Abstract
Spatial binary coverage maps, which are images with binary-valued pixels, can represent the pattern of presence and absence of vegetation and other substances. This thesis applies statistical theory for random sets to lacunarity indices, box-counting fractal dimension estimators and other quantitative descriptions of binary coverage maps that are popular in landscape ecology. The application potential of the well-known gliding box lacunarity index is greatly expanded by new estimators
that tolerate irregularly-shaped spatial domains. Further benefits include associations between quantitative descriptions, statistically-efficient estimators, and estimators of estimator variance. Hazard rate models for inference of random sets are also investigated.
that tolerate irregularly-shaped spatial domains. Further benefits include associations between quantitative descriptions, statistically-efficient estimators, and estimators of estimator variance. Hazard rate models for inference of random sets are also investigated.
Original language | English |
---|---|
Qualification | Doctor of Philosophy |
Awarding Institution |
|
Supervisors/Advisors |
|
Thesis sponsors | |
Award date | 26 Sept 2019 |
DOIs | |
Publication status | Unpublished - 2019 |