CCSD3ZF0000100000001NJPL3IF0PDS200000001 = SFDU_LABEL RECORD_TYPE = FIXED_LENGTH RECORD_BYTES = 80 SPACECRAFT_NAME = MAGELLAN TARGET_NAME = VENUS OBJECT = TEXT PUBLICATION_DATE = 1991-09-03 NOTE = "COMMENTS ON MAGELLAN MIDR CD-ROMS" END_OBJECT END COMMENTS ON MAGELLAN MIDR CD-ROMS 1. Purpose of This Document This document is a cumulative summary of comments concerning the Magellan MIDR data set published on CD-ROMs. It contains comments on all the MIDR CD-ROMs published so far. Comments include information of interest to users of the data and notes about any known errors on the CD-ROMs. 2. General Comments 2.1 Normalization of Data Note that the eight-bit digital numbers (DN's) for each MIDR pixel on the CD-ROM's were computed by taking the estimated backscatter cross sections, dividing the values by the value for the Muhleman scattering law at the same incidence angles, converting to decibels, and scaling to bytes. The Muhleman Law has the following expression: sigma naught = A cos(i) / (sin(i) + B cos(i)) ** 3 Values for A and B were meant to be 0.0188 and 0.111, respectively Through an error, the A value used for the Muhleman Law estimate was 0.0118. As a result of this mistake, the pixel values are systematically about 10 DN (2 dB) greater than they should be according to the planned algorithm. Also, because of an indexing error, the incidence angles used in the above equation were 0.5 degrees lower than the actual angles. 2.2 VICAR2 Labels of Browse Images The browse versions of MIDRs were created by reducing the original 7168 by 8192 MIDR by a factor of 8. The VICAR2 labels for the browse images do not accurately reflect the size reduction, in that parameters PROJSAMP, PIXSIZ, and SPECLINE still pertain to the original MIDR size. 2.3 Data Feathering MIDR products on the CD-ROM's, in some cases, have had a feathering algorithm applied to suppress banding due to BIDR differences. The file CONTENTS.TAB includes a column (column no. 31) that specifies whether or not feathering has been applied. A "C" in this column indicates correction by feathering, while an "R" means not feathered. 3. Comments on volume MG_0001, version 2 3.1 Difference between MG_0001 file structure and Standard file structure The files MCUMDIR and CONTENTS are in the root directory of MG_0001. They are in subdirectory INDEX of MG_0002 and subsequent CD's. In addition, MG_0001 has a label file PMIDRDS.LBL in the LABEL directory. This file is replaced by NPMIDRDS.LBL and SPMIDRDS.LBL in the LABEL subdirectory of MG_0002 and subsequent CD's. For MG_0002 and subsequent CD's, a number of other label and text files (including AAREADME.TXT) have been updated to reflect these and a few other minor changes. 3.2 Difference between Standard and Engineering F-MIDRs The ten F-MIDRs on this CD-ROM are not standard products; rather, they are engineering test products generated before delivery of operational software. There are three primary differences between Engineering F-MIDRs and standard F-MIDRs. First, Engineering F-MIDRs consist of 20 orbits whereas standard F-MIDRs contain 24 orbits. Second, Engineering F-MIDRs are not necessarily complete on the trailing edge and may contain missing orbits within the body of the mosaic that have since been processed. The gap at the trailing edge of some of the F-MIDRs is a start of mapping gap and is present because processing began with orbit 376. Gaps in the bodies of the F-MIDRs are present because the Image Data Processing Team (IDPT) was instructed not to wait until all orbits had completed the processing queue. Therefore, orbits that the SAR Data Processing Team (SDPT) had trouble processing appear as gaps in the Engineering F-MIDRs even though the orbit has now been processed. 3.3 Error in VICAR2 Keyword PROJSAMP This CD-ROM has an error in the VICAR2 labels for the mosaic framelets. Specifically, the value for PROJSAMP does not vary from framelet to framelet. 3.4 VICAR2 Labels of MIDR Framelets Some of the MIDR framelet files have a VICAR2 label embedded at the end of the file, in addition to the normal VICAR2 label at the beginning of the file. This extra label is not documented in the PDS detached label. 3.5 Product ID in VICAR2 Label In two of the data subdirectories, the product ID given in the VICAR2 embedded labels is not correct. Specifically, the VICAR2 label has an N for north latitude instead of an S for south latitude. The correct IDs are as follows: directory product ID in VICAR2 labels correct product ID --------- --------------------------- ------------------ [F05S335] F-MIDR.05N335 F-MIDR.05S335 [F40S342] F-MIDR.40N342 F-MIDR.40S342 3.6 Error in BROWSE.LBL Files The PDS label files for the browse images on this CD-ROM (all named BROWSE.LBL) contain incorrect entries in the IMAGE object. The keyword LINES, which shows a value of 1024, should have the value 896. The keyword LINE_SAMPLES, which shows a value of 896, should have the value 1024. 4. Comments on Volume MG_0003, version 1 The F-MIDR products, F50S345, F55N337, F60N334, F65N330 and F75N332, on CD-ROM MG_0003 are the standard product replacements for five of the engineering test product F-MIDRD's on MG_0001. 5. Comments on GEO.LBL for Volumes MG_0001 to MG_0022 The PDS label file GEO.LBL on Magellan CD's volumes 1 to 22 contain incorrect DESCRIPTION keyword entries for the objects MINIMUM_LONGITUDE and MAXIMUM_LONGITUDE. The correct entries for these objects are: OBJECT = COLUMN NAME = MINIMUM_LONGITUDE DATA_TYPE = REAL UNIT = DEGREE START_BYTE = 19 BYTES = 8 DESCRIPTION = "The minimum_longitude element specifies the westernmost (left_most) longitude of the feature." END_OBJECT = COLUMN OBJECT = COLUMN NAME = MAXIMUM_LONGITUDE DATA_TYPE = REAL UNIT = DEGREE START_BYTE = 28 BYTES = 8 DESCRIPTION = "The maximum_longitude element specifies the easternmost (right_most) longitude of the feature." END_OBJECT = COLUMN Previously, the DESCRIPTION keyword for MINIMUM_LONGITUDE referred to "easternmost" and the DESCRIPTION keyword for MAXIMUM_LONGITUDE referred to "westernmost". 6. Comments on Volume MG_0012 Framelet 56 for C1-MIDR 60N014 has zeros in approximately 6.5 records starting at line 605 sample 513 and continuing to line 611 sample 1024. This will result in a horizontal black band across the framelet and may cause problems with some automated processing algorithms. 7. Comments on Volume MG_0014 Framelet 36 for C1-MIDR 00N009 has zeros in approximately 7 records starting at line 316 sample 513 and continuing to line 323 sample 512. This will result in a horizontal black band across the framelet and may cause problems with some automated processing algorithms. 8. Comments on Volume MG_0015 Framelet 22 for F-MIDR 20N097 has zeros in approximately 6 records starting at line 360 sample 513 and continuing to line 366 sample 512. This will result in a horizontal black band across the framelet and may cause problems with some automated processing algorithms.