Using Isoscapes to Model Probability Surfaces for Determining Geographic Origins
Using Isoscapes to Model Probability Surfaces for Determining Geographic Origins
复制标题
使用等景观对概率曲面进行建模以确定地理起源
DOI：
10.1007/978-90-481-3354-3_12
复制
发表时间：
2010
影响因子：
2.2
通讯作者：
Michael B. Wunder
中科院分区：
文献类型：
--
作者：
Michael B. Wunder
本章介绍了一个建模框架，使用isoscape来描述采样材料的可能地理来源。该方法首先反转isoscape模型，然后添加已知或估计的方差分量，以创建单个连续的后验概率密度，该后验概率密度描述任何给定位置作为原点的概率。根据方差结构的复杂性，后验密度可以通过解析或蒙特卡罗方法来定义。该模型必须用已知来源的材料进行训练，并可用于确定单个样品地理来源的空间显式概率密度。个体密度可用于总体上找到样本群体的地理起源中的空间结构。为了说明这种方法，我探索模拟数据的一个假设的例子，从野生动物法医学涉及的跨季节的连接性在候鸟物种的确定。
This chapter describes a modeling framework for using isoscapes to describe probable geographic origins of sampled material. The approach first inverts an isoscape model and then adds known or estimated components of variance to create a single continuous posterior probability density that describes the probability of any given location as the origin. The posterior density is defined analytically or by Monte Carlo methods, depending on the complexity of the variance structure. The model must be trained with material of known origin and can be used to determine spatially-explicit probability densities for geographic origin of individual samples. Individual densities can be used in aggregate to find spatial structure in the geographic origins for populations of samples. To illustrate the approach, I explore simulated data for a hypothetical example from wildlife forensics involving the determination of cross-seasonal connectivity in a migratory bird species.