Author:

Publisher:

Published: 1993

Total Pages: 324

ISBN-13:

DOWNLOAD EBOOK

While many questions regarding human impact on tundra ecosystems are regional in spatial extent, the patch level is the largest scale at which experimental validation is possible. Since the individual organism ultimately responds to perturbations, it is necessary to scale up to higher levels. This in turn requires an understanding of spatial pattern that can be observed at a landscape scale. In this thesis, relationships between the spatial pattern of the physical environment and vegetation pattern of an arctic tundra landscape in the foothills of the Brooks Range, Alaska, are analyzed by testing the hypothesis that the spatial pattern of plant communities can be quantified using topography as the only spatial variable. The hypothesis is first tested by examining the spatial relationship between patterns of the normalized difference vegetation index (NDVI) and the water regime. Using gridded elevation data, a model (T-HYDRO) is developed to generate a 2-dimensional water flow field for the watershed. The results show that pattern of water flow can account for about 43% of the spatial variance in NDVI, supporting the hypothesis. Secondly, the G-model concept is developed to predict tundra community vegetation patterns based on topographic gradients. Maps showing patterns of slope and discharge were used to generate quantitative gradient models. The models predicted vegetation pattern at Imnavait creek (10% of a larger mapped region) with an accuracy of 70%. Validation of models based on the relationships developed at Imnavait Creek watershed resulted in an accuracy in predicted vegetation pattern of about 60% for the entire region; again supporting the hypothesis. The spatial pattern of prediction errors revealed the influence of landscape age and snow drifts. The appendix presents a software toolkit for modeling using spatial data. It is designed to enable access to spatial data using the most modern and widely used programming language C++. The system enables input and output of file formats used by different geographic information systems, comfortable and efficient access to entire layers and single pixels, and includes some fundamental GIS functionality such as map overlay. The usage of the routines is illustrated by several example programs.

Download or read book Spatial Explicit Modeling of Arctic Tundra Landscapes written by and published by . This book was released on 1993 with total page 324 pages. Available in PDF, EPUB and Kindle. Book excerpt: While many questions regarding human impact on tundra ecosystems are regional in spatial extent, the patch level is the largest scale at which experimental validation is possible. Since the individual organism ultimately responds to perturbations, it is necessary to scale up to higher levels. This in turn requires an understanding of spatial pattern that can be observed at a landscape scale. In this thesis, relationships between the spatial pattern of the physical environment and vegetation pattern of an arctic tundra landscape in the foothills of the Brooks Range, Alaska, are analyzed by testing the hypothesis that the spatial pattern of plant communities can be quantified using topography as the only spatial variable. The hypothesis is first tested by examining the spatial relationship between patterns of the normalized difference vegetation index (NDVI) and the water regime. Using gridded elevation data, a model (T-HYDRO) is developed to generate a 2-dimensional water flow field for the watershed. The results show that pattern of water flow can account for about 43% of the spatial variance in NDVI, supporting the hypothesis. Secondly, the G-model concept is developed to predict tundra community vegetation patterns based on topographic gradients. Maps showing patterns of slope and discharge were used to generate quantitative gradient models. The models predicted vegetation pattern at Imnavait creek (10% of a larger mapped region) with an accuracy of 70%. Validation of models based on the relationships developed at Imnavait Creek watershed resulted in an accuracy in predicted vegetation pattern of about 60% for the entire region; again supporting the hypothesis. The spatial pattern of prediction errors revealed the influence of landscape age and snow drifts. The appendix presents a software toolkit for modeling using spatial data. It is designed to enable access to spatial data using the most modern and widely used programming language C++. The system enables input and output of file formats used by different geographic information systems, comfortable and efficient access to entire layers and single pixels, and includes some fundamental GIS functionality such as map overlay. The usage of the routines is illustrated by several example programs.