...
| Panel | ||
|---|---|---|
| ||
Change the value of You might add the mslp or z500 fields to this plot. e.g. Or produce two plots, one with PV at 320K, the other with z500 or MSLP and put them side by side on the screen. |
| Panel | ||||
|---|---|---|---|---|
| ||||
Compare From the animation of the z500 and mslp fields with : (as in Figure 1. from Pantillon et al.) Q. When does the cut-off low form (see z500)? |
Task 3: Changing geographical area
...
Use the hres_to_an_diff.mv icon and plot the difference map between the HRES forecast and the analysis first for z500 and then mslp (change plot1 from z500 to mslp).
| Panel | ||
|---|---|---|
| ||
Q. What differences can be seen? |
...
Visualising ensemble forecasts can be done in various ways. During this exercise we will use a number of visualisation techniques in order to understand the errors and uncertainties in the forecast,
Key parameters: MSLP, z500, and z500. And total precipitation (tp) over France. We suggest concentrating on viewing these fields. If time, visualize other parameters (e.g. PV320K).
...
Right-click the ens_rmse.mv icon, select 'Edit' and plot the curves first for 'mslp' and then for 'z500' (change the param field to mslp, run the macro and then change to z500 and run again).
Change 'expID' for your choice of ensemble.
...
Use the ens_to_an.mv icon and first plot the MSLP and then z500 (set param to mslp, run the macro, then change param to z500 and run again).
This will produce plots showing: the mean of of all the ensemble forecasts, the spread of the ensemble forecasts, the operational HRES deterministic forecast and the analysis.
Change 'expId' if required.
Animate this plot to see how the spread grows.
...
Use the ens_to_an_runs_spag.mv icon. Plot and animate either the MSLP and or z500 fields using your suitable choice for the contour level. Find a value that highlights the low pressure centres. Note that not all members may reach the low pressure set by the contour.
| Info |
|---|
If the contour value is not set correctly, no lines will appear. Use the 'cursor data' icon |
The red contour line shows the control forecast of the ensemble.
...
There are two icons to use, stamp.mv and stamp_diff.mv.
Use stamp.mv to first plot the MSLP and then z500 fields in the ensemble .(set param='mslp', run the macro, then change to 'z500' and run again).
The stamp map is slow to The stamp map is slow to plot as it reads a lot of data. Rather than animate each forecast step, a particular date can be set by changing the 'steps' variable.
...
It is usual to create clusters from z500 as it represents the large-scale flow and is not a noisy field. However, for this particular case study, the stamp map of 'tp' (total precipitation) over France is also very indicative of the distinct forecast scenarios. You might also try using other fields, such as 'mslp' or 'pv320K' to compare.
| Panel | ||
|---|---|---|
| ||
Right-click 'ens_oper_cluster.example.txt' and select Edit (or make a duplicate) The file contains two example lines:
The first line defines the list of members for 'Cluster 1': in this example, members 2, 3, 4, 9, 22, 33, 40. The second line defines the list of members for 'Cluster 2': in this example, members 10, 11, 12, 31, 49. Change these two lines!. You can create multiple cluster definitions by using the 'Duplicate' menu option to make copies of the file for use in the plotting macros.. The filename is important! |
...
| Panel | ||
|---|---|---|
| ||
Use the clusters of ensemble members you have created in Change If you are looking at the 2016 reforecast, then make sure your file is called ens_2016_cluster.example.txt. Replot ensembles:RMSE: plot the RMSE curves using Stamp maps: the stamp maps will be reordered so the ensemble members will be grouped according to their cluster. Applies to Spaghetti maps: with clusters enabled, two additional maps are produced which show the contour lines for each cluster. |
...
A quantitative way of clustering an ensemble is by a principal component analysis using empirical orthogonal functions. These are computed from the differences between the ensemble members and the ensemble mean, then computing the eigenvalues and eigenfunctions of these differences (or variances) over all the members such that the difference of each member can be expressed as a linear combination of these eigenfuctions, also known as empirical orthogonal functions (EOFs).
Although geopotential height at 500hPa at 00 24/9/2012 is used in the paper by Pantillon et al. as it gives the best results, the steps described below can be used for any parameter at any step.
...
The eof.mv macro computes the EOFs and the clustering.
| Warning |
|---|
Always first use the Otherwise cluster_to_an.mv and other plots with clustering enabled will fail or plot with the wrong clustering of ensemble members. If you change step or ensemble, recompute the EOFS and cluster definitions using eof.mv. Note however, that once a cluster has been computed, it can be used for all steps with any parameter. |
...
| Panel | ||
|---|---|---|
| ||
Q. How similar is the PCA computed clusters to your manual clustering? |
| Panel | ||||||
|---|---|---|---|---|---|---|
| ||||||
To change the number of clusters created by the EOF analysis, find the file in the folder 'base' called base_eof.mv. Edit this file and near the top, change:
to
then select 'File' and 'Save' to save the changes. Now if you run the You can use the 3 clusters in the
would plot the mean of the members in the first and the third clusters (it's not possible to plot all three clusters together). |
| Panel | ||
|---|---|---|
| ||
For those interested: The code that computes the clusters can be found in the Python script: This uses the 'ward' cluster method from SciPy. Other cluster algorithms are available. See http://docs.scipy.org/doc/scipy/reference/generated/scipy.cluster.hierarchy.linkage.html#scipy.cluster.hierarchy.linkage The python code can be changed to a different algorithm or the more adventurous can write their own cluster algorithm! |
...