FM 92-XII GRIB General Regularly-distributed Information in Binary Form

1. A reference value is normally the minimum value of the data set which is represented.



2. For grid-point values, complex packing features are intended to reduce the whole size of the GRIB message (data compression without loss of information with respect to simple packing). The basic concept is to reduce data size thanks to local redundancy. This is achieved just before packing, by splitting the whole set of scaled data values into groups, on which local references (such as local minima) are removed. It is done with some overhead, because extra descriptors are needed to manage the groups characteristics. An optional pre-processing of the scaled values (spatial differencing) may also be applied before splitting into groups, and combined methods, along with use of alternate row scanning mode, are very efficient on interpolated data.
3. For spectral data, complex packing is provided for better accuracy of packing. This is because many spectral coefficients have small values (regardless of sign), especially for large wave numbers. The first principle is to not pack a subset of coefficients, associated with small wave numbers so that the amplitude of the packed coefficients is reduced . The second principle is to apply an operator to the remaining part of the spectrum: with appropriate tuning it leads to a more homogeneous set of values to pack.
4. The original data value Y (in the units of code table 4.2) can be recovered with the formula:
   
   Y * 10D= R + (X1+X2) * 2E
   For simple packing and all spectral data
   E = Binary scale factor,
   D = Decimal scale factor,
   R = Reference value of the whole field,
   X1 = 0,
   X2 = Scaled (encoded) value.
   
   For complex grid point packing schemes, E, D, and R are as above, but X1 = Reference value (scaled integer) of the group the data value belongs to, X2 = Scaled (encoded) value with the group reference value (XI) removed.

1. If octet 6 is not zero, octets 15-xx (15-nn if octet 11 is zero) may not be supplied. This should be documented with all bits set to 1 (missing value) in Grid Definition Template Number.



2. An optional list of numbers defining number of points is used to document a quasi-regular grid, where the number of points may vary from one row to another (row being defined as adjacent points in a coordinate line, so this is dependent from data layout). In such a case, octet 11 is non zero, and gives the number of octets on which each number of points is encoded. For all other cases, such as regular grids, octets 11 and 12 are zero and no list is appended to the Grid Definition Template.



3. If a list of numbers defining number of points is present, it is appended at the end of Grid Definition Template (or directly after Grid Definition Template Number if template is missing), the length of the list is given by the grid definition. When the Grid Definition Template is present, the length is given according to bit 3 of scanning mode flag octet (length is Nj or Ny for flag value 0). List ordering is implied by data scanning.



4. Depending on code value given in octet 12, the list of numbers defining number of points corresponds either to the coordinate lines as given in the grid definition, or to a full circle.

1. Coordinates values are intended to document the vertical discretisation associated with model data on hybrid coordinate vertical levels. A number of zero in octets 6-7 indicates that no such values are present. Otherwise, the number corresponds to the whole set of values.



2. Hybrid systems, in the context, employ a means of representing vertical coordinates in terms of a mathematical combination of pressure and sigma coordinates. When used in conjunction with a surface pressure field and an appropriate mathematical expression, the vertical coordinate parameters may be used to interpret the hybrid vertical coordinate.



3. Hybrid coordinate values, if present, should be encoded in IEEE 32-bit floating point format. They are intended to be encoded as pairs.

1. If octet 6 is not zero, the length of the Section is 6 and octets 7-nn are not present.

1. Basic angle of the initial production domain and subdivisions of this basic angle are provided to manage cases where the recommended unit of 10-6 degrees is not applicable to describe the extreme longitudes and latitudes, and direction increments. For these last six descriptors, unit is equal to the ratio of the basic angle and the subdivisions number. For ordinary cases, zero and missing values should be coded, equivalent to respective values of 1 and 106 (10-6 degrees unit).
2. For data on a quasi-regular grid, in which all the rows or columns do not necessarily have the same number of grid points, either Ni (Octets 31-34) or Nj (Octets 35-38) and the corresponding Di (Octets 64-67) or Dj (Octets 68-71) shall be coded with all bits set to 1 (missing). The actual number of points along each parallel or meridian shall be coded in the octets immediately following the Grid Definition Template (Octets [xx+1] nn), as described in the description of the Grid Definition Section.
3. A quasi-regular grid is only defined for appropriate grid scanning modes. Either rows or columns, but not both simultaneously, may have variable numbers of points. The first point in each row (column) shall be positioned at the meridian (parallel) indicated by Octets 47-54. The grid points shall be evenly spaced in latitude (longitude).

9. Basic angle of the initial production domain and subdivisions of this basic angle are provided to manage cases where the recommended unit of 10-6 degrees is not applicable to describe the extreme longitudes and latitudes, and direction increments. For these last six descriptors, unit is equal to the ratio of the basic angle and the subdivisions number. For ordinary cases, zero values should be coded, equivalent to respective values of 1 and 106 (10-6 degrees unit).



10. The stretching is defined by three parameters:
1. The latitude in degrees (measured in the model coordinate system) of the "pole of stretching";
2. The longitude in degrees (measured in the model coordinate system) of the "pole of stretching"; and
3. The stretching factor C in units of 10-6 represented as an integer. The stretching is defined by representing data uniformly in a coordinate system with longitude Y and latitude X1, where:
   
   

                                (1 - C2) + (1 + C2) sin X
                       X1 = sin-1 -----------------------------
                                (1 + C2) + (1 - C2) sin X 
   

   
   and Y and X are longitude and latitude in a coordinate system in which the "pole of stretching" is the northern pole. C = 1 gives uniform resolution, while C > 1 give enhanced resolution around the pole of stretching.

1. Basic angle of the initial production domain and subdivisions of this basic angle are provided to manage cases where the recommended unit of 10-6 degrees is not applicable to describe the extreme longitudes and latitudes, and direction increments. For these last six descriptors, unit is equal to the ratio of the basic angle and the subdivisions number. For ordinary cases, zero values should be coded, equivalent to respective values of 1 and 106 (10-6 degrees unit).
2. See Note (2) under Grid Definition Template 3.1 - Rotated Latitude/longitude (or equidistant cylindrical, or Plate Carree)
3. See Note (2) under Grid Definition Template 3.2 - Stretched Latitude/longitude (or equidistant cylindrical, or Plate Carree)

1. Limited to the range of 0 to 90 degrees; if the angle of orientation of the grid is neither 0 nor 90 degrees, Di and Dj must be equal to each other.
2. Grid lengths are in units of 10-3 m, at the latitude specified by LaD.

1. The resolution flag (bit 3-4 of Flag table 3.3) is not applicable.
2. LoV is the value of the meridian which is parallel to the Y-axis (or columns of the grid) along which latitude increases as the Y-coordinate increases (the orientation longitude may or may not appear on a particular grid).
3. Grid length is in units of 10-3 m at the latitude specified by LaD.
4. Bit 2 of the projection flag is not applicable to the polar stereographic projection.

1. Grid lengths are in units tenths of 10-3 m, at the latitude specified by LaD.
2. If Latin 1 = Latin 2, then the projection is on a tangent cone.
3. The resolution flags (bits 3-4 of Flag Table 3.3) are not applicable
4. LoV is the value of the meridian which is parallel to the Y-axis (or columns of the grid) along which latitude increases as the Y-coordinate increases (the orientation longitude may or may not appear on a particular grid).

1. Basic angle of the initial production domain and subdivisions of this basic angle are provided to manage cases where the recommended unit of 10-6 degrees is not applicable to describe the extreme longitudes and latitudes, and direction increments. For these last six descriptors, unit is equal to the ratio of the basic angle and the subdivisions number. For ordinary cases, zero values should be coded, equivalent to respective values of 1 and 106 (10-6 degrees unit).
2. The number of parallels between a pole and the equator is used to establish the variable (Gaussian) spacing of the parallels; this value must always be given.

1. Basic angle of the initial production domain and subdivisions of this basic angle are provided to manage cases where the recommended unit of 10-6 degrees is not applicable to describe the extreme longitudes and latitudes, and direction increments. For these last six descriptors, unit is equal to the ratio of the basic angle and the subdivisions number. For ordinary cases, zero values should be coded, equivalent to respective values of 1 and 106 (10-6 degrees unit).
2. The number of parallels between a pole and the equator is used to establish the variable (Gaussian) spacing of the parallels; this value must always be given.
3. See Note (2) under Grid Definition Template 3.1 - Rotated Latitude/longitude grid (or equidistant cylindrical, or Plate Carree)

1. Basic angle of the initial production domain and subdivisions of this basic angle are provided to manage cases where the recommended unit of 10-6 degrees is not applicable to describe the extreme longitudes and latitudes, and direction increments. For these last six descriptors, unit is equal to the ratio of the basic angle and the subdivisions number. For ordinary cases, zero values should be coded, equivalent to respective values of 1 and 106 (10-6 degrees unit).
2. The number of parallels between a pole and the equator is used to establish the variable (Gaussian) spacing of the parallels; this value must always be given.
3. See Note (2) under Grid Definition Template 3.2 -Stretched Latitude/longitude (or equidistant cylindrical, or Plate Carree)

1. Basic angle of the initial production domain and subdivisions of this basic angle are provided to manage cases where the recommended unit of 10-6 degrees is not applicable to describe the extreme longitudes and latitudes, and direction increments. For these last six descriptors, unit is equal to the ratio of the basic angle and the subdivisions number. For ordinary cases, zero values should be coded, equivalent to respective values of 1 and 106 (10-6 degrees unit).
2. The number of parallels between a pole and the equator is used to establish the variable (Gaussian) spacing of the parallels; this value must always be given.
3. See Note (2) under Grid Definition Template 3.1 -Rotated Latitude/longitude (or equidistant cylindrical, or Plate Carree)
4. See Note (2) under Grid Definition Template 3.2 -Stretched Latitude/longitude (or equidistant cylindrical, or Plate Carree)

1. The pentagonal representation of resolution is general. Some common truncations are special cases of the pentagonal one:
   Triangular M = J = K
   Rhomboidal K = J + M
   Trapezoidal K = J, K > M
   

1. See Note (1) under Grid Definition Template 3.50 - Spherical harmonic coefficients
2. See Note (2) under Grid Definition Template 3.1 - Rotated Latitude/longitude grid (or equidistant cylindrical, or Plate Carree)

1. See Note (1) under Grid Definition Template 3.50 - Spherical harmonic coefficients
2. See Note (2) under Grid Definition Template 3.20 - Stretched Latitude/longitude grid (or equidistant cylindrical, or Plate Carree)

1. It is assumed that the satellite is at its nominal position, i.e., it is looking directly at its sub-satellite point.
2. Octets 46-49 shall be set to all ones (missing) to indicate the orthographic view (from infinite distance)
3. It is the angle between the increasing Y-axis and the meridian 80E if the sub-satellite point is the North Pole; or the meridian 0 if the sub-satellite point is the South Pole.
4. The apparent angular size of the Earth will be given by 2 * Arcsin (106 )/Nr).
5. For orthographic view from infinite distance, the value of Nr should be encoded as missing (all bits set to 1).
6. The horizontal and vertical angular resolutions of the sensor (Rx and Ry), needed for navigation equation, can be calculated from the following:
   Rx = 2 * Arcsin (106 )/Nr)/ dx
   Ry = 2 * Arcsin (106 )/Nr)/ dy
   

1. For more details see appendix II to the Manual of Codes, Vol. I, Part B- definition of the triangular grid based on an icosahedron
2. The origin of the grid is an icosahedron with 20 triangles and 12 vertices. The triangles are combined to nd quadrangles, the so-called diamonds (e.g. if nd = 10, two of the icosahedron triangles form a diamond, and if nd = 5, 4 icosahedron triangles form a diamond). There are two resolution values called n2 and n3 describing the division of each triangle side. Each triangle side is divided into ni equal parts where ni = 3**n3 * 2**n2 with n3 either equal to 0 or to 1. In the example of appendix II, the numbering order of the rectangles is anti-clockwise with a view from the pole point on both hemispheres. Diamonds 1 to 5 are northern hemisphere and diamonds 6 to 10 are Southern Hemisphere.
3. The exponent of 3 for the number of divisions of triangle sides is used only with a value of either 0 or 1.
4. The total number of grid points for one global field depends on the grid point position. If e.g. the grid points are located at the vertices of the triangles nt = (ni + 1) * (ni + 1) * nd since grid points at diamond edges are contained in both adjacent diamonds and for the same reason the pole points are contained in each of the five adjacent diamonds.

1. A data bin is a data point representing the volume centered on it.

1. Hours greater than 65534 will be coded as 65534.

1. Hours greater than 65534 will be coded as 65534.

1. Hours greater than 65534 will be coded as 65534.

1. Hours greater than 65534 will be coded as 65534.

1. Hours greater than 65534 will be coded as 65534.

1. Hours greater than 65534 will be coded as 65534.

1. Hours greater than 65534 will be coded as 65534.

1. Hours greater than 65534 will be coded as 65534.

1. Hours greater than 65534 will be coded as 65534.
2. The reference time in section 1 and the forecast time together define the beginning of the overall time interval.
3. An increment of zero means that the statistical processing is the result of a continuous (or near continuous) process, not the processing of a number of discrete samples. Examples of such continuous processes are the temperatures measured by analogue maximum and minimum thermometers or thermographs, and the rainfall measured by a rain gauge.
4. The reference and forecast times are successively set to their initial values plus or minus the increment, as defined by the type of time increment (one of octets 46, 58, 70 ...). For all but the innermost (last) time range, the next inner range is then processed using these reference and forecast times as the initial reference and forecast time.

1. This form of representation enables a matrix of values to be depicted at each grid point; the two dimensions of the matrix may represent coordinates expressed in terms of two elemental parameters (e.g. direction and frequency for wave spectra). The numeric values of these coordinates, beyond that of simple subscripts, can be given in a functional form, or as a collection of explicit numbers.
2. Some simple coordinate functional forms are tabulated in Code Table 5.2. Where a more complex coordinate function applies, the coordinate values shall be explicitly denoted by the inclusion of the actual set of values rather than the coefficients. This shall be indicated by a code figure 0 from Code Table 5.2; the number of explicit values coded shall be equal to the appropriate dimension of the matrix for which values are presented and they shall follow octet 36 in place of the coefficients.
3. Matrix bit maps will be present only if indicated by octet 22. If present, there shall be one bit map for each grid point with data values, as defined by the primary bit map in Section 6, each of length (NR*NC) bits: a bit set to 1 will indicate a data element at the corresponding location within the matrix. Bit maps shall be represented end-to-end, without regard for octet boundaries; the last bit map shall, if necessary, be followed by bits set to zero to fill any partially used octet.
4. Matrices restricted to scanning in the + i direction (left to right) and in the -j direction (top to bottom).

1. Group lengths have no meaning for row by row packing, where groups are coordinate lines (so the Grid Description Section and possibly the Bit-map Section are enough); for consistency associated field width and reference should then be encoded as 0.
2. For row by row packing with a bit-map, there should always be as many groups as rows. In case of rows with only missing values, all associated descriptors should be coded as zero.
3. Management of widths into a reference and increments, together with management of lengths as scaled incremental values, are intended to save descriptor size (which is an issue as far as compression gains are concerned).
4. Management of explicitly missing values is an alternative to bit-map use within Section 6; it is intended to reduce the whole GRIB message size.
5. There may be two types of missing value(s), such as to make a distinction between static misses (for instance, due to a land/sea mask) and occasional misses.
6. As an extra option, substitute value(s) for missing data may be specified. If not wished (or not applicable), all bits should be set to 1 for relevant substitute value(s).
7. If substitute value(s) are specified, type of content should be consistent with original field values (floating-point -and then IEEE 32-bit encoded-, or integer).
8. If primary missing values are used, such values are encoded within appropriate group with all bits set to 1 at packed data level.
9. If secondary missing values are used, such values are encoded within appropriate group with all bits set to 1, except the last one set to 0, at packed data level.
10. A group containing only missing values (of either type) will be encoded as a constant group (null width, no associated data) and the group reference will have all bits set to 1 for primary type, and all bits set to 1, except the last bit set to 0, for secondary type.
11. If necessary, group widths and/or field width of group references may be enlarged to avoid ambiguities between missing value indicator(s) and true data.
12. The group width is the number of bits used for every value in a group.
13. The group length (L) is the number of values in a group.
14. The essence of the complex packing method is to subdivide a field of values into NG groups, where the values in each group have similar sizes. In this procedure, it is necessary to retain enough information to recover the group lengths upon decoding. The NG group lengths for any given field can be described by Ln = ref + Kn * len_inc, n = 1,NG, where ref is given by octets 38-41 and len_inc by octet 42. The NG values of K (the scaled group lengths) are stored in the Data Section, each with the number of bits specified by octet 47. Since the last group is a special case which may not be able to be specified by this relationship, the length of the last group is stored in octets 43-46.

1. Spatial differencing is a pre-processing before group splitting at encoding time. It is intended to reduce the size of sufficiently smooth fields, when combined with a splitting scheme as described in Data Representation Template 5.2. At order 1, an initial field of values f is replaced by a new field of values g, where g1 = f1, g2 = f2 f1, , gn = fn fn-1. At order 2, the field of values g is itself replaced by a new field of values h, where h1 = f1, h2 = f2, h3 = g3 g2, , hn = gn gn-1. To keep values positive, the overall minimum of the resulting field (either gmin or hmin) is removed. At decoding time, after bit string unpacking, the original scaled values are recovered by adding the overall minimum and summing up recursively.
2. For differencing of order n, the first n values in the array that are not missing are set to zero in the packed array. These dummy values are not used in unpacking.

1. Removal of the real part of (0,0) coefficient from packed data is intended to reduce the variability of the coefficients, in order to improve packing accuracy.
2. For some spectral representations, the (0,0) coefficient represents the mean value of the parameter represented.

1. The unpacked subset is a set of values defined in the same way as the full set of values (on a spectrum limited to JS , KS and MS), but on which scaling and packing are not applied. Associated values are stored in octets 6 onwards of Section 7.
2. The remaining coefficients are multiplied by (n*(n+1))P, scaled and packed. The operator associated with this multiplication is derived from the laplacian operator on the sphere.
3. The retrieval formula for a coefficient of wave number n is then: Y = (R+X*2E)*10-D* (n*(n+1))-P where X is the packed scaled value associated with the coefficient

1. Group descriptors mentioned above may not be physically present; if associated field width is 0.

1. Group descriptors mentioned above may not be physically present; if associated field width is 0.
2. Group lengths have no meaning for row by row packing; for consistency associated field width should then be encoded as 0. So no specific test for row by row case is mandatory at decoding software level to handle encoding/decoding of group descriptors.
3. Scaled group lengths, if present, are encoded for each group. But the true last group length (unscaled) should be taken from Data Representation Template.
4. For groups with a constant value, associated field width is 0, and no incremental data are physically present.

1. Referring to the notation in Note (1) of Data Representation Template 5.3, at order 1, the values stored in octets 6-ww are g1 and gmin. At order 2, the values stored are h1, h2, and hmin.
2. Extra descriptors related to spatial differencing are added before the splitting descriptors, to reflect the separation between the 2 approaches. It enables to share software parts between cases with and without spatial differencing.
3. The position of overall minimum after initial data values is a choice that enables less software management.
4. Overall minimum will be negative in most cases. First bit should indicate the sign: 0 if positive, 1 if negative.

1. Values ordering within the unpacked subset is defined according to the source of grid definition associated with the data
2. Number of octets associated with each value of the unpacked subset (I) is defined in Code Table 5.7, according to the actual value in octet 35 of Data Representation Template 5.51
3. Values ordering within the packed data is done according to the source of grid definition, skipping the values processed in the unpacked subset

1. i direction: west to east along a parallel or left to right along an X-axis
2. j direction: south to north along a meridian, or bottom to top along a Y-axis
3. If bit number 4 is set, the first row scan is as defined by previous flags

Flag Table 3.5: Projection Centre

Code Table 3.6: Spectral data representation type

Not Defined

Code Table 3.7: Spectral data representation mode

Code table 3.8: Grid point position

Flag table 3.9:
Numbering order of diamonds as seen from the corresponding pole

Flag table 3.10: Scanning mode for one diamond

Code table 3.11: Interpretation of list of
numbers defining number of points

CODE AND FLAG TABLES USED IN SECTION 4

• 4.0 Product definition template number
• 4.1 Category of parameters by product discipline
• Product Discipline 0: Meteorological products
• Product Discipline 1: Hydrological products
• Product Discipline 2: Land surface products
• Product Discipline 3: Space products
• Product Discipline 10: Oceanographic products
• 4.2 Parameter number by product discipline and parameter category
• Meteorological products
• Parameter Category 0: Temperature
• Parameter Category 1: Moisture
• Parameter Category 2: Momentum
• Parameter Category 3: Mass
• Parameter Category 4: Short-wave Radiation
• Parameter Category 5: Long-wave Radiation
• Parameter Category 6: Cloud
• Parameter Category 7: Thermodynamic stability indices
• Parameter Category 13: Aerosols
• Parameter Category 14: Trace Gases
• Parameter Category 15: Radar
• Parameter Category 18: Nuclear/radiology
• Parameter Category 19: Physical atmospheric properties
• Parameter Category 253: Physical atmospheric properties
• Land surface products
• Parameter Category 0: Vegetation/Biomass
• Parameter Category 2: Soil products
• Space products
• Parameter Category 0: Image format products
• Parameter Category 1: Quantitative products
• Oceanographic products
• Parameter Category 0: Waves
• Parameter Category 1: Currents
• Parameter Category 2: Ice
• Parameter Category 4: Surface properties
• Parameter Category 5: Sub-surface properties
• 4.3 Type of generating process
• 4.4 Indicator of unit of time range
• 4.5 Fixed surface types and units
• 4.6 Type of ensemble forecast
• 4.7 Derived forecast
• 4.8 Clustering Method
• 4.9 Probability Type
• 4.10 Type of statistical processing
• 4.11 Type of time intervals
• 4.12 Operating mode
• 4.13 Quality control indicator
• 4.14 Clutter filter indicator
• 4.201 Precipitation type
• 4.202 Precipitable water category
• 4.203 Cloud type
• 4.204 Thunderstorm coverage
• 4.205 Aerosol Type
• 4.206 Volcanic ash
• 4.207 Icing
• 4.208 Turbulence
• 4.209 Planetary boundary layer regime
• 4.210 Contrail intensity
• 4.211 Contrail engine type
• 4.212 Land use
• 4.213 Soil type

Code Table 4.0: Product Definition Template Number