Work
Package 3 (WP3): Data collection
Start
date or starting event: Month 9
Objectives
and input to workpackage
-
The
task of WP3 is to collect the data needed to construct the conditional
probability tables for the case study belief networks. Each network
will include a wide range of environmental, economic, social and
political factors. Existing data collection will, therefore, cover
a wide range of disciplines drawing data from a large number of
sources. This data will be input directly into WP2.
Methodology
/ Description of work
-
Collection
of existing data at each site will be carried out by the organisation
based in that area (i.e. partner Nos 1,3,4, and 5). Overall responsibility
for completion of the task falls to GEUS. The main phase
of data collection will begin at month 9, though some basic information
that is certain to be needed (e.g. rainfall), can be obtained before
this time. Each partner will have a team of environmental, economic
and social science specialists to identify and collect the information
required in each of these fields. For the social and economic aspects
of the work each participant has a named expert in this field. The
conditional probability tables constructed for each variable should
be based on the best information available. This may be in the form
of an extensive existing set of measurements such as long-term river
flow records or economic statistics.
-
Tables
may also be constructed using the output from models. For example,
a model may be used to calculate groundwater recharge; an economic
model to predict land use change based on changing market and policy
conditions; or an ecological model to simulate habitat diversity.
Part of the data collection exercise will be to identify the range
of models whose output can be used to help generate conditional
probability tables.
-
In
some instances the data available for a particular link will be
limited or even non-existent. In these cases it may be necessary
to fall back on 'expert opinion'. Raising the price of water, for
example, will tend to lower demand in areas where it is metered.
But there may be no data or model to indicate the level of demand
reduction. An expert opinion may thus be introduced, which reflects
the best estimate available. Such a prediction will inevitably be
associated with a much higher level of uncertainty than other approaches,
but this uncertainty will be explicitly expressed in the output.
-
Collection
of data in the four areas will not be a trivial task but a complex
and time consuming operation, which involves making contact with
a large number of organisations. The networks will make use of existing
data and models; no additional field studies are planned.
|
