By Thomas LeFebvre
Expressing the weather forecast as a series of gridded fields fundamentally changes the method by which forecasters do their job. Since this new methodology has not been thoroughly tested in an operational environment, it poses significant risks to National Weather Service operations. With no guarantee that it will improve forecast accuracy or increase productivity, what is the motivation for implementing such a massive change in the way forecasts are generated? The answer is that weather forecast information can be much better communicated to users when the forecast can be visualized graphically via modernized products (as in Figure 1).
Defining the weather forecast digitally offers great benefits to users who need comprehensive weather information in order to make better decisions. Once the forecast is expressed as a set of gridded weather elements, users can view this information in virtually unlimited ways, and tailor products to match their specific needs. For example, people traveling by car are interested in weather parameters that affect driving conditions such as precipitation, wind, and visibility. Anyone interested in recreational sailing would be concerned with wind and waves. A system that disseminates forecast data in this way not only gives users the information they need, but also avoids cluttering products with irrelevant information that might be confusing.
Even after the Grapical Forecast Editor (GFE) is operational nationwide, weather forecast offices will continue to disseminate alpha-numeric products based on these grids on a regular schedule. However, with recent advances in high-speed communications such as the World Wide Web and wireless technology, an opportunity arises to fundamentally change the way forecast information is delivered to consumers. Rather than accessing a fixed set of products from each forecast office, users interested in the forecast could invoke interactive applications that query a digital forecast database and be able to precisely specify the type and location of any forecast information desired, and then generate a graphical or textual representation of the future state of the weather. In this article, we discuss some of these applications from the user’s perspective.
Because gridded datasets contain a large amount of detail, graphical products are a natural way to view them since they can accurately represent this detail. Graphical products can take on many forms, such as image, contour analysis, time-series, or as combinations of these types.
Image Products – The simplest way to convey digital gridded data is via an image or picture. Figure 1 is an example of such a product, generated by the ifpGIF program. In this example, the color of each picture element represents a temperature value. Using the map background as a reference, a user of this product can locate the area of interest and deduce the temperature. A simple time animation of temperature images informs the user of how the temperature is expected to change with time.
Vector Graphics – Vector graphics can denote magnitude and direction on the display. This type of product has the added advantage, in that graphics can be overlaid on top of a different image product to produce a combined display of different weather elements. Figure 2 displays a wind field as a graphic plotted on top of an ocean wave height field. This is useful to a sailor in locating the best combination of moderate winds and low wave height.
Grids – Some customers need detailed predictions of weather conditions as input to complex algorithms. These "high-end" customers would receive the gridded data directly, in order to run their applications that require high-resolution surface weather elements. Examples of this type of user include a laboratory running a fire behavior model, a state transportation department interested in predictions of road conditions, and any agency that predicts pollution dispersion.
Interactive Products – One of the best advantages of expressing the forecast in digital form is that customers need not be limited (as they are today) to a specified set of products. The digital nature of the forecast information makes it possible for interactive applications to generate forecast displays of weather-specific user needs. An interface prompts the user for location, time, and weather element type and the application displays the information on the screen. Users can see precise information of interest, and the display is not cluttered with data they consider irrelevant. A few examples follow to illustrate this concept.
Point Forecast in Time – Many times users are interested in knowing what time is ideal for planning a particular activity that could be affected by the weather. For instance, a user who might be planning an outdoor gathering can specify the location on a map and a time period to a point forecast application. The application then fetches the data interactively and generates a display of the weather forecast at that location. The display in Figure 3 illustrates the forecast as a series of graphical objects to indicate weather conditions as a function of time.
Traveling Forecast – A traveler would not want a forecast at a fixed point, but one that moves in space and time. The traveling forecast allows the user to indicate a route and a time of departure and arrival, and the traveling forecast application retrieves the forecast data just like the point forecast, but with location and time as a variable. Figure 4 shows a depiction of the weather as the traveler is enroute.
Graphical Forecast Viewer – At FSL we have developed a Java-based application, the Graphical Forecast Viewer (GFV), that combines several of the examples discussed above. The GFV displays plan (map) view images that animate over time, generates a point forecast as a time-series plot, and allows the user to probe a specific point or area to get more detailed weather and geographical information. (Figure 4 shows one possible interface to the GFV.) The GFV is based on the same type of software modules as those comprising the FSL-developed Local Data Acquisition and Dissemination (LDAD) system [see the November 1997 issue of the FSL Forum]. Users of the GFV can look at any component of the forecast in several different formats. Weather information can be overlaid in any combination, giving the user maximum flexibility for obtaining weather forecast information. In combination with its geographical information, the GFV is a very powerful tool for millions of people (everyone at some time or another) whose decisions are based on the weather forecast.
Audio Products – Interactive products are not limited to those that can be visualized. An automated telephone system could be used to disseminate weather information as well. After dialing, a caller ID system could identify the caller’s location, retreive the appropriate weather information, and deliver the weather forecast for that area in the form of a computer-generated voice. Customers interested in a forecast for a location other than their own could enter a zip code to identify a new location.
Interactive Text – Despite the gridded approach to forecasting, text products will continue to be distributed from every forecast office. This does not necessarily imply that every product received by users originates from the prescribed list. The concept of querying the forecast database still applies, even for text products.
The primary motive for changing the method by which forecasters express the weather forecast from text to digital is so that users of these datasets can make faster and more accurate weather-related decisions. The digital nature of this information allows applications to disseminate the data in a variety of ways so that customers with varying degrees of requirements can be easily accommodated. The National Weather Service has invested great resources into its modernization over the past 10 years. This move toward digital forecasts and modernized products is the last step toward delivering better products and services to the public.
(Tom LeFebvre can be reached at email@example.com or (303)497-6582.)