A survey of remote sensing techniques used to monitor a dynamic wetland complex, Cheyenne Bottoms, Kansas.

摘要

Cheyenne Bottoms, a renowned interior wetland in central Kansas, is prone to dramatic fluctuations in water level, as well as the invasion of non-indigenous species of vegetation. In order to preserve and manage this sensitive ecosystem, it is imperative to monitor conditions of the wetland and its immediate surroundings. Large-scale field observations of this nature are most practically accomplished through the use of remotely sensed data.;Four types of remotely sensed images were used to explore the spatial and temporal dynamics of the Cheyenne Bottoms wetland area. The types of data analyzed included historical aerial photography, Landsat TM imagery, Ikonos imagery, and small-format aerial photography. Images spanned more than five decades in acquisition date, beginning in 1957, and continuing until 2009. Corresponding weather data were also gathered for the years studied to examine the visual relationship between precipitation levels and wetland features. It was concluded that each type of imagery offers unique benefits when monitoring an unpredictable wetland environment; for this reason, it is often beneficial to use multiple kinds of remotely sensed data when analyzing this type of dynamic ecosystem.;Landsat TM images have an advantageous cost-benefit ratio, yet the large pixel size of these scenes is prohibitive for detailed analysis of wetland phenology. Ikonos images provide an ideal combination of spectral and spatial resolution, but the cost-benefit ratio is quite high, as scenes cost thousands of dollars, and cover a limited surface area when compared to Landsat TM data. The multispectral capabilities and fine spatial resolution of Ikonos imagery facilitate more detailed analyses of wetland features when compared to either Landsat images or small-format airphotos alone. However, advanced image processing techniques would require further ground truth verification and spectral refinement.;Small-format aerial photography has been shown to be particularly useful for documenting the dynamics of interior wetlands. The extremely fine spatial resolution enables differentiation of wetland phenology; this is more difficult using other types of remotely sensed imagery, as emergent vegetation exhibits inherent small-scale variations.;In order to thoroughly interpret the environmental changes that take place in a dynamic wetland ecosystem, it is most advantageous to gather multiple types of remotely sensed data. Small-format aerial images, in particular, can be used to augment analysis results of Landsat TM or Ikonos scenes. By providing in situ verification of ground cover conditions, small-format airphotos could improve accuracy of advanced image processing techniques, such as automated cluster classification. Success of such techniques could ameliorate the effectiveness of large-scale remotely sensed images used to assess wetland dynamics.