Dear all,
This year's joint St Andrews/Highlands RSS group meeting will take place in the Maths Building, St Andrews, Lecture Theatre B on November 21st. Please see the schedule and abstracts below.
We would like to get a rough idea of how many people will come to the meeting, for catering purposes. If you are planning to attend please email Janine Illian ([log in to unmask]) by Monday 12th of November to confirm your attendance.
Best wishes
Janine
------------------------------------------------------------------------------------------------------------------------
schedule:
2:00: Welcome
2:05 - 2:50: Tony O'Hagan,University of Sheffield: "Aspects of Model Uncertainty"
3:00 - 3:45: Mark Brewer, BIOSS, Aberdeen: "Source distribution modelling for compositional analysis in hydrology "
coffee break
4:15 - 5:00: Alan Gelfand, Duke University, Durham, USA: "Point pattern modeling for degraded presence-only data over large regions"
abstracts:
Tony O'Hagan (University of Sheffield):
The MUCM project (http://mucm.ac.uk) has been developing Bayesian statistical methods for quantifying and managing uncertainty in the outputs of complex mechanistic models. Such models are used in numerous fields to predict, understand and control complex physical systems, and are typically built out of differential equations that represent the best available scientific knowledge. For users of such models a key question is how accurate they are as predictors of the real-world system - this is the issue that MUCM addresses. One cause of inaccuracy is errors in the parameters that are input to the model, but another key factor is model discrepancy. "All models are wrong", and this is as true of mechanistic science-based models as it is of statistical models. Even when the correct, true values of input parameters are used, the model will not predict reality correctly, and the difference is model discrepancy.
In this talk, I will introduce the methods developed in MUCM for tackling uncertainty in the predictions of mechanistic models. I will look particularly at the complications caused by
model discrepancy, and will go on to consider the implications for model error in statistical models.
Mark J Brewer (Biomathematics and Statistics Scotland):
End-member mixing (EMM) is a method in hydrology for attempting to define the runoff sources in river catchments. It involves estimation of the relative proportions of water from different sources, and is often recorded as a time series. Given regular measurements of a chemical tracer on the target water body and, in addition, corresponding measurements for samples of known sources, it is possible to perform end-member mixing using compositional analysis taking a Bayesian random effects approach in a hierarchical framework, including covariates if appropriate.
This talk considers the case where there are no separate data available for the source components, and develops a model for source distributions via nonlinear regression on the tracer/flow relationship and nonparametric density estimation. We allow these source component distributions to vary from year to year and apply the model to a data set from two streams in central Scotland, comprised of weekly or fortnightly readings of alkalinity over seventeen years. We conclude there is evidence of a change in source distribution over time; that corresponding to low flow conditions exhibits a gradual increase in alkalinity for both of two streams studied, whereas for high flow conditions alkalinity appeared to be rising for only one stream.
Alan Gelfand (Duke University):
Explaining species distribution using local environmental features is a long standing ecological problem. Often, available data is collected as a set of presence locations only thus precluding the possibility of a presence-absence analysis. We propose that it is natural to view presence-only data for a region as a point pattern over that region and to use local environmental features to explain the intensity driving this point pattern. This suggests hierarchical modeling, treating the presence data as a realization of a spatial point process whose intensity is governed by environmental covariates. Spatial dependence in the intensity surface is modeled with random effects involving a zero mean Gaussian process. Highly variable and typically sparse sampling effort as well as land transformation degrades the point pattern so we augment the model to capture these effects. The Cape Floristic Region (CFR) in South Africa provides a rich class with such species data. The potential, i.e., nondegraded presence surfaces over the entire area are of interest from a conservation and policy perspective.
Our model assumes grid cell homogeneity of the intensity process where the region is divided into 37, 000 grid cells. To work with a Gaussian process over a very large number of cells we use predictive process approximation. Bias correction by adding a heteroscedasticerror component is implemented. The model was run for a number of different species. Model selection was investigated with regard to choice of environmental covariates. Also, comparison is made with the now popular Maxent approach, though the latter is much morelimited with regard to inference. In fact, inference such as investigation of species richness immediately follows from our modeling framework.
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|