5. Conclusion

Wetlands are productive and diverse ecosystems providing numerous ecological services that are biologically important as well as playing a key role in surface water hydrology and flood risk. Wetlands are and have been threatened by land-use conversion, increased urbanization, industrial development, and climate change, resulting in more than half of the world's wetlands threatened, damaged, or destroyed. Earth observation provides a new cost-effective approach to mapping wetlands to aid in their management especially in remote and difficult to access regions. A combination of optical and SAR data provides adequate input data to use an object-based classification with machine learning algorithms such as Random Forest resulting in classification accuracies exceeding 90% for study sites in Newfoundland/Labrador.

For more details on some of the information discussed in this chapter, please refer to our published papers [3, 26, 38, 40–42, 45–48, 50, 51, 61, 64, 68, 69, 98–100]. While [42] is a literature review paper on the use of interferometric synthetic aperture radar (InSAR) data for water level monitoring of wetlands, the rest mainly introduces new machine learning methods for wetland classification using optical, SAR data, or the combination of both.
