AVHRR Files - record format

Word # ---- Parameter --- Description

1 ---- LDBGN
Record number of the first row of the field (currently 2 for all fields since the documentation record requires only one record).
2 ---- SMGLAT
Minimum latitude included in field which is the bottom edge and first row of field.
3 ---- AXLAT
Maximum latitude included in field which is the top edge and last row of field.
4 ---- SMLONG
Minimum longitude included in field which is the left edge and first column of field.
5 ---- AXLONG
Maximum longitude included in field which is the right edge and last column of field (excluding the I.D. column).
6 ---- RES
Number of latitude/longitude degrees between each grid point.
7 ---- SMHOUR
Youngest time, in hours of the year, of observations used during last analysis, which becomes the oldest time allowed for the next analysis. If the difference between this time and time of next analysis is greater than the maximum time gap allowed, SMHOUR for the beginning of the next analysis is reduced to make the difference equal to the maximum time gap.
8 ---- HOURS
Oldest time, in hours of the year, of observations used during last analysis.
9 ---- TIMGAP
Number of hours between youngest and oldest times of observations used in analysis.
10 ---- MAXDAT
Maximum number of hours allowed in time period for observation times to be included in analysis.
11 ---- SMREL
Minimum reliability of observations to be used in analysis.
12 ---- AXREL
Maximum reliability of observations to be used in analysis.
13-22 ---- SORC(10)
List of source codes of observations to be used in analysis.
23-32 ---- OBTYPE(10)
List of observation types allowed to be used in analysis.
33 ---- NROWS
Number of rows (latitudinal parallels) included in field, excluding documentation record.
34 ---- NCOLS
Number of columns (longitudinal meridians) in field, including the I.D. column.
35 ---- IBLK
Number of rows or logical records per physical block.
36 ---- NWRDS
Number of fullwords (32 bits) allocated to each grid point.
37 ---- ISZ
Number of rows to be maintained in an array in core for optical thickness and analysis and calculation of gradients.
38 ---- ICENT
Center line within the array upon which calculations will be performed.
39-41 ---- LWT, LNT, LBT
Word number, length in bits, and starting bit location of optical thickness within a grid intersection information unit of an SST Field.
42-44 ---- LWG, LNG, LBG
Word number, length in bits, and starting bit location of Average Gradient.
45-47 ---- LWGXP, LNGXP, LBGXP
Word number, length in bits, and starting bit location of Gradient X+ direction.
48-50 ---- LWGXN, WNGXN, LBGXN
Word number, length in bits, and starting bit location of Gradient X+ direction.
51-53 ---- LWGYP, LNGYP, LBGYP
Word number, length in bits, and starting bit location of Gradient Y+ direction.
54-56 ---- LWGYN, LNGYN, LBGYN
Word number, length in bits, and starting bit location of Gradient Y- direction.
57-59 ---- LWPD, LNPD, LBPD
Word number, length in bits, and starting bit location of Physiographic Descriptor.
60-62 ---- LWNO, LNNO, LBNO
Word number, length in bits, and starting bit location of Number Observations.
63-65 ---- LWAGE, LNAGE, LBAGE
Word number, length in bits, and starting bit location of Age Recent Observation
66-68 ---- LWREL, LNREL, LBREL
Word number, length in bits, and starting bit location of Reliability.
69-71 ---- LWCLS, LNCLS, LBCLS
Word number, length in bits, and starting bit location of Class 1 Coverage.
72-74 ---- LWSXP, LNSXP, LBSXP
Word number, length in bits, and starting bit location of Spatial Covariance in the positive X direction.
75-77 ---- LWSXN, LNSXN, LBSXN
Word number, length in bits, and starting bit location of Spatial Covariance in the negative X direction.
78-80 ---- LWSYP, LNSYP, LBSYP
Word number, length in bits, and starting bit location of Spatial Covariance in the positive Y direction.
81-83 ---- LWSYN, LNSYN, LBSYN
Word number, length in bits, and starting bit location of Spatial Covariance in the negative Y direction.
84-86 ---- LWIND, LNIND, LBIND
Word number, length in bits, and starting bit location of Independent Temperature.
87-96 ---- GRDWTS(10)
Weight assigned to each grid unit, according to its distance from the grid intersection for which gradients are being calculated.
97 ---- NP
Number of grid points to be used in calculation of gradients.
98-117 ---- KMDST(10,2)
Look up table of gradient values and corresponding distances to be used in determining the search area for analysis.
118 ---- MKM
Number of paired entries in KMDST.
119-138 ---- H(10,2)
Look up table of gradient values and corresponding factors to be used in determining the new weight assigned to the observation temperature for analysis.
139 ---- MH
Number of paired entries in H.
140 ---- EXP
Exponent used in temperature analysis.
141 ---- FDX
Factor used in determining new weight assigned to the optical thickness observation for analysis.
142 ---- XCLASS
Factor used to place gradients into a class for Gradient Class Summary.
143 ---- DEL
Maximum number of optical thickness units that the new analysis temperature may differ from the previous optical thickness field value.
144 ---- MF
Factor applied to the previous optical thickness and reliability to determine the final optical thickness and its reliability.
145 ---- MSTAR
Factor applied to the combined observations optical thickness and weight in determining the new analysis optical thickness.
146 ---- MNSRCH
Minimum distance in kilometers to be searched for analysis observations.
147 ---- MXSRCH
Maximum distance in kilometers to be searched for analysis observations.
148 ---- BDEL
Maximum difference allowed between new analysis optical thickness and the previous one for the Class 1 Coverage Bit to be set to 1.
149 ---- FCWT
Maximum value that can be assigned as the reliability of the new analysis optical thickness.
150 ---- IYYY
Year of youngest time of observation data used (0-99).
151 ---- IYMM
Month of youngest time of observation data used (1-12).
152 ---- IYDD
Day of youngest time of observation data used (1-31).
153 ---- IYHH
Hour of youngest time of observation data used (0-23).
154 ---- IOYY
Year of oldest time of observation data used (0-99).
155 ---- IOMM
Month of oldest time of observation data used (1-12).
156 ---- IODD
Day of oldest time of observation data used (1-31).
157 ---- IOHH
Hour of oldest time of observation data used (0-23).
158 ---- ICURTM
Last time used in analysis.
Values are stored as real (IBM floating-point) or integer according to the format implied by the first letter of their label. (Parameters beginning with I,J,K,L,M,N are integer values.)

IBM360 Floating points

This floating point representation is not compatible with the IEEE floating point specification which is used by most modern computers. The IBM360 float was native to the now obsolete IBM 360/370 series of computers, and a few others. Since documentation of this representation is getting difficult to find, a summary is provided here.

The 4-byte floating point representation is divided into one byte (SEXP) for sign and exponent, followed by a 3-byte MANTISSA. The mantissa should be thought of as a six digit hexadecimal number, with a radix point preceding the first digit. The numeric value of the MANTISSA is the value of the integer with the same hex representation, divided by 0x1000000, and ranges from 0.0 to just less than 1.0. For a non-zero, normalized float, the first hex digit will range from 0x1 to 0xf.

In the SEXP byte, the high order bit is the sign of the number, 0 for plus and 1 for minus. The remaining seven bits are a biased exponent. After subtracting 0x40, the result is the number of hexidecimal digits (powers of 16) to shift the radix point right (if positive) or left (if negative). This means the range of the representation is from 16^63 to 16^(-64).


Table of Contents