Elements for Life

[

] 189

ORECASTING AND OTHER

weather-related information

provision has improved immensely during recent

decades. The improvement in forecasting means that it

is possible to describe coming weather events in terms of

attributes such as intensity, location and duration. Due to this

more complete set of variables and probabilities, not only

weather variables but also other related variables are identifi-

able. New means of communication and visualization are

contributing to improved possibilities for forecasting services,

and other important decision-making materials can be more

easily integrated.

A prerequisite for positive forecasting outcomes is that deci-

sions are made over and over again based on the content of

good forecasts. The predictability of weather is such that, while

the outcome of individual forecasts might be inaccurate, the

integrated value over time should be accurate.

Quality is not only an academic issue; it has to embrace

several dimensions in order to develop an optimised decision-

making process. It is important to take into account

continuously developing technologies and techniques to

enhance the value of forecasting, both in economic terms and

in terms of mitigating damage from predicted strong weather

events. This view of forecasting will hopefully provide some

new thinking on how to optimise service quality and how to

improve it in an adequate parallel to the developing sophisti-

cation of meteorology as a science, where the state-of-the-art

is now based on supercomputer and space techniques.

On quality

Verification measures are expected to reveal the quality of fore-

casts. However due to the breadth and varied skill levels of its

audience, an accurate forecast may still be confusing for one

end user while providing a lot of useful information to another.

A forecast can be considered to exhibit value if it helps the end

user to make decisions on the basis of that particular forecast,

regardless of skill.

A service meeting its users’ expectations is not necessarily

100 per cent accurate. If it provides an acceptable mean to facil-

itate decision making, it may still be a satisfactory service. The

main goal of validation is to authenticate and quantify the

delivered products, so that users can be informed on the quality

and limitations (and therefore the applicability) of the infor-

mation that they are receiving.

Quality definitions

The outcomes of validation processes will be a measure of the

quality of the service/products, in its broadest sense. The

concept of quality should be clearly broken down, since all

these quality components are identified as part of a successful

service.

The product value should depend on technical quality (TQ)

and functional quality (FQ). Operational quality (OQ) can be

seen as apart of the FQ, but also as a separate part of the full

quality concept.

Technical quality

– TQ is directly related to the service’s tech-

nical specifications. It gives information on the accuracy and

scientific maturity of the products. The TQ of a categorical or

probabilistic forecast is a measure of the accuracy of the fore-

cast statement, with accuracy measured using the relevant

range of metrics that quantify how close the forecast was to

the observed value or the analysed value it was intended to

predict. TQ might describe how well the predicted precipita-

tion, temperature, water level etc. corresponded to the actual

measurements. It might also be described as the skill involved

in the forecast. However, TQ is understood in quite different

ways by different users. The requirements associated with it

could differ between those of a user wishing to overview a year

and one looking at a specific hazardous event. The quantifica-

tion of TQ must be performed by taking into account the

nature of the service and the type of information relevant to

each thematic domain.

The value of weather forecasts:

quality, decision-making and outcome

Erik Liljas, Swedish Meteorological and Hydrological Institute

Quality components of a successful service

Source: From the validation concept of PREVIEW – a project within the 6th Framework

Programme of the European Union.

Accuracy,

Correctness

TECHNICAL

QUALITY

FUNCTIONAL

QUALITY

Operational

Quality

Operability,

Portability,

Training support,

Other support

Availability,

Usability,

Credibility,

Understandability

VALUE