Please find below details of the next meeting of the Merseyside local
RSS group.
Wednesday 22nd November 2006 2.00pm
(Tea & Coffee in Common room afterwards, 3rd Floor)
Venue : Penthouse, Maths & Oceanography building University of Liverpool
Please contact Ashley Jones ([log in to unmask]) for further
details.
---------------------------
Geostatistics, Preferential Sampling and Environmental Monitoring
Peter Diggle (Lancaster University)
Geostatistics involves the fitting of spatially continuous models to
spatially discrete data (Chiles and Delfiner, 1999). Preferential
sampling arises when the process which determines the data-locations and
the process being modelled are stochastically dependent. Conventional
geostatistical methods assume, if only implicitly, that sampling is
non-preferential, even when the context makes this assumption
implausible: for example, in a pollution monitoring network, monitors
will typically be placed close to likely sources of pollution.
We first review the model-based approach to geostatistics under
non-preferential sampling (Diggle,Moyeed and Tawn, 1998). We then use an
idealised model to investigate the consequences of incorrectly assuming
that sampling is non-preferential, and to illustrate some of the
difficulties which arise in fitting preferential sampling models.
Finally, we describe two environmental monitoring applications in which
sampling is known to be preferential, and discuss possible analysis
strategies.
Chiles, J-P. and Delfiner, P (1999). Geostatistics: Modeling Spatial
Uncertainty. New York: Wiley.
Diggle, P.J., Moyeed, R.A. and Tawn, J.A. (1998). Model-based
geostatistics
(with Discussion). Applied Statistics, 47, 299-350.
|