Service Delivery and Public Weather Services—an Overview

by Gerald Fleming*



Over the past couple of years, the concept of “service delivery” has moved centre-stage within WMO. But what is service delivery? What does it encompass and what does it imply? How do National Meteorological and Hydrological Services (NMHSs) need to change and develop in order to properly address this challenge and attain good Service Delivery? This talk will attempt to explore these issues.

Lessons from the pizza parlour

It is Friday night and we are at home, tired after the week’s work. Hungry too, but we don’t feel like donning the apron and setting to at the cooker. So, we phone for a pizza. We place our order, are given the time for delivery and sit down to wait, uncorking a nice bottle of red wine to begin an evening winding down. So, we have commissioned a service. In terms of this service, there are four attributes that will need to be satisfied if our primary purpose—to be fed while enjoying a relaxing evening—is to be attained. These four attributes are:

  • The pizza as delivered should be the one we ordered; if we wanted a calzone, we should get a calzone;
  • The food should be good quality; proper crispy dough, nice tomato sauce, generous toppings;
  • It should arrive hot;
  • It should arrive on time.

If we now consider the chef at the pizza parlour; the person who “produces the goods” as it were, he/she has control over only one of these attributes—the second one. We could put the best chef in the world into the kitchen, turning out high-quality pizzas, but unless the other three attributes are satisfied, our experience of the service is not good.

In order to provide us with the service we desire, the management of the pizza parlour, as well as hiring a good chef, must ensure that the other necessities of good service delivery are in place. They must establish a clear and unambiguous ordering system, whereby the details of the clients and their desired pizza remain closely linked. The system should allow the person taking the order to quickly make a realistic estimate of when the pizza will be ready and how long the delivery will take, so that they can provide a delivery time to the client. The management needs to provide transport with warmed boxes so that the product can be delivered hot and in good condition. They need to ensure that their delivery drivers are well-versed in the geography of their area of business, so that they don’t get lost and the delivery thus delayed.

The point here is that there are lots of things that need to be organized in the background in order to maintain a good service to the client. At home on a Friday night, however, you are blissfully unaware of all this; you are just happy that you have a tasty hot pizza, as ordered. You do not need, nor do you want, to know about how it all works behind the scenes. Enough about pizza. What lessons does this have for meteorology?

Service delivery—what is it all about?

One view of service delivery is that it encompasses all the bits that are “added on” to the meteorology in order to provide a product or service; that it is the window dressing. This window dressing would include concepts like: marketing, branding public relations and presentation and indeed good service delivery will include elements of all of these. However, if this were the entire story, all NMHSs would need to do would be to hire marketing people, give them their product, and get them to “package” it for the client or public.


Service delivery— lessons from the pizza parlour for public weather services!


The contention of this paper is that good service delivery is a much broader concept, and one that reaches back into the way we do our meteorology. It is not just an “add-on”; it should be integral to the way in which we organize ourselves and our science. Like the management of the pizza parlour, we must regard the chef (in our case the forecasters, or the system of numerical weather prediction (NWP) models and post-processing) of just one part of a complex system that is designed, from the start, to address and satisfy the needs of the clients.

Let us take a fresh look at what we actually do. We provide information, analysis and forecasts concerning the weather. The information comes from instruments that read or sense the weather, ranging from the humble thermometer in the Stevenson screen to the radiometers on the weather satellites which provide us with images over a multiplicity of spectral channels. Some of this information is straightforward (“The temperature today is 15 degrees”), some of it requires interpretation. Analysis encompasses the synthesizing of all the available information into some sensible conceptual view of what the atmosphere is doing: perhaps through the identification of a phenomenon as a tropical cyclone, a thunderstorm, a cold front, or whatever. It allows order to emerge from the chaos of thousands of individual weather observations. Forecasting is then a statement of how we expect the atmosphere to behave over a coming period of time; that period could stretch from minutes and hours to months or even years, depending on the phenomenon under consideration.

There is an analogy with medicine, to which we will come back later. You don’t feel well. The doctor makes observations regarding your symptoms; once more these could be simple (the thermometer again!) or complex (maybe a CAT scan). Their analysis (or diagnosis, in medical terms) gathers together this information into a coherent concept of what is happening to your body. Then, drawing on their knowledge and experience, they predict how that somewhat abstract concept (virus, infection, etc.) will evolve and what the consequences will be for you. Finally, drawing on this forecast, they may make a prescription or provide suitable advice. The advice may not be what you want to hear; the information may not be welcome, but (in general) it is better to get that knowledge earlier rather than later.

Back to the weather: the idea of “service delivery” is a concept that can be defined around a number of “abilities”:

Availability—is the information relevant to the client, and there when the client needs it?

Dependability—can the client expect the information to be delivered on-time and without fail?

Usability—is the information presented in a manner which enables the client to fully understand it?

Credibility—does the client have faith in what is (frequently) no more than a professional opinion?

To help focus on these “abilities”, we might ask the following questions about our products and services:

  • Will the information help the client to solve their particular problem; answer their particular question; make their particular decision?
  • Do we know what are the clients’ problems, questions and potential decisions?
  • And if we don’t, should we not ask the client directly?

Looking back to look forward

We might ask at this juncture what the point of meteorology is anyway. The modern approach to meteorology was first conceived in response to a problem—the loss of many sailing ships in wrecks around the coasts of western and northern Europe. We might pause to note that the impetus was probably not humanitarian in origin but concern with the loss or diminution of military power and with the loss of assets manifested in the merchant ships and their cargoes. So, meteorology was initially developed to address a particular problem.

Since then, some of the major advances in the science and organization of meteorology have come from the need to address other problems, specifically the growth of aviation through the early and middle years of the 20th century and (at various intervals) needs of a military nature. In more recent decades, as expertise in forecasting many days ahead was developed, the needs of agriculture became a major focus of NMHSs. It is interesting to note that these needs always existed; the cycle of tilling, planting, husbanding and harvesting has existed for centuries. However, meteorology was simply not relevant to this business until it had developed the capacity to provide reliable information on a time-scale which was useful to agriculture; the time-scale of days to weeks ahead.

Today, we are faced, in the developed world at least, with a much more complex society. The needs of this society—the problems to be addressed to which meteorology can contribute—are not as obvious as they were heretofore. The developing world has its own set of problems and challenges and addressing them is rendered more difficult by the widening gap between developed and developing. Technologies appropriate to the former may be completely unsuitable for the latter. So:

  • The problems with which society has to deal are many and multi-faceted;
  • The connections with meteorology are not as obvious as they once were;
  • Meteorology needs to work harder to embed itself, its services and products, in business and in society.

Let us look back at our “abilities” in more detail and try to define what they mean in the context of service delivery.


What do users need? Do they know what they need? Does the NMHS understand the nature of the problem/decision which the user has to make? Does the NMHS appreciate how the information and expertise at its disposal can help the client? Does the user appreciate the extent to which the NMHS can provide useful information and advice?

Answering these questions means consultation and discussion; it means the NMHS personnel taking time to get to know the business of the client. It means the preparation of sample products and services which will help the client to understand the extent to which the NMHS (or other meteorological service provider) can offer assistance. Availability can reach right back into the infrastructure of meteorology, resulting in the siting of an observation system at a location where it will provide readings relevant to a client, or perhaps running an NWP model at a resolution or over a domain which will match a client’s specific needs.


If a client is going to use weather information in an organized and coherent manner he/she needs to have it delivered in a timely fashion, or perhaps made available quickly and easily on-demand. This implies that attention be given not just to the production of the product or service, but also to the means of delivery (remember the pizza?). Sending information out is one thing, but ensuring that it reaches its intended destination is a step further. If a client is going to depend on a product or service, all the steps in the chain that lead from service provider to user need to be tested and not found wanting.


Once the client has the information in his/her hands, can he/she use it? This depends on many factors. Are the appropriate meteorological parameters provided? Are they presented in a manner which allows to client to extract the information they need quickly and easily? Has the user been trained sufficiently to realize the full significance of the information and how it might apply to his or her unique situation? Does the client have a contact point in the NMHS with whom he/she can follow up on receipt of the forecast and ask any supplementary questions as he/she feels necessary?

The concept of usability has many consequences for the presentation of the weather information. The style in which a forecast is written, for example, may affect the degree to which it is usable by the client. For visual representation, the skills of graphic artists and designers are often required in order to present the information clearly and unambiguously. Our earlier concept of “window-dressing” is relevant here too, although not central to the job of presenting information, it helps if weather information presented graphically also looks well, in terms of composition and use of colour. Indeed, in a competitive environment, such as the World Wide Web, attractive presentation is probably more important, in terms of attracting users, than the quality of the underlying information.


Of all the “abilities”, this is probably the most important—certainly in the case of forecast information. In order for information to be of use to a client in making a decision, he/she needs to believe in it. How do we foster this belief? Certainly, the underlying quality of the information is an important element here. However, all weather information is qualified to some degree, so it is important for the client to understand the limitations of the service provided. The client needs to understand that a forecast will, on occasion, be wrong. While a “meteorological” explanation of why a particular forecast went awry can be useful, this is of no real assistance to the client in using and interpreting future forecasts. What is relevant here is the establishment of verification scores, allied to a quality management framework.

Verification scores in this context are not about some abstract scoring of purely meteorological phenomena. Verification scores have relevance only if they are a tool to identify weaknesses in the system; a process which will allow those weaknesses to be addressed and the ultimate service improved. User-based verification scores are part of a feedback process within the forecast system itself. A properly defined verification score, however, will also allow the client to acquire a good understanding of how best to use forecast information in his/her specific instance (which will depend on the relative costs of protective or preventive action that might be taken and the losses that might otherwise ensue).

A properly established Quality Management Framework on the other hand, will provide NMHS management with the tools to identify weaknesses in their systems and to apply resources in such a manner as to strengthen the weak areas effectively.

For all that these are necessary, however, credibility is most often invested by humans in other humans, not in systems. Going back to our medical analogy: if we were to be ill today we could probably type our symptoms into Google and get back a listing of the probable causes; refining our search might indeed lead to a “diagnosis”. But would we do his? Most of us would instead go to see our doctor, whom we know and trust. This is despite the fact that the combined wisdom and knowledge available on the Web is many times greater than that which our doctor could possible retain. However, in order to follow a particular course of treatment, we need to believe that this will be effective (even though medical statistics indicate clearly that there is a significant probability that it will not) and this belief derives from trust in the judgement of another.

The lesson for meteorology is that credibility in a product or service is very much tied in with the people who deliver that product or service; in the example of our client, the individual who personifies the service is the contact point within the NMHS. Persons who are contact points carry the brand of the NMHS with them when they meet with, or speak to, clients; if they perform poorly, the brand suffers accordingly. For an NMHS, the implications are careful selection and training of those whom it sends out into the public, or client’s, eye.

Meteorology, public service and the media

The delivery of meteorological services (including public weather services) through the media deserves special consideration in the context of service delivery. Much of what is discussed above is relevant to a single client or group of clients, who have a definable need for certain weather services and products. When dealing with public services, though, other considerations apply. These are typically “push” services—they are placed out in the public domain, through the media or otherwise, and there is no strong feedback mechanism which would allow an NMHS to gain a sense of how their services are being received.

This feedback gap can be addressed through the commissioning of opinion surveys, etc., but this is not a common practice within the meteorological community, in part because of a lack of resources and in part because it was never seen as part of the work of a “scientific” organization. The situation is further complicated when the public service is delivered (as it so often is) through the media. The media organization—be it broadcaster, newspaper publisher or otherwise—now lies between the NMHS and the public; this organization is partly a client and partly a medium to reach the ultimate clients. The NMHS thus has to satisfy two completely different sets of requirements if it is to deliver a good and effective service. These can be defined as follows:


Media needs

Timely delivery
Presentation standards
New skill sets
Cross-promotion within media
Media branding

Public needs

Clear information
Predictability of coverage

Of course, these differing requirements are not mutually exclusive; they often support each other. The media organization, for example, will be just as concerned with clarity of information and predictability of coverage as the members of the public, while adding other elements (branding, exclusive use) which are primarily aimed at serving the media organization rather than the public.

However, credibility—listed above under “public needs”—is just as important to the media organization, although there will be a strong tendency to personify this credibility in the presenter (in broadcasters) or writer (in publishers) rather than in the NMHS from which the information emanates.

    Map picture

The growth of Websites as a source of weather information has weakened somewhat this tendency for credibility in weather to be associated with a particular personality; use of Websites appears to be more driven by ease of use/access and quality graphical presentation of information rather than by any other parameter.

As broadcast and Internet technologies converge and the bandwidth available to the public increases (in the developed world at least), however, it is likely that personality-driven weather information will strengthen once again. The weather man or woman familiar from the television will be coming soon to a (small) screen near you, courtesy of improved video compression techniques and the rapidly increasing computing power available in portable computers and hand-held devices.

Back to credibility

In discussing credibility earlier in the case of an individual client, we laid emphasis on the need for the client to understand the limitations of the service, and to have ready access to the verification statistics which would allow them to make rational weather-based decisions in respect of their own business needs. Clearly, this implies a sophisticated understanding of the boundaries of weather forecasting, and an acceptance that the losses incurred when the forecast goes wrong are more than balanced by the gains when the forecast is right. This represents a high level of awareness, and it would be unrealistic to expect the public to attain this, just as it would be unrealistic to expect the NMHSs to allow, in its public forecast service, for every possible activity in which members of the public might indulge. Public service forecasts will go wrong from time to time, yet the NMHS must remain credible in the public eye if its services are to be of value. How can this be achieved?

While the level of scientific knowledge and understanding within a community will vary substantially, an NMHS nevertheless has a duty to raise the level of meteorological knowledge and to increase the understanding of the scope and limitations of the services which it provides. Thus, the exhortation by Fifteenth World Meteorological Congress (2007) to NMHSs to “Engage in education, awareness and preparedness activities aimed at helping citizens make the best use of forecasts and warnings information, understand the potential impacts of severe weather, and be aware of the appropriate mitigating actions”.

This exhortation represents a significant challenge for NMHSs that takes them well beyond their traditional roles as passive providers of weather information. Yet it is a challenge that must be met if the public are to invest proper credibility in forecast products and services. If this credibility is not generated and maintained, the weather services on offer are greatly devalued.


We have travelled some distance with our pizza delivery. Service delivery is not an add-on; it is not window dressing. It is much more than these. It encompasses an attitude which starts with defining the needs of society and of individual clients, which then uses those needs to develop the fundamentals of our systems and drive the progress in our science. It remembers that a forecast service is a means to an end; the end being the safety and security of our citizens and the sustainable development of our societies. It recognizes that the skills of many other professions must be joined with those of meteorology to achieve the maximum benefit from what we know and understand. It realizes that the point of it all is not those within the world of meteorology, but those without.

* Met Éireann (Irish National Meteorological Service) and Chair, WMO Commission for Basic Systems Open Programme Area Group on Public Weather Services


Share this page