Clusters, or groups,
of similar forecasts are computed within the ensemble of 51 (i.e.
the 50 perturbed forecasts plus the control forecast); For five
different areas of clustering - domains
- General European Area (75.00N 20.00W 30.00N 45.00E)
- North West Europe (70.00N 27.50W 40.00N 10.00E)
- North East Europe (72.50N 0.00W 50.00N 45.00E)
- South West Europe (57.50N 15.00W 32.50N 17.50E)
- South East Europe (57.50N 2.50E 32.50N 42.50E)
Up to 6 clusters are
produced, depending on the overall spread of the ensemble. The
measure of similarity between individual forecasts is based on the
geopotential at 500hPa over the particular domain area for the
period from day 5 to day 7, so that the clusters are fully
consistent within that range. For each cluster, the mean and
standard deviation of its elements are computed every 12 hours from
day 3 to day 7 for 500 hPa geopotential, 500 hPa temperature, 850
hPa temperature, and 1000 hPa geopotential (Table 5.3).
Extensions to the GRIB
code enable cluster products to contain details of the cluster
number and the numbers of the ensemble members from which a cluster
was formed (an ensemble number of 0 refers to the control
forecast).
The number of
clusters generated varies from day to day. Note that only the
generated clusters will be disseminated; There will be no "empty"
clusters.
* On pressure levels 1000/500 ** On
pressure levels 850/500
Table 5.3 Cluster
Mean and Cluster Standard Deviation products
|
FIELD CODE
|
MARS ABBREV.
|
FIELD NAME
|
UNITS
|
NO. OF BITS/VAL.
|
|
129
|
Z
|
Geopotential*
|
m2s-2
|
16
|
|
130
|
T
|
Temperature**
|
K
|
12
|
Horizontal
resolution
- 1.5 x 1.5 degrees
latitude-longitude grid, or multiples thereof
Verifying times
- T+072 to T+168
(inclusive) at 12 hour intervals
Cluster
Mean and Cluster Standard Deviation requirements
The following command
language parameters are specific for the cluster mean and cluster
standard deviation requirements. Other DIS parameters are listed
under the general EPS requirements above.
|
TYPE
|
=
|
CLUSTER MEAN
CLUSTER STANDARD
DEVIATION
|
[CM]
[CS]
|
|
DOMAIN
|
=
|
General European
Area
North West
Europe
North East
Europe
South West
Europe
South East
Europe
|
[G]
[A]
[B]
[C]
[D]
|
(75.0 -20.0 30.0
45.0)
(70.0 -27.5 40.0
10.0)
(72.5 0.0 50.0
45.0)
(57.5 -15.0 32.5
17.5)
(57.5 2.5 32.5
42.5)
|
|
NUMBER
|
=
|
OFF
|
|
LEVTYPE
|
=
|
PRESSURE LEVEL
|
[PL]
|
|
|
LEVELIST
|
=
|
1000/500
850/500
|
(PARAM=Z only)
(PARAM=T only)
|
|
GRID
|
=
|
a/b
(latitude-longitude grid)
a= grid resolution
(deg.) along a latitude line (i.e. East/West)
b= grid resolution
(deg.) along a longitude line (i.e. North/South)
1.5/1.5 or multiples
thereof;
|
|
STEP
|
=
|
* (hours)
l gives forecast
time step(s);
72/84/96/108/120/132/144/156/168
NB ALL, is not
permitted.
|
|
PARAM
|
=
|
Field name
abbreviation
*
NB ALL is not
permitted.
Available
parameters are listed in the Table 5.2.
|
GRIB
section 1 definition for Cluster Mean and Cluster Standard
Deviation products
By convention, ECMWF
always uses octet 41 to define which set of local definitions
follow. The set of local definitions for EPS Cluster Mean and
Cluster Standard Deviation fields:
- Table 5.4 GRIB
section 1 definition for Cluster Mean and Cluster Standard
Deviation products
-
|
Octet
|
Ksec1
element
|
Contents
|
|
41
|
37
|
ECMWF local GRIB use
definition identifier.
2 = Cluster means
and standard deviations
|
|
42
43
44-45
46-49
|
8
39
40
41
|
Class:
1 = Operations
Type:
14 = Cluster
means
15 = Cluster
standard deviations
Stream:
1035 = Ensemble
forecasts
Version number or
experiment identifier.
(4 ASCII characters,
right justified) Version number or experiment identifier.
(4 ASCII characters,
right justified)
|
|
50
51
52
|
42
43
44
|
Number: Cluster
number
Total: Total number
of clusters.
Clustering
method:
1 = maximum
linkage
2 = mixed
3 = small
linkage
|
|
54-55
56-57
|
45
46
|
Start timestep
considered when clustering
(units as defined
for forecast timesteps)
End timestep
considered when clustering
|
|
58-60
61-63
64-66
67-69
|
47
48
49
50
|
Northern latitude
(of domain of clustering in millidegrees)
Western longitude
(of domain of clustering in millidegrees)
Southern latitude
(of domain of clustering in millidegrees)
Eastern longitude
(of domain of clustering in millidegrees)
Number of cluster
to which the operational forecast
belongs
|
|
70
71
|
51
52
|
Number of cluster
to which the operational forecast
belongs
Number of cluster
to which the control forecast
belongs
|
|
72
73 - 328n1-nN
|
53
54 - (53+N)
|
N, the number of
forecasts belonging to the cluster
List of the N
ensemble forecast numbers used in the
cluster (one octet
per number); if the control forecast
has been used, this
will be listed as number 0. Any
unused octets at the
end of the list will be set to 0.
|
|