Right-click an empty spot in the Metview desktop and choose Create New… Link to File and type /home/ectrain/trx/mv_data/ztuv.grb into the File name box (or navigate there and select the file with the mouse). Note that the text label under the new icon is in italics, and there is an arrow in the bottom-left corner - these indicate that this is a link; if you hover the mouse cursor over the icon, the link details will be shown in the status bar. The padlock in the bottom-right corner tells us that this file is read-only. This file contains analysis and 1, 2, 3, 4 and 5-day forecasts for geopotential, temperature and wind at various pressure levels – visualise it to see for yourself.

You can create links to individual files or complete folders. This facility can be used to share folders between users.

Retrieving data from MARS

Metview has the ability to retrieve data directly from the ECMWF MARS archive (or indeed any other MARS archive installed outside ECMWF). Metview can also retrieve data from certain data sets without much special setup using the MARS Web API.

Create a new Mars Retrieval icon. Edit it and set only the following:

This ensures that the data is transformed onto a regular 1.5 degree grid. Much of Metview's functionality relies on GRIB data being on a grid (regular or quasi-regular) and not stored as spherical harmonics.

If you look at the other parameters in the icon editor, you will see that by default, it will retrieve analysis data (Type = An) of geopotential (Param = Z) on six pressure levels (Levelist = 1000/850/700/500/400/300) for yesterday (Date = -1).

Save the icon, then visualise it. The data should be retrieved, then visualised. From the titles of the different fields in the plot you can confirm that it is the correct data.

We will see in Processing Data how we can manipulate the data returned from MARS.

Icon Feedback

The Mars Retrieval icon gives us the opportunity to explore another feature of Metview. The colour of an icon’s text label tells us which state it is in:

The icon should currently be green, meaning that if you visualise it (or perform any other action on it), the data will not be retrieved again - the cached copy will be used instead. There are two ways to un-cache the data:

1. edit the icon and change at least one parameter (in this case, the cached data will no longer match the retrieval request, so it is deleted)
2. right-click on the icon and select Clear result

Use the second technique, then execute the icon and observe as the colour of its name changes.

Icon Output

By now you should have generated some log message, probably when you performed the MARS retrieval. Each icon has its own text output which can be viewed by selecting Log from the icon’s right-click menu. This is only enabled when there is output for that icon, and is reset when the icon’s contents change. Have a look at the log messages for the Mars Retrieval icon (ensure that you have retrieved data first).

You can view a complete history of output from all icons by selecting Log from the Tools menu in any Metview desktop.

For an even more detailed output, you can start Metview on the command line with the ‘-slog’ option - this will write lots of information to your terminal window. This information can be useful when reporting a problem to the Metview team! ‘metview -h’ gives a list of all useful command line options and environment variables.

Downloading files from URLs

Metview's Download module provides the means to download a file from an arbitrary URL.

Create a new Download icon, edit it and set:

Visualise the icon to retrieve the file and plot it. If there was a problem with the URL then look at the icon's log messages.

Saving results to disk

The data files cached by these icons will be removed when the current Metview session is terminated. To store a permanent copy, right-click on the icon and select Save result. Try this with either your Mars Retrieval or your Download icon.

Field Data in GRIB Files

ECMWF's model output fields are stored in GRIB format, so that is where much of Metview's functionality lies. The following sections will introduce some of the data inspection facilities available.

Examining GRIB Meta-data

Right-click and Examine the Mars Retrieval icon (if for some reason the retrieval did not work, or you do not have access to MARS, then use ztuv.grb instead).

GRIB file. In this case we have 6 fields (messages) in the file, each represented by a row in the message list on the left-hand side. The right-hand side shows detailed meta-information for the selected message, presented in a number of different ways (try changing between Tree view and Plain text; try different Dump modes). You can sort the fields by clicking on the different column headers. The GRIB Examiner can be customised – see the extra tasks for this chapter, as this is an advanced topic.

Filtering GRIB Data

Metview provides powerful data filtering capabilities. Let’s take our ztuv.grb file and extract the forecast and analysis data separately from it.

Create a new GRIB Filter icon. This time we’ll rename it from within the icon editor (just to show an alternative way to rename an icon). Edit the icon, and notice the button to the right of the information panel; click it and change the name of the icon to t_3day_fc - we will use this icon to extract only the 3-day forecast data for temperature.

First, we specify the input data. Drag the ztuv.grb icon into the Data field of the editor. This is an icon field – an area where you can drop other icons. Now set the following parameters to extract just the 3-day (72-hour) temperature forecast:

Visualise this icon and verify that it returns only the data we expect.

Now create a new GRIB Filter icon, rename it t_an and use it to extract only the temperature analysis data:

It is quite often the case that GRIB data comes as several fields in the same file, and using the GRIB Filter icon is an easy way to extract just the fields you want without making copies of the file. This icon also has some parameters to perform some post-processing on the data, which we will cover in Processing Data.

Plotting Grid Values

We will now plot the actual grid values. Create a new Contour icon and rename it grid_10x10. Edit it and find the set of parameters close to the bottom of the editor which control the plotting of grid values. Activate grid value plotting, set it to plot both values and markers, and set the lat/lon frequency each to 10. Visualise a scalar field (temperature or geopotential) from ztuv.grb and apply the new visdef icon to it - you will see every 10th grid point plotted. If you wish, you can also deactivate the isolines by setting Conotur to Off.

If you zoom into smaller areas, you may want to see every grid point - duplicate grid_10x10 and call it grid_1x1. Set the lat/lon frequency to 1 - one fast way to do this is to click on the blue ‘revert’ button next to the parameter. This button does two things: it indicates that a parameter has been altered from its default, and it restores the parameter to its default when clicked.

Cursor Data

As already seen in A Simple Visualisation, the Cursor Data can also be used to inspect grid point values. Activate it and check it against the plotted values!