We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
A SIMULATION APPROACH TO USING LANDSAT 8 IMAGERY TO DETERMINE A THRESHOLD FOR DETECTING CHANGES ALONG A STREAMSIDE MANAGEMENT ZONE: A CASE STUDY IN LOUISIANA.
- Authors
GAY, D. C.; CROSBY, M. K.; HOLDERIEATH, J. J.; MCCONNELL, T. E.
- Abstract
Satellite data is often employed to assess land use/land cover changes, particularly over larger areas. However, little attention is given to how much area can change before a given land use/cover classification is detected using satellite data. This is an important consideration, particularly in the use of image classifications to assess best management practices (BMPs). To determine these changes, and their corresponding impacts on land cover classification, Landsat 8 data was acquired and an area selected where two land cover classes meet (i.e., forest and field). The Landsat pixels were subset into 900 one square meter (1 m²) pixels and the average pixel values for grass were utilized to simulate tree/forest removal. The objective is to determine how many pixels would be converted from forest to field before an unsupervised classification detected the change. Approximately 25 percent of the area changed before one Landsat pixel (30m) changed classes and 43% of pixels changed before a row, representing a streamside management zone (SMZ), changed. This indicates that image resolution should be considered when using satellite imagery to assess BMPs/land cover changes.
- Subjects
LOUISIANA; LANDSAT satellites; RIPARIAN areas management; LAND cover; PIXELS
- Publication
Mathematical & Computational Forestry & Natural Resource Sciences, 2020, Vol 12, Issue 2, p74
- ISSN
1946-7664
- Publication type
Article