179 / 218
production costs, decision making, or similar economic
processes. For such a model to be useful for estimating the
value of information, information itself has to play a role in
the model’s operation. Then we can use the model, with differ-
ent kinds or amounts of information available within it, and
observe how the different information available changes the
model results. From this, we can infer the economic benefits
of different kinds, amounts, or quality of information.
An example of the economic modelling approach is
presented in a recent estimate of the cost savings in US elec-
tricity generation from using 24-hour temperature forecasts to
plan production operations.
2
Particularly in the southern
United States during the summer air-conditioning season, next-
day temperature has an important influence on next-day
electricity demand.
An electricity generator typically has many alternative gener-
ating units, and any combination of these units could be used
to serve the next day’s demand for electricity. Among the
various units, lead times and cost characteristics differ. For
example, preparing different units for service may require
different lead times (with varying costs), and operational costs
may differ as well once a unit is operational. As a result, the
costs of generation to meet a given demand for electricity can
be reduced if a good next-day forecast is available.
In the example study, the investigators used three levels of
modelling to produce an estimate of cost savings from tempera-
ture forecasts in electricity generation. First was a model that
chose, in advance, the best set of generating units to prepare for
next-day use, given a forecast of next-day electricity demand.
Second was a model that made appropriate ‘real-time’ adjust-
ments on the next day to compensate for error in the previous
day’s forecast. This modelling step produced a relationship
between the degree of accuracy in the electricity demand fore-
cast and the ultimate cost of meeting the next-day electricity
demand. The third step in the modelling procedure involved
[
] 177
Restructuring (MAR) Programme to significantly upgrade
observing and forecasting systems at a cost of approximately
USD4.5 billion over 20 years. Suppose, that the improvements
specified in the US households study would cost twice as much
and would also take 20 years to complete. Further, assume that
maintenance and operation costs would begin in the tenth year
at a rate of USD60 million per year. Finally, assume that one
tenth of the benefits (USD0.187 billion) begin in the tenth year
and increase linearly over ten years to the full level estimated
in the study, USD1.87 billion a year. Using a 5 per cent rate of
discount, the net present value of such a program (over a 100-
year time horizon) is USD13.5 billion. This means that the
present value of the benefits of this programme is about 3.02
times as much as the present value of the capital, maintenance,
and operation costs over the time period. This illustration
demonstrates how the values estimated from research such as
the US household study could be used in policy decision-
making to evaluate weather forecast improvement programs.
The value of temperature forecasts in
electricity generation
In some situations, we can use a model to directly represent
economic activities. By a model, economists indicate a set of
equations or relationships that is used to describe behaviour,
Agriculture
A recent study of the potential benefits of improved NOAA
hydrological information by the Office of the NOAA Chief
Economist examined the potential economic value of soil
moisture information for private irrigation management in the
semi-arid Great Plains.
The study estimated significant benefits to farmers which, if
aggregated for the states of Nebraska and Kansas, are worth
USD55 million per year and potentially over USD200 million
per year.
About 45 per cent of these benefits result from more profitable
irrigation and 55 per cent from the opportunity value of
conserved groundwater. Other private or public benefits of soil
moisture data would add to these advantages.
Source: Supalla, R, Martin, D, Adams, R and Weiher, R.
Potential economic value of soil
moisture data for irrigation management in the central Great Plains
, October 2005:
www.economics.noaa
The agricultural industry saves money if given accurate
predictions of wind and rain
Photo: NOAA