| Octet No. | Contents |
|---|---|
| 10 |
Parameter category (see Code
table 4.1) |
| 11 |
Parameter number (see Code table 4.2 and Notes 2 and 3) |
| 12 |
Partition Table Number (PTN) (see Notes 1 and 3) |
| 13 |
Number of Partitions (NP) (see Note 1) |
| 14 - (14+2NP-1) | Partition set (list all partition numbers in the partition) (see Code table 4.PTN and Note 1)) |
| (14+2NP) - (15+2NP) | Partition number (PN) (see Code table 4.PTN and Note 3)) |
| (16+2NP) | Type of generating process (see
Code table 4.3) |
| (17+2NP) | Background generating process identifier (defined by originating centre) |
| (18+2NP) | Analysis or forecast generating process identifier
(see Code ON388 Table A) |
| (19+2NP) - (20+2NP) | Hours after reference time of data cut-off after reference time (see Note 1) |
| (21+2NP) | Minutes of observational data cut-off after reference time |
| (22+2NP) | Indicator of unit of time range
(see Code table 4.4) |
| (23+2NP) - (26+2NP) | Forecast time in units defined by octet 18 |
| (27+2NP) | Type of first fixed surface (see Code table 4.5) |
| (28+2NP) | Scale factor of first fixed surface |
| (29+2NP) - (32+2NP) | Scaled value of first fixed surface |
| (33+2NP) | Type of second fixed surfaced (see Code table 4.5) |
| (34+2NP) | Scale factor of second fixed surface |
| (35+2NP) - (38+2NP) | Scaled value of second fixed surfaces |
| (39+2NP) | Type of ensemble forecast |
| (40+2NP) | Perturbation number |
| (41+2NP) | Number of forecast in ensemble |
| Notes: (1) A single partition with code value PN from the partition set composed by the NP partitions is represented in the template. The code values of the NP partitions are expressed in octets 14 to 14+2NP-1. The NP partitions are linked by the normalization formula stating the sum of all the NP partitions must be equal to a normalization term (N) on each point of the grid. (2) Only parameters expressing fractions or percentages can be used in this template. Code tables shall state clearly that they are meant to be used in partitioned parameters context. (3) The word "fraction" or the word "percentage" has to be explicitly used in the name of the parameter to refer to a normalization term N=1 in the case of "fraction" and N=100 in the case of percentage. |