Table of ContentsTable of Contents 
Previous Chapter
The forecaster then examines the derived weather elements on the 12Z eta worksheet and decides that its solution for the surface temperature from hours 2 through 10 is better than the current forecast. Using an interface depicted in Figure 6, the forecaster selects the 2- through 10-hour temperature forecasts and drags them over the forecast worksheet. The selected 12Z eta model guidance replaces the data currently in the worksheet. The forecaster can then examine and edit the worksheet until the desired solution is shown,(1) and then save it in the database as the official forecast.
The tool kit will be designed initially to support only two dimensions plus time. The structure of the tool kit, graphical depictions, and tools is such that the concepts can be applied to three-dimensional fields plus time when appropriate.
The tool kit will contain tools that work with all general weather elements. It will not contain forecast-product-specific tools, such as zone combinations. One of the principal purposes of AFPS is to have all products come from the same source, i.e., the forecast database. Forecasters will not have time to concentrate on individual products, nor should they be forced to do so.
A continuous field is used to represent those weather elements whose values are defined over the entire domain. A contour representation can be used to depict a one-dimensional field (e.g., temperature), while vectors or barbs can be used for a two-dimensional field (e.g., wind).
Each bounded area has an associated value, represented by a series of tags that can be interpreted by the product generators. For example, a bounded area with a value of 3R- would be interpreted as 3 miles visibility and light rain.
During all edit operations, the changes made to the field are displayed immediately along with their original unchanged values. Once the edited changes are saved, the contours and lines representing the original unchanged value are removed from the display.
Examples of the spatial depiction tools are shown in Figure 10. Forecasters will be able to draw new contours, delete contours, and modify existing contours. The last will probably be the most used of this set. In like fashion, it will be possible to add or delete bounded areas, and reshape existing areas. Bounded areas may also be divided or combined.
A forecaster may select an area of continuous field and apply a new value to that area or specify a change to be made. A user-selected weighting function will determine how the modified data blend with surrounding points.
The move/copy tools allow the forecaster to identify a bounded area or to outline an area on a continuous field and move it to another area of the screen. The associated field values are moved to the new location and the hole left behind is filled in using interpolation or extrapolation.
A weather entry calculator may be used to enter or edit the value of a bounded area.
Time series lines (for continuous fields) will be edited in the same way as contours. Appropriate tools for bounded area tags will also be developed.
It will be possible to set or change selected gridpoint values explicitly. The same technique will be used for data presented in matrix form.
Several related tools that the user controls but does not use for depiction editing are contained within the tool kit. These tools perform time interpolation and linking, consistency checks, and worksheet management.
Forecasters are familiar with viewing many elements simultaneously using a station model plot. Although not quite a "tool" (it will be for viewing only), this way of viewing data will be useful in many circumstances. A common way of using this will be in time-series form (Figure 11).
Combining the various depictions provides an additional set of tools, such as the interactive query and bounded area initialization hybrid tools. One such interactive query tool will allow the forecaster to move and click the mouse over the spatial depiction and view the time series or station model plot for the selected point (Figure 12). The bounded area initialization tool (Figure 13 on page 44) will allow the forecaster to vary the value of a bounded area over time.
The capability of overlaying depictions also provides an additional set of tools. Multiple weather elements valid for the same time on the spatial depiction or for the same location on a time series can be stacked or overlaid. An example is shown in Figure 14 on page 45, a possible aviation terminal forecast tool including both observed and forecast time periods for temperature/dew point, winds, clouds, and precipitation. The temperature and dew point are overlaid, while the others are stacked. Similarly, multiple locations can be presented on a time series.
Forecasters will wish to view the change from time period to time period; AWIPS animation capabilities may be used for this purpose. Multiple times can also be presented on the spatial depiction to allow a forecaster to view such changes. For example, bounded area depictions of the precipitation pattern at multiple times may be simultaneously displayed and edited, as illustrated in Figure 15 on page 45. Adjusting the arrows on the time bar will display the corresponding depiction.
A forecaster may have several windows open on the display, each containing one or more depictions. Modifications made on one depiction will result in changes in any related windows, allowing the forecaster to interleave operation sequences as appropriate.
All forecasters are confronted with areas that consistently experience conditions different from surrounding areas. Examples include lake-effect snow areas and fog-prone valleys. Means will be provided for forecasters to define and store special areas of influence to assist with forecasting in these regions.
The user interface necessary to operate the tool kit will be determined through investigative work including prototype testing and risk reduction exercises. A conceptual example is shown in Figure 16.
Once the forecaster has initialized the forecast worksheet as discussed in Section 6.1, she will use the various tools in the tool kit to modify the forecast until it accurately represents the desired forecast solution. The forecaster may first produce some annotated overlays by displaying conventional model depictions (e.g., eta 850-500 hPa thickness, 700 hPa temperature) and drawing the rough positions of the fronts, troughs, cloud areas, and precipitation areas. She may correct these annotations in terms of position and speed based on her knowledge of the models and the particular weather situation. These annotated overlays will probably be produced by standard AWIPS capabilities and will not be a unique feature of the forecast preparation system. Once the annotations are completed, they may be overlaid with spatial depictions of weather elements.
The forecaster first chooses to modify the cloud depictions. She selects the cloud element for the present time and overlays it on a satellite image for comparison. Satisfied that they agree, she selects tonight's cloud element depiction, overlays tonight's annotation and model relative humidity fields, and discovers that the depicted cloud field is farther west than she expects it to be. Using the bounded area modification tool, she redraws the cloud field in the correct location. At the same time, she decides to overlay the "weather" weather element and correct that field. Skipping ahead to tomorrow's late afternoon forecast, she redraws the cloud and precipitation fields to accurately represent their position. Upon command, the interpolation tools fill in all cloud and precipitation depictions between tonight's and tomorrow's late afternoon forecast. The forecaster rapidly scans through the entire cloud and precipitation sequence using standard AWIPS animation features. Satisfied with the results, she proceeds to the temperature and dew point fields.
A cold front is expected to pass through the area tomorrow and drop the temperature by about 15°. The forecaster decides to see how the guidance represents this by displaying temperature time series for three points in her forecast area of responsibility. Realizing that the front is moving faster than the model predicted, the forecaster modifies the three time lines to represent this change. Referring back to the spatial depiction, she draws an area of influence around each of the three points, then clicks the interpolate button. She has now modified many temperature depictions for her entire area of responsibility by editing three time lines. Interpolation routines handle the area of influence, the boundaries between the three outlines, and the time interpolation.
The forecaster quickly checks the wind forecast. Displaying the spatial depiction for the time the cold front will pass through the area, she modifies the data to accurately represent the gusty winds and wind shift expected with the front. Satisfied, she copies the winds in the region around the front to a different area and time, thus keeping the winds around the front consistent.
The areal forecast is nearly complete. Selecting several key locations in her area of forecast responsibility, the forecaster displays station model plots by time. Satisfied, she then turns to those tasks requiring more specific forecasts such as aviation.
The forecaster selects her terminal forecast sites using the locator map for the time series depiction, and displays the aviation terminal forecast tool shown in Figure 14. The weather elements have been initialized from the general weather elements with which she has been working. Knowing that one particular airport always retains low stratus after a front longer than other sites since it is surrounded by low hills, she changes the clouds for that site. Forecasts for the other sites are accurate, so no other changes are necessary.
Satisfied that the depictions accurately represent her view of the forecast, the forecaster saves her worksheet as the official forecast in the database.
The interface consists of a list of forecast products and several status boxes. The forecaster checks the product(s) to generate(2) and selects "Go." The display will indicate the status of the product generation, such as in progress, completed, or released at 1527.
The software will prevent the forecaster from selecting a product to be generated if the forecast database contains inadequate information for the product. For example, there could be field inconsistencies or interpolations to be performed.
Table of Contents 
Next Chapter