Pasar al contenido principal

page search

Biblioteca Sample-based estimation of “contagion metric” using line intersect sampling method (LIS)

Sample-based estimation of “contagion metric” using line intersect sampling method (LIS)

Sample-based estimation of “contagion metric” using line intersect sampling method (LIS)

Resource information

Date of publication
Diciembre 2015
Resource Language
ISBN / Resource ID
AGRIS:US201500212806
Pages
239-248

Quantification of landscape pattern is of primary interest in landscape ecological studies. For quantification purposes, a large number of landscape metrics have been developed, with definitions based on measurable patch attributes. Calculation of these metrics is commonly conducted on wall-to-wall maps, whereas a new interest is to use sample data. It is argued that a sample survey takes less time and results are more reliable. The overall objective in this paper was to present the potential of the line interest sampling method for estimating a special contagion metric. The specific objective was to assess statistical properties in terms of root mean square error (RMSE) and bias of the contagion metric estimator. This study was conducted on 50.1 km² already manually delineated land cover maps from the National Inventory of Landscape in Sweden. Monte-Carlo sampling simulation was employed to assess the statistical properties of the estimator. The simulation was conducted for different combinations of two sampling designs, four sample sizes, five lines transect configurations, three lines transect lengths, and two classification systems. The systematic sampling design resulted in lower RMSE and bias compared to a simple random one. Both RMSE and bias of the contagion estimator tended to decrease with increasing sample size and line transect length. We recommend using a combination of systematic sampling design, straight line configuration and long line transect. We conclude that there is no need to use mapped data and thus polygon delineation errors can considerably be reduced or eliminated.

Share on RLBI navigator
NO

Authors and Publishers

Author(s), editor(s), contributor(s)

Ramezani, Habib
Holm, Sören

Publisher(s)
Data Provider
Geographical focus