192 / 218
[
] 190
These technical-functional-operational dimensions are
complementary and partially interwoven, and all have to be
continuously validated. Correctness, accuracy, functionality,
reliability, efficiency and usability determine the total quality
for an end user in an operational context.
Only when weather dependence is fully understood and
the relevant FQ is fully deployed will there be a direct corre-
lation between TQ and outcome. We can then say that
weather-dependence has been transformed into weather-
information-dependence.
Forecasting and other varieties of weather information have
improved immensely during the last decades, even if tradi-
tional verification shows only moderate improvements. But
have the benefits followed the same trend? Yes, in some
branches where the weather-information-sensitivity is so high
that all quality aspects are immediately updated to maximize
the outcome. In other cases, the availability and immediacy of
information online has had the impact of prioritising simplis-
tic data at the expense of more complex insights.
Too much high value information remains in temporary data-
bases at meteorological institutes. Many institutes are certified
according to the ISO 9001 manual and procedures. However,
this is no guarantee that weather information is converted to
savings in terms of safety, economy or disaster mitigation. The
chain of processes has to be thoroughly investigated. Quality
aspects should improve in such a way that a direct link between
technical quality and outcome is possible. In parallel, the deci-
sion-making process should focus on recognizing what
auxiliary information is needed to make optimal decisions.
The ideas presented here are mainly relevant for sophisticated
users of weather information, but not only commercial
customers. Civil protection authorities, local and central, are
perhaps the most important targets for this enhanced quality
and decision making. The developing sophistication of meteo-
rological science, where the state-of-the-art has supercomputers
and space techniques as integral factors, also requires a more
thorough look into the world of the users of our service. This
is central for further stimulation, feedback, justification and
funds.
Functional quality
– TQ is not a guarantee of FQ. For
example, a perfect forecast that is communicated to the user too
late has zero functional quality although it is technically
correct. A less accurate forecast that is communicated to the
user early enough to allow protective action to reduce poten-
tial losses, is technically less correct but functionally more
valuable. This example illustrates the fact that the distinction
between technical and functional quality is not academic, but
reflects the real-time use of the forecasts.
FQ is mainly related to the ‘quality in use’ of the products:
it includes both subjective judgment and understanding by
the user, and technical capabilities regarding service provi-
sion.
The former is mostly user-centred, and can be defined as the
usability of the service/product – that is, the extent to which
a product can be used by specified users to achieve specified
goals with effectiveness, efficiency and satisfaction in a context
of use. The latter globally identifies the context of use for the
service, encompassing the following parameters: availability
of the service, frequency of delivery, means of delivery, perfor-
mance, timeliness, understandability and learnability.
To sum up, FQ measures the efficiency of the service to meet
the users’ needs, resulting in user satisfaction and productiv-
ity. It is directly related to the capability of the service to be
understood, delivered and used in accordance with users’
expectations.
Operational quality
– Even if a service meets users’ expecta-
tions (FQ) and the delivered products are technically/scientifically
correct (TQ), it does not necessarily mean that its operational
deployment satisfies the user. For example, if the service is not
accessible at the right time, if it is too often unavailable because
of insufficient reliability, or if the user encounters too many prob-
lems regarding training or user support for the service, then it
will not be useful in an operational sense.
OQ measures the capabilities required for a successful oper-
ational deployment, such as reliability, operability, efficient
support, maintenance, training, interoperability, security and
portability. OQ also depends on TQ and FQ values, in the sense
that a zero TQ or a bad FQ would mean a poor OQ.
Shipping has to a great extent a well developed weather-information-sensitivity
Photo: SMHI image archive