

Abstract

Seagrass ecosystems are well studied and seagrass is recognised as a vital contributor to overall ecosystem health and productivity. However, a significant gap in knowledge exists in terms of the large scale temporal and spatial dynamics of cover level and distribution of seagrass communities. Remotely sensed satellite imagery offers a means to map seagrass cover and distribution over large temporal and spatial scales. At present, no operational methods have been produced to map seagrass on large spatio-temporal scales (> 100km2). This study presents a combined per-pixel/object-based method to rapidly map seagrass cover and distribution from a full Landsat archive, from 1972–2010 (MSS, TM and ETM+), with no in-situ data and at accuracies as good or better than existing mapping methods. The products provide management agencies with a baseline assessment as well as the capacity to continue to map seagrass distribution and predict changes in the future.