A brief review of remote sensing data and techniques for wetlands identification

Abstract

This paper reviews some data issues and associated techniques which have been implemented in the field of remote sensing for the identification of wetlands. Wetlands are found in diverse geographic areas, however in situ measurements from wetlands may often be hard to obtain. A review of some data sources and associated techniques revealed that satellite data is a good means of identifying wetlands. Further, various data sources may be complimentary in nature, and by combining data from various sources (including ancillary data), better identification of wetlands can be achieved. While multi-temporal imageries aid in the detection wetlands, the date(s) of imagery could have a large effect on the accuracy of the classification of wetlands. Various techniques for wetland identification exist, some of which are statistically based while others may be rule-based or machine learning based. Hybrid techniques combining unsupervised and supervised classification are gaining popularity along with rule-based methods due to their higher accuracies in classification.