The accurate estimation of atmospheric water vapour and the subsequent derivation of surface spectral reflectance from hyperspectral VNIR-SWIR remotely sensed data is important for many applications. A number of algorithms have been developed for estimating water vapour content from remotely sensed hyperspectral data that do not require in-situ measurements. Two algorithms, the Continuum Interpolated Band Ratio (CIBR) and the Atmospheric Precorrected Differential Absorption (APDA) have proven to be highly effective at estimating atmospheric water vapour. Although highly successful, the two methods still exhibit unwanted or spurious results when challenging conditions are encountered. Such conditions include the estimation of atmospheric water vapour over dark targets, when uncorrected atmospheric aerosols are present and over surfaces with complex spectral signatures.A differential absorption method called the Transmittance Slope Ratio (TSR) has been developed that negates these problems. The TSR method is comprised of a weighted mean radiance that is defined between two atmospheric water absorption features which is divided by a reference channel radiance to produce a measurable ratio value. This, is turn, may be related to a reference curve, such that, the TSR value may be expressed as an atmospheric water vapour content. To test the TSR method over real terrains, AVIRIS and HyMap measured hyperspectral radiometric data were used. Three test sites were used in total with each site allowing different aspects of the water vapour estimation to be critically examined. The sites are, Jasper Ridge and Moffett Field in California and Brukunga in South Australia.The TSR method is found to significantly improve estimated atmospheric water vapour over dark targets (with less than 3.5 % error for reflectances as low as 0.5 %), improvement over nonlinear surfaces, and finally, ++ / improvement in water vapour estimation when atmospheric aerosol conditions are not well known. In the final case the TSR method is found to estimate atmospheric water vapour with an error of less than 2 % when a 5 km visibility is assumed to be 25 km. The final result is at least an order of magnitude better than the CIBR and APDA methods.

Magister Scientiae (Biodiversity and Conservation Biology) / Coastal upwelling is one of the most important oceanographic processes relating to ecosystem function at local and global spatial scales. To better understand how changes in upwelling trends may occur in the face of ongoing anthropogenically induced climate change it is important to quantify historical trends in climatic factors responsible for enabling coastal upwelling. However, a paucity of conclusive knowledge relating to patterns concerning changes in upwelling across the world’s oceans over time makes such analyses difficult. In this study I aimed to quantify these patterns by first identifying when upwelling events occur using a novel method for predictingthe behaviours of coastal upwelling systems over time. By using remotely sensed SST data of differing resolutions as well as several wind variables I was able to identify and quantify upwelling signals at several distances away from the coastline of various upwelling systems. Using this novel method of determining upwelling, I then compared upwelling patterns within all Eastern Boundary Upwelling Systems (EBUS) over a period of 37 years, with the assumption that climate change was likely to have driven variable wind patterns leading to a more intense upwelling over time. Overall, upwelling patterns and wind variables did not intensify overtime. This method of identifying upwelling may allow for the development of predictive capabilities to investigate investigate investigate upwelling trends in the future.

The sagebrush shrub steppe ecosystem is of great concern to researchers, conservationists, and the general public because of the documented declines associated with it. Monitoring in the past has generally been point-based and lacking in long-term data. To overcome these deficiencies, an automated method of monitoring was developed using GIS and remote sensing. Geospatial layers of vegetation, soils, fire history, roads, streams, and springs were acquired and processed to characterize selected monitoring locations. A temporal set of Landsat satellite imagery for the past 30 years was normalized to reduce the effects of sun angle, haze, and sensor change. After normalization, a Tasseled Cap Transformation was adapted with local coefficients to provide a landscape metric which was sensitive to actual ground conditions and meaningful at management level. The Tasseled Cap outputs of brightness and greenness are a relative measure of bare ground and plant productivity, respectively. When measured over time, brightness and greenness provided diagnostic trends and condition of treated big sagebrush communities

京都大学 / 0048 / 新制・論文博士 / 博士(地球環境学) / 乙第12828号 / 論地環博第8号 / 新制||地環||24(附属図書館) / 31315 / 京都大学大学院地球環境学舎地球環境学専攻 / (主査)准教授 西前 出, 教授 渡邉 紹裕, 教授 小方 登 / 学位規則第4条第2項該当 / Doctor of Global Environmental Studies / Kyoto University / DFAM

To address accuracy concerns of the National Land Cover Dataset (NLCD), this case study compares impervious surface from the NLCD to a Detailed Land Cover Dataset (DLCD) for the Town of Blacksburg, Virginia over two time periods (2005/2006 and 2011) at spatial aggregation scales (fine to coarse) and scopes (site-specific to area-extent). When comparing the total impervious surface area, the NLCD overestimated the DLCD by appreciable amounts (12-27%) for the entire town and across all specified land use zones for both time periods examined. A binary pixel-wise accuracy assessment of impervious surface revealed that the NLCD performed well for all scopes except for the single family land use zone (user accuracy <40%). The spatial aggregation of pixels to 90-m led to improved agreement between the two datasets. Using the DLCD as a reference, an empirical normalization equation was successfully applied to the NLCD to further reduce overestimation and data skewness. / Master of Science

The Army National Guard Bureau has implemented a cooperative project with Utah State University to help with the use, display, and evaluation of environmental data for maintaining land condition. Camp Grayling, Michigan, is comprised of deciduous and evergreen forest types. Use of remote sensing for classification has been limited in this region due to the difficulty of species-level classification using single-date remote-sensing techniques . Also, remote sensing has traditionally focused on mapping homogenous zones rather than vegetation boundaries, while one of the concerns for land managers is the nature of vegetation edges (ecotones). This study analyzed each season and band from multiseasonal satellite imagery for their contribution to separating vegetation type and density classes. Then spectral reflectance values for each vegetation and density class were used in discriminant models that define vegetation cover types and densities. These models were then tested against points within 200 m of vegetation boundaries to determine the performance of the models at edges of vegetation types . The reflectance values for vegetation types on Landsat Thematic Mapper (TM), Landsat MultiSpectral Sensor (MSS), and Advanced Very High Resolution Radiometer (AVHRR) imagery were used. Single-band separability decreased with decreasing resolution of the remote sensing data, and the number of spectral bands that could separate means of vegetation and density cover classes was much greater than expected . Winter bands provided more separability than expected for density classes . A VHRR data were shown to provide very little separation and were not included in the discriminant analysis. In the evaluation of the discriminant models, both resubstitution and crossvalidation tests showed that TM and MSS were nearly equal in their ability to discriminate cover types and densities. At the vegetation boundary zones, classification accuracy increased with increasing distance from the edge. These results are encouraging for future classification and monitoring of ecotones using satellite imagery, as picture elements (pixels) of ecotones generally exhibit the characteristics of a mixing of the boundary vegetation types. Further investigation into fuzzy set classification and ecotone classification and monitoring appears warranted.

This thesis explores the idea of a liminal space, as being dreamlike, suspended in time and physically unlocatable. It questions and exploits the boundary between abstraction and figuration in painting. This investigation has been considered from a subjective viewpoint allowing a distancing of space to illuminate new perceptions and experiences through the language of painting. The project has sought to explore the relationship between the natural world and seeing, to deepen and emphasize the other worldliness of an in-between space. This third space has been evoked by a process of abstracting pictorial content, juxtaposition of elements, colour and composition. The thesis is constituted of practice-based 80%, accompanied by an exegesis 20%.

Over the last 50 to 150 years there has been widespread conversion of grassland to shrubland throughout the western United States. A major management concern on the Edwards Plateau is the encroachment of Ashe Juniper (Juniperus ashei). To facilitate brush management programs, I investigated relationships of two endangered species, the black-capped vireo (Vireo atricapillus) and the golden-cheeked warbler (Dendroica chrysoparia), with their habitats at the landscape level. GIS (Geographic Information Systems) and remotely sensed data, such as Landsat imagery, DEMs (Digital Elevation Maps), and DOQQs (Digital Ortho Quarter Quads) were used to evaluate vegetative and geomorphic features within both 100m- and 400m-radius areas surrounding occupied and (assumed) unoccupied sites. Stepwise-logistic regression was used to develop probability models for each species within a catchment and was then applied to the entire Leon River Watershed and evaluated for accuracy. Golden-cheeked warblers were identified in areas with mean juniper cover greater than 70%, mean departure from North (aspect), and maximum slope. For black-capped vireos, mean shrub cover, mean departure from North, and mean slope were important in habitat selection. Variables at the 400m spatial scale best identified areas of probable occurrence for both species, indicating that features of landscape surrounding a territory may play an important role in habitat selection.

Over the last 50 to 150 years there has been widespread conversion of grassland to shrubland throughout the western United States. A major management concern on the Edwards Plateau is the encroachment of Ashe Juniper (Juniperus ashei). To facilitate brush management programs, I investigated relationships of two endangered species, the black-capped vireo (Vireo atricapillus) and the golden-cheeked warbler (Dendroica chrysoparia), with their habitats at the landscape level. GIS (Geographic Information Systems) and remotely sensed data, such as Landsat imagery, DEMs (Digital Elevation Maps), and DOQQs (Digital Ortho Quarter Quads) were used to evaluate vegetative and geomorphic features within both 100m- and 400m-radius areas surrounding occupied and (assumed) unoccupied sites. Stepwise-logistic regression was used to develop probability models for each species within a catchment and was then applied to the entire Leon River Watershed and evaluated for accuracy. Golden-cheeked warblers were identified in areas with mean juniper cover greater than 70%, mean departure from North (aspect), and maximum slope. For black-capped vireos, mean shrub cover, mean departure from North, and mean slope were important in habitat selection. Variables at the 400m spatial scale best identified areas of probable occurrence for both species, indicating that features of landscape surrounding a territory may play an important role in habitat selection.

As well as the many benefits associated with the evolution of multispectral sensors into hyperspectral sensors there is also a considerable increase in storage space and the computational load to process the data. Consequently the remote sensing ommunity is investigating and developing statistical methods to alleviate these problems. / The research presented here investigates several approaches to reducing the dimensionality of hyperspectral remotely sensed data while maintaining the levels of accuracy achieved using the full dimensionality of the data. It was conducted with an emphasis on applications in maximum likelihood classification (MLC) of hyperspectral image data. An inherent characteristic of hyperspectral data is that adjacent bands are typically highly correlated and this results in a high level of redundancy in the data. The high correlations between adjacent bands can be exploited to realise significant reductions in the dimensionality of the data, for a negligible reduction in classification accuracy. / The high correlations between neighbouring bands is related to their response functions overlapping with each other by a large amount. The spectral band filter functions were modelled for the HyMap instrument that acquires hyperspectral data used in this study. The results were compared with measured filter function data from a similar, more recent HyMap instrument. The results indicated that on average HyMap spectral band filter functions exhibit overlaps with their neighbouring bands of approximately 60%. This is considerable and partly accounts for the high correlation between neighbouring spectral bands on hyperspectral instruments. / A hyperspectral HyMap image acquired over an agricultural region in the south west of Western Australia has been used for this research. The image is composed of 512 × 512 pixels, with each pixel having a spatial resolution of 3.5 m. The data was initially reduced from 128 spectral bands to 82 spectral bands by removing the highly overlapping spectral bands, those which exhibit high levels of noise and those bands located at strong atmospheric absorption wavelengths. The image was examined and found to contain 15 distinct spectral classes. Training data was selected for each of these classes and class spectral mean and covariance matrices were generated. / The discriminant function for MLC makes use of not only the measured pixel spectra but also the sample class covariance matrices. This thesis first examines reducing the parameterization of these covariance matrices for use by the MLC algorithm. The full dimensional spectra are still used for the classification but the number of parameters needed to describe the covariance information is significantly reduced. When a threshold of 0.04 was used in conjunction with the partial correlation matrices to identify low values in the inverse covariance matrices, the resulting classification accuracy was 96.42%. This was achieved using only 68% of the elements in the original covariance matrices. / Both wavelet techniques and cubic splines were investigated as a means of representing the measured pixel spectra with considerably fewer bands. Of the different mother wavelets used, it was found that the Daubechies-4 wavelet performed slightly better than the Haar and Daubechies-6 wavelets at generating accurate spectra with the least number of parameters. The wavelet techniques investigated produced more accurately modelled spectra compared with cubic splines with various knot selection approaches. A backward stepwise knot selection technique was identified to be more effective at approximating the spectra than using regularly spaced knots. A forward stepwise selection technique was investigated but was determined to be unsuited to this process. / All approaches were adapted to process an entire hyperspectral image and the subsequent images were classified using MLC. Wavelet approximation coefficients gave slightly better classification results than wavelet detail coefficients and the Haar wavelet proved to be a more superior wavelet for classification purposes. With 6 approximation coefficients, the Haar wavelet could be used to classify the data with an accuracy of 95.6%. For 11 approximation coefficients this figure increased to 96.1%. / First and second derivative spectra were also used in the classification of the image. The first and second derivatives were determined for each of the class spectral means and for each band the standard deviations were calculated of both the first and second derivatives. Bands were then ranked in order of decreasing standard deviation. Bands showing the highest standard deviations were identified and the derivatives were generated for the entire image at these wavelengths. The resulting first and second derivative images were then classified using MLC. Using 25 spectral bands classification accuracies of approximately 96% and 95% were achieved using the first and second derivative images respectively. These results are comparable with those from using wavelets although wavelets produced higher classification accuracies when fewer coefficients were used.