Local definition 19: Extreme forecast index data
Local definitions
- Local definition 1
- Local definition 2
- Local definition 3
- Local definition 4
- Local definition 5
- Local definition 6
- Local definition 7
- Local definition 8
- Local definition 9
- Local definition 10
- Local definition 11
- Local definition 12
- Local definition 13
- Local definition 14
- Local definition 15
- Local definition 16
- Local definition 17
- Local definition 18
- Local definition 19
- Local definition 20
- Local definition 21
- Local definition 22
- Local definition 23
- Local definition 24
- Local definition 25
- Local definition 26
- Local definition 27
- Local definition 28
- Local definition 29
- Local definition 30
- Local definition 31
- Local definition 32
- Local definition 36
- Local definition 37
- Local definition 38
- Local definition 39
- Local definition 49
- Local definition 50
- Local definition 190
- Local definition 191
- Local definition 192
Octets | Key | Type | Contents |
41 | localDefinitionNumber | unsigned | ECMWF local GRIB use definition identifier. 19 = Extreme forecast index data |
42 | class | codetable | Class = 1, Operations |
43 | type | codetable | Type: 27 = extreme forecast index 28 = extreme forecast index control |
44-45 | stream | codetable | Stream = 1035 (ensemble forecast) |
46-49 | experimentVersionNumber | ksec1expver | Version number or experiment identifier. (4 ASCII characters, right justified) |
50 | number | unsigned | Zero, for compatibility with Mars labelling |
51 | ensembleSize | unsigned | Ensemble size. |
52 | versionNumberOfExperimentalSuite | unsigned | Before Feb 2006: Power of 10 used to scale climate weight = F Before Mar 2008: Version number of the Experimental Suite (1 - 255) From March 2008: No further change |
53-56 | implementationDateOfModelCycle | unsigned | Before Feb 2006: Weight * 10**F applied to climate month 1, stored as an integer Before Mar 2008: Implementation date of the model cycle (YYYYMMDDHH) From March 2008: No further change |
57-59 | numberOfReforecastYearsInModelClimate | unsigned | Before Feb 2006: First month used to build climate month 1, YYYYMM Before Mar 2008: Basetime of the EFI computation, HHMM (0 or 1200) From March 2008: Number of re-forecast years in the model climate (Currently set to 18) |
60-62 | numberOfDaysInClimateSamplingWindow | unsigned | Before Feb 2006: Last month used to build climate month 1, YYYYMM Before Mar 2008: Number of re-forecasts each year (XX) and time difference as number of days applied in the operational generation of the re-forecasts (YY), separated by 0, XX0YY From March 2008: Length (in number of days) of the climate sampling window centred over the day of the climate run (Currently set to 31) |
63-65 | sampleSizeOfModelClimate | unsigned | Before Feb 2006: First month used to build climate month 2, YYYYMM Before Mar 2008: First year of the climate period From March 2008: Sample size of the model climate (Number of re-forecast fields constituting the climate) |
66-68 | versionOfModelClimate | unsigned | Before Feb 2006: Last month used to build climate month 2, YYYYMM Before Mar 2008: Last year of the climate period From March 2008: Version of the model climate (Currently set to 0 = No climate files prepared for days between two re-forecast runs) |
69 | efiOrder | unsigned | EFI order For EFI this is set to 0 but for SOT (Shift of Tails) it has an important meaning. It shows the reference model climate percentile used for SOT computation. Therefore, for the lower tail SOT efiOrder=1 and for the upper tail SOT efiOrder=99. Having the octet 50 (number) we can figure out how SOT is computed, e.g. for the upper tail SOT efiOrder=99 and number=90 which shows that for SOT computation we use 90th and 99th model climate percentiles and 90th percentile of the forecast distribution |
70 | Set to 0 for this version of EFI(prior to March 2008) | ||
71-80 | Set to zero |