My current projects include:


1) the Urban Ecology { W E B S T U D I O } come visit!


2) a research project on wildlife habitat modeling, using different ways of constructing categories for mapping and statistical analysis. OK, that's a little more esoteric.... it's like this:

February 1996, K. Hill

Method Overview:


This project will compare four means of representing definitional uncertainty in suitability analysis:

1. Boolean categories (which have crisp upper and lower bounds, and for which membership is represented numerically with either a 0 or a 1);

2. indexed categories (which use a range between 0 and 1 as a representation of membership);

3. probabilistic categories (which have crisp upper and lower bounds, and which represent membership as a likelihood varying between 0 and 1); and

4. fuzzy categories (which use a "coefficient" ranging between 0 and 1 to describe degrees of membership, and which are defined as overlapping sets).


Each of these four approaches will be used to represent categories and criteria in three suitability analyses, which evaluate a single landscape according to separate (and potentially conflicting) objectives:

A. conservation of bird habitat;

B. optimization of timber harvest; and

C. development of the best 200 acres for mixed-density housing.


The objective of this comparative analysis is, firstly, to determine which means of representation (from alternatives numbered 1-4 above) provides the best spatial approximation of actual bird habitat as measured in field surveys, and secondly, to determine in what ways the spatial pattern of conflicts between the three objectives (habitat, timber harvest, and housing) varies as a function of category representation in a suitability analysis.