The reliability of vegetation indices for monitoring saltmarsh vegetation cover
Abstract. The dynamic nature of saltmarshes poses a challenge for coastal zone monitoring; natural variability in the sediment reflectance and atmospheric visibility within saltmarsh environment affects the reliability of regression relations between vegetation cover and vegetation indices. Analysis of field data indicates that regression equations should only be applied to sparse canopies. Mixture modelling and signal-to-noise analysis show that resistance to sediment reflectance variability is more important than resistance to atmospheric visibility changes. Therefore MSAVI and GEMI are the best indices to use for saltmarsh vegetation cover monitoring.