المستودع الأكاديمي جامعة المدينة

A mathematical transformation of multi-angular remote sensing data for the study of vegetation change

أعرض تسجيلة المادة بشكل مبسط

dc.contributor Nolin, Anne W.
dc.contributor Kimerling, A. Jon
dc.contributor Cohen, Warren B.
dc.date 2007-02-07T21:04:41Z
dc.date 2007-02-07T21:04:41Z
dc.date 2006-12-01
dc.date 2007-02-07T21:04:41Z
dc.date.accessioned 2013-10-16T07:44:00Z
dc.date.available 2013-10-16T07:44:00Z
dc.date.issued 2013-10-16
dc.identifier http://hdl.handle.net/1957/3928
dc.identifier.uri http://koha.mediu.edu.my:8181/xmlui/handle/1957/3928
dc.description Graduation date: 2007
dc.description Vegetation change is an important factor affecting the global carbon cycle, land-atmosphere interaction, and terrestrial ecology. The study of vegetation change on a global scale can be used to evaluate the impact of global climate change on terrestrial ecosystems. Satellite remote sensing can monitor vegetation change at the global scale, providing continuous samples of radiation reflected by vegetated surfaces with a temporal resolution of days. The MISR instrument offers the potential to sample the specular anisotropy of the Earth from up to 9 angles. Characterization of the specular anisotropy of vegetated surfaces on a global scale will provide information on the physical characteristics of vegetation affecting anisotropy not available from nadir-view only remote sensing. The objective of this study is to develop a Principal Components Analysis (PCA) transformation of multi-angular measurements of the Earth’s surface acquired by the MISR instrument, to examine the feasibility of quantifying the structural characteristics of different vegetation communities at a global scale. This transformation will be applied to a time-series analysis of the Kalmioposis Wilderness in the Siskiyou National Forest in Southwestern Oregon to better understand the changes in spectral and angular reflectance of a forest stand during re-growth after a stand-replacing disturbance. A sample encompassing a full phenologic cycle, of the red bands only from MISR cameras Ca – Cf, at scaled surface reflectance, provided the template on which PCA was performed. The sample of MISR data was created using imagery collected from 2001 – 2005 to provide a wide variety of vegetation and soils reflectance over a phenologic cycle. Sample data was rotated to the principal components as the new axes using the coefficients of rotation from an un-standardized PCA. Samples were evaluated at various latitudes, differing topography, and varying vegetation density and land cover to determine the properties of the scene controlling the range and magnitude of the principal components. Principal component 1 was found to have high negative correlation to NDVI. Principal component 2 was found to have high positive correlation to both the solar zenith angle and the relative azimuth angle between the MISR sensor and the path of incident radiation. Principal component 3 could not be correlated to any available metric, although evidence suggests that component 3 may carry useful information. The PCA transformation proved useful at relating the changes in vegetation after a fire at the Biscuit Complex. The changes in the BRDF as sampled by MISR were expressed through the principal components, but these changes could not be directly related to changing structural characteristics of the vegetation. The goal of assessing structural characteristics of vegetation through the PCA transformation to a single metric of vegetation structure was unsuccessful. The PCA transformation of the MISR sample successfully yielded a transformation where different classes of vegetation occupied distinct and unique regions of PCA space. The first two principal components were successfully correlated to measurable and definable metrics of vegetation and solar illumination. The third principal component, for which a correlation could not be found, was suggestive of carrying unique information and merits further investigation. The transformation of multi-angular red band reflectance as presented in this study may prove to be a valuable method of estimating biomass at a global scale. With principal components correlated to measures of biomass in NDVI and to the shadowing of the ground through the angle of solar illumination, the PCA relates characteristics of vegetated scenes in a minimum of bands.
dc.language en_US
dc.subject MISR
dc.subject PCA
dc.subject vegetation
dc.subject multi-angle
dc.title A mathematical transformation of multi-angular remote sensing data for the study of vegetation change
dc.type Thesis


الملفات في هذه المادة

الملفات الحجم الصيغة عرض

لا توجد أي ملفات مرتبطة بهذه المادة.

هذه المادة تبدو في المجموعات التالية:

أعرض تسجيلة المادة بشكل مبسط