Following on from the recent discussions on the list I have some questions on
the distributional properties of pollutants. The answers will help me and may
possibly help to summarise some of the current discussions.
Environmental pollutants such as heavy metals or radionuclides are often
found to have approximately lognormal distributions (frequency of
occurrence v. sample concentration).
These pollutants may also show spatial correlation (e.g. a relationship
between concentrations measured at locations 'close' to each other).
1) Do you get a reasonable representation of the underlying distribution (eg
lognormal) only when when random sampling at greater than the spatial
2) Is there a method of estimating the underlying distribution of pollutant (eg
lognormal) when the sample values available are spatially correlated ?
3) If the underlying distribution is assumed to be known (eg lognormal)
does this place any constraints on the form of the spatial correlation?
4) If the lognormal distribution is the result of a large number of random
processes acting multiplicitvely (from the central limit theorem) what is the
interpretation to put on other environmental distributions such as the weibull?
5) There may be many different scales of spatial correlation. How will
these combine to affect the distribution?
Direct Line 44-1235 822764
Fax 44-1235 833891
*To post a message to the list, send it to ai-geostats@...
*As a general service to list users, please remember to post a summary
of any useful responses to your questions.
*To unsubscribe, send email to majordomo@...
with no subject and
"unsubscribe ai-geostats" in the message body.
DO NOT SEND Subscribe/Unsubscribe requests to the list!