Determining uncertainties and their propagation in dynamic change detection based on classified remotely-sensed images

Shi, Wenzhong; Ehlers, Manfred

doi:10.1080/01431169608949103

Cited by 24 publications

(12 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For successful change detection, a co-registration error of less than half a pixel is desired (Dai and Khorram 1998). It has been shown that even registration accuracy of better than one-fifth of a pixel can still lead to a CD error of approximately 10% (Dai and Khorram 1998;Khorram et al 1998;Shi and Ehlers 1996;Townshend et al 1992). However, it is highly unrealistic to achieve co-registration accuracies smaller than one pixel.…”

Section: Class Selection and Classification Methodsmentioning

confidence: 94%

Plant community dynamics relative to the changing distribution of a highly invasive species, Eichhornia crassipes: a remote sensing perspective

et al. 2011

View full text Add to dashboard Cite

Eradicating or controlling invasive alien species has frequently had unintended consequences, such as proliferation of other invasive species or loss of ecosystem function. We explore this problem using a case study of a highly invasive floating aquatic macrophyte, water hyacinth (Eichhornia crassipes), in the Sacramento-San Joaquin Delta of California. We used 5 years of remote sensing data to perform change detection analysis to study plant community dynamics contemporaneous with changes in water hyacinth cover. Our results show that as water hyacinth cover decreased, submerged aquatic plant (SAP) cover increased and vice versa. This effect was strongest in large patches of water hyacinth. We found no evidence that the native floating aquatic species, pennywort (Hydrocotyle umbellata), benefitted from reducing cover of water hyacinth. In most years, pennywort cover either showed no trend or followed the same trajectory as water hyacinth cover. In this study a decrease in cover of water hyacinth most often resulted in colonization by SAP species with some habitat returning to open water.

show abstract

Section: Class Selection and Classification Methodsmentioning

confidence: 94%

Plant community dynamics relative to the changing distribution of a highly invasive species, Eichhornia crassipes: a remote sensing perspective

et al. 2011

View full text Add to dashboard Cite

show abstract

“…There are many change detection approaches developed for remotely sensed images (Singh 1989 Great Salt Lake dynamics 1817 1995, Dimyait et al 1996, Jensen 1996, Shi and Ehlers 1996, Mas 1999, Morisette et al 1999. Among these digital change detection techniques, the post-classification comparison method is particularly attractive because the nature of change can be identified.…”

Section: General Methodologymentioning

confidence: 98%

Mapping and visualizing the Great Salt Lake landscape dynamics using multi‐temporal satellite images, 1972–1996

Hung

2005

International Journal of Remote Sensing

View full text Add to dashboard Cite

This study focuses on monitoring and visualizing the Great Salt Lake dynamics from satellite images. Objectives of this research are to identify the Great Salt Lake and vicinity areas land cover types, to detect and quantify water and wetland dynamics, and to visualize these dynamics. Satellite imagery, including Landsat Multi-Spectral Scanner (MSS) and Thematic Mapper (TM) images, were obtained for the extent of the Great Salt Lake and vicinity from 1972 to 1996, with a three-year interval, except 1978 and 1981. These satellite images were first classified to map five land cover types: water, wetland, salt, developed land, and undeveloped land. These classified images were compared with each other to identify areas of change. Documentation, including tables and charts, was generated to report changes. In the final stage, computer animation files were created from both classified images and satellite images to visualize and simulate the Great Salt Lake and vicinity land cover changes and landscape dynamics, respectively. The implication of this study is that it is easy, with the right educational materials, for land managers and planners, as well as the public, to quickly comprehend that the Great Salt Lake is very dynamic. Satellite images, visualization techniques, and computer animation files prove useful for demonstrating these dynamic characters.

show abstract

“…This is particularly true in highly dynamic landscapes where it is less reasonable to make specific assumptions about the rate of land-cover change and the observed land-cover change trajectories in the context of the magnitude of classification errors. Because of the inherent error in landcover classification maps derived from satellite imagery, special attention is needed in the development of products that are measurements of land-cover change (Shi and Ehlers 1996;Carmel et al 2001). A classification for a single time point includes some pixels that are correctly classified and some pixels that are incorrectly classified.…”

Section: Uncertainty Precision and Accuracy In Carbon Assessmentsmentioning

confidence: 99%

National forest carbon inventories: policy needs and assessment capacity

2008

View full text Add to dashboard Cite

Previous research has identified the importance of the role of land cover in the global carbon cycle. In particular, forests have been identified as a significant carbon sink that can mitigate the rate of global climate change. Policy makers are faced with complex and difficult challenges in getting timely and useful information in monitoring global forest resources. Recent advances in the tools and methods of forest carbon accounting have produced new, innovative approaches to forestbased carbon inventories. But it is important as new tools are developed that scientists understand the needs of policy makers and that policy makers understand the capabilities and limitations of forest inventory methods. This paper explores four different policy applications that rely, or could benefit from, national carbon inventories. The goal is to help build a bridge between the communities of climate policy makers and scientists specialized in forest carbon inventories. To this end, we pursue three specific objectives: First we provide an overview for policy makers about approaches to forest carbon inventories, paying particular attention to the contributions of remote sensing technologies. Second, we outline the issues particularly relevant to forest inventory scientists who are interested in responding to public policy needs. We then discuss the tradeoffs between information cost, accuracy, precision, transparency and timeliness that need to be balanced in long-term monitoring of forest carbon. Finally, the article concludes with a series of observations and 70 Climatic Change (2009) 93:69-101 recommendations for the implementation of forest carbon inventories as increasingly central components of global climate change policy.

show abstract

Determining uncertainties and their propagation in dynamic change detection based on classified remotely-sensed images

Cited by 24 publications

References 3 publications

Plant community dynamics relative to the changing distribution of a highly invasive species, Eichhornia crassipes: a remote sensing perspective

Plant community dynamics relative to the changing distribution of a highly invasive species, Eichhornia crassipes: a remote sensing perspective

Mapping and visualizing the Great Salt Lake landscape dynamics using multi‐temporal satellite images, 1972–1996

National forest carbon inventories: policy needs and assessment capacity

Contact Info

Product

Resources

About