Vegetation mapping is the process of delineating the geographic distribution, extent, and landscape patterns of vegetation types. Traditionally, riparian vegetation has been mapped using field-based methods, which can be time consuming, expensive, and difficult to apply to extensive riparian systems. Remote sensing methods offer an economical and effective alternative. While not as robust as field-based methods, remote sensing techniques can be applied to extensive riparian systems, allowing resource managers to establish the extent of riparian areas relatively quickly—with an acceptable level of accuracy, and at a reasonable cost.
An integral part of vegetation mapping is the application of an effective vegetation classification system to the mapping effort. Vegetation classification is the grouping of similar vegetation based on selected, shared characteristics, such as physiognomic and floristic similarities. A successful vegetation mapping effort is dependent on a well-designed classification system and robust training data. For more information on these issues, please review the Vegetation Mapping Prerequisites.
It should be noted that the term “classification” can mean different things to the remote sensing analyst and to the ecologist. This can lead to confusion when discussing the application of remote sensing to vegetation mapping. In the ecology world, classification refers to the grouping of similar vegetation based on selected, shared characteristics such as physiognomic and floristic similarities. In the remote sensing world, classification refers to techniques that categorize the pixels in an image into similar groups or themes. In general, this site uses the term “mapping” when referring to remote sensing classification. However, when referring to specific remote sensing methods such as “unsupervised classification,” the term is retained. In general, the reader should be mindful of the context in which the term is used.
Mapping riparian vegetation from remote sensing imagery involves using limited knowledge of vegetation in known areas to extrapolate vegetation type in unknown areas through human interpretation or by computer-aided image analysis. The methods for mapping riparian vegetation using remote sensing imagery are varied, but they generally fall somewhere along a continuum—with human interpretation at one end and computer interpretation at the other. Human-based methods rely on an analyst’s interpretation. Computer-based methods rely on user-provided training data that teach the algorithm of what each vegetation type “looks” like.
The methods listed below are generally categorized by which end of the spectrum they fall on. Click on each method for more information.
- Human Interpretation
- Computer Interpretation
- Object-based Classification
- Classification and Regression Tree (CART) Analysis
- RSAC Riparian Mapping Toolbar—This tool integrates the valley bottom and vegetation mapping into a single workflow based on CART analysis.
The methods presented here are not limited to riparian vegetation mapping; they are applicable to all types of vegetation mapping.