|
|
Point processes on linear networks
Moravec, Jan ; Prokešová, Michaela (advisor) ; Pawlas, Zbyněk (referee)
The central theme of this thesis is the theory of point processes on linear net- works, in particular two kinds of the network K-function. The first part is devoted to the theory of stationary point processes in the plane, including the K-function and its estimator. The second part is concerned with the theory of point proces- ses on linear networks. There is defined the Okabe-Yamada network K -function and its estimator, the geometrically corrected network K-function, including its estimator, and there are explained their theoretical properties. In the third part we examine the ability of these two kinds of the network K-function to detect clustering or regularity in point processes on linear networks. There is explained the envelope test, the refined envelope test and the deviation tests. The software environment R with library spatstat is used for simulations.
|
|
|
Alternative K-functions for stationary point processes
Koňasová, Kateřina ; Dvořák, Jiří (advisor) ; Prokešová, Michaela (referee)
The main theme of this thesis is the theory of stationary point processes, in particular the directional K-function. In the first chapter we explain the essentials of planar point process theory including the classical definition of K-function and its estimator. The second chapter introduces two types of the directional K-function: cylindrical K-function whose structural element is a cylinder and directional K-function using double spherical cones. The third chapter presents the comparison of directional K-function and its estimator on an anisotropic version of Thomas process. We also illustrate the major contribution of directional K-function in orientation analysis of point patterns. We introduce a heuristic method for detecting anisotropies in clustered or regular data. 1
|
|
|
Alternative K-functions for stationary point processes
Koňasová, Kateřina ; Dvořák, Jiří (advisor) ; Prokešová, Michaela (referee)
The main theme of this thesis is the theory of stationary point processes, in particular the directional K-function. In the first chapter we explain the essentials of planar point process theory including the classical definition of K-function and its estimator. The second chapter introduces two types of the directional K-function: cylindrical K-function whose structural element is a cylinder and directional K-function using double spherical cones. The third chapter presents the comparison of directional K-function and its estimator on an anisotropic version of Thomas process. We also illustrate the major contribution of directional K-function in orientation analysis of point patterns. We introduce a heuristic method for detecting anisotropies in clustered or regular data. 1
|
|
|
Point processes on linear networks
Moravec, Jan ; Prokešová, Michaela (advisor) ; Pawlas, Zbyněk (referee)
The central theme of this thesis is the theory of point processes on linear net- works, in particular two kinds of the network K-function. The first part is devoted to the theory of stationary point processes in the plane, including the K-function and its estimator. The second part is concerned with the theory of point proces- ses on linear networks. There is defined the Okabe-Yamada network K -function and its estimator, the geometrically corrected network K-function, including its estimator, and there are explained their theoretical properties. In the third part we examine the ability of these two kinds of the network K-function to detect clustering or regularity in point processes on linear networks. There is explained the envelope test, the refined envelope test and the deviation tests. The software environment R with library spatstat is used for simulations.
|
|
|
The analysis of forest vegetation on the Hallasan volcano: evaluation of impact of disturbances and competition on growth and survival of tree populations along altitudinal gradient
ALTMAN, Jan
The study was conducted on Mt. Hallasan (1950 a.s.l.) in South Korea. Forest plots were established along altitudinal gradient (950-1770 a.s.l.) from oak-hornbeam (Quercus serrata, C. laxiflora), to mixed-oak (Quercus mongolica, Taxus cuspidata), and fir (Abies koreana, Sorbus commixta) forests. Vascular plants were identified to species, and their total cover and covers of individual species in herb ({>} 1 m), shrub (1 - 5 m) and tree ({<} 5 m) layers estimated using Braun-Blanquet scale. Dentrended Correspondence Analyses was used to analyze relationship between environment and species composition, lifeforms spectra and plant diversity. The spatio-temporal variations in stem dispersion, radial growth, and mode of competition among trees were examined using the mapped location of trees and tree-rings. Changes in radial-growth were used to analyse stand disturbance history using 36647 measured tree-rings. Histories of radial-growth of individuals trees were further examined using visual classification to 5 different groups of growth trends. Ripley´s K-function was used to determine whether the spatial distribution pattern of trees was random, regular or clumped. Spatial size autocorrelations were used to determine the mode and extant of neighbor interactions. Spatial associations of different groups of trees and random mortality test were also computed for tree stands where Abies koreana was a dominant species.
|
guest ::
