DESCRIPTION = "This file describes the format of the TABLE in the GVXIF.TAB file. See the GVXIF.LBL file for the full PDS label." OBJECT = COLUMN NAME = SAMPLE_COUNT DATA_TYPE = MSB_UNSIGNED_INTEGER START_BYTE = 1 BYTES = 2 DESCRIPTION = "The total number of image framelets used to compute the estimates of radar scattering properties given in this row of the table. The quantities given in later columns pertain only these framelets. Any framelet that is partially or completely contained inside this pixel is included. Each framelet is a slice of image data from the F-BIDR or C-BIDR files and is approximately 2 km in size along the image track. The framelets used in this row share a common observational geometry, given explicitly by the columns AZIMUTH_ANGLE, INCIDENCE_ANGLE, and POLARIZATION_ANGLE, and implicitly by the pixel address. This pixel may have entries in other rows, and if so at least one of the three angles will be different from this row." END_OBJECT = COLUMN OBJECT = COLUMN NAME = AZIMUTH_ANGLE DATA_TYPE = MSB_UNSIGNED_INTEGER START_BYTE = 3 BYTES = 2 OFFSET = 0 SCALING_FACTOR = 0.00549367 UNIT = DEGREE VALID_MINIMUM = 0 VALID_MAXIMUM = 360 DESCRIPTION = "The average azimuthal angle of the incident radiation of all image framelets used in this row. (An image framelet is a single SAR observation.) The azimuthal angle is defined in two ways, depending on the image map projection in use. For the Sinusoidal and Mercator map projections, it is defined as the local azimuth direction toward the spacecraft when viewed by an observer at the boresight intercept point on the planet surface, in degrees clockwise from North. For example, if the spacecraft appears to be due east of the observer, the azimuth angle is 90 degrees. Since this definition becomes useless near the poles, this field is set to zero above 85 degrees of latitude for the Sinusoidal projection. The Mercator projection does not extend to such high latitudes. For the Polar Stereographic map projection, the azimuth angle is expressed in the cartesian map coordinates rather than in geographic coordinates. This makes the azimuth angle more useful for interpretation because its meaning no longer varies with position. First, the azimuth direction is computed as above. Then, this direction is transformed to a direction in map coordinates; the transformed vector is parallel to the vector originating at the framelet and pointing in the azimuth direction. This vector is expressed in degrees clockwise from the +Y (up) direction on the map. For example, if the north polar projection has 0 degrees of longitude at the bottom, then a framelet at 90 degrees of longitude with a true azimuth of 90 degrees (spacecraft to the east) has a transformed azimuth of 0 degrees. The vector from the framelet to the spacecraft appears to point in the +Y direction on the map. The relationship between true azimuth and transformed azimuth is simple. For the north polar projection with 0 degrees longitude at the bottom, CARTESIAN_AZIMUTH = GEOGRAPHIC_AZIMUTH - LONGITUDE and for the south polar projection with 0 degrees longitude at the top, CARTESIAN_AZIMUTH = GEOGRAPHIC_AZIMUTH + LONGITUDE As discussed in the GVHDR.LBL file, the group of SAR observations used in this row forms a 'cohort'. Within a cohort, the azimuth angle of each observation falls within a single interval of size 360/N, where N is the value of XIF_COHORT_AZIMUTH_COUNT from the GVHDR file. We can reconstruct that interval by noting that the average azimuth angle of all the observations (given in this column) falls within the same interval. Specifically, we find a value of I that satisfies I * 360.0 / N <= AZIMUTH_ANGLE < (I+1) * 360.0 / N where N = XIF_COHORT_AZIMUTH_COUNT I = integer between 0 and N-1 inclusive The azimuth angles of all observations in the cohort lie in the interval [I*360 , (I+1)*360), and their average value is given in this column." END_OBJECT = COLUMN OBJECT = COLUMN NAME = INCIDENCE_ANGLE DATA_TYPE = MSB_UNSIGNED_INTEGER START_BYTE = 5 BYTES = 2 OFFSET = 0 SCALING_FACTOR = 0.00137342 UNIT = DEGREE VALID_MINIMUM = 0 VALID_MAXIMUM = 90 DESCRIPTION = "The average incidence angle of the incident radiation of all image framelets used in this row. (An image framelet is a single SAR observation.) The incidence angle is the angle between the local mean surface normal (based on a pre-Magellan topography model and provided with the BIDR data) and the direction of the incoming radiation (based on geometric factors and a pre-Magellan atmospheric model of refraction, also provided with the BIDR data). Normal incidence is thus 0 degrees, and grazing incidence is 90 degrees. As discussed in the GVHDR.LBL file, the group of SAR observations used in this row forms a 'cohort'. Within a cohort, the incidence angle of each observation falls within a single interval of size 90/N, where N is the value of XIF_COHORT_INCIDENCE_COUNT from the GVHDR file. We can reconstruct that interval by noting that the average incidence angle of all the observations (given in this column) falls within the same interval. Specifically, we find a value of I that satisfies I * 90.0 / N <= INCIDENCE_ANGLE < (I+1) * 90.0 / N where N = XIF_COHORT_INCIDENCE_COUNT I = integer between 0 and N-1 inclusive The incidence angles of all observations in the cohort lie in the interval [I*90 , (I+1)*90), and their average value is given in this column." END_OBJECT = COLUMN OBJECT = COLUMN NAME = POLARIZATION_ANGLE DATA_TYPE = MSB_UNSIGNED_INTEGER START_BYTE = 6 BYTES = 1 OFFSET = -90 SCALING_FACTOR = 0.72 UNIT = DEGREE VALID_MINIMUM = -90 VALID_MAXIMUM = 90 DESCRIPTION = "The average polarization angle of the incident radiation of all image framelets used in this row. The polarization angle is defined to be +/- 90 degrees for H-H polarization (+90 is nominal for mission cycle 1) and 0 degrees for V-V polarization. Nearly all orbits maintained an almost constant H-H or V-V polarization angle; the values of 90 or 0 reported for these orbits are only nominal and are not based on actual geometric calculations. Only values that differ from 90 or 0 have been actually computed." END_OBJECT = COLUMN OBJECT = COLUMN NAME = HISTOGRAM_LOWER_KNEE DATA_TYPE = MSB_UNSIGNED_INTEGER START_BYTE = 7 BYTES = 1 DESCRIPTION = "The average DN (still to do: we should change this to absolute dB, free of Muhleman normalization) of the lower knee of the histogram of all image framelets used in this row. For each framelet, the histogram of image pixel values was computed, and the abscissa to the right of 15.87% of the pixel values was recorded. For a normal distribution, this percentile is one standard deviation below the mean; note that the pixel distribution is rarely normal and is expressed on a logarithmic abscissa. The values found for each histogram were finally averaged to form the value here." END_OBJECT = COLUMN OBJECT = COLUMN NAME = HISTOGRAM_MEDIAN DATA_TYPE = MSB_UNSIGNED_INTEGER START_BYTE = 8 BYTES = 1 DESCRIPTION = "The average DN (we should change this to absolute dB, free of Muhleman normalization) of the median of the histogram of all image framelets used in this row. For each framelet, the histogram of image pixel values was computed, and the abscissa to the right of 50% of the pixel values was recorded. For any symmetric distribution, this value is equal to the mean; note that the pixel distribution is rarely normal and is expressed on a logarithmic abscissa. The values found for each histogram were finally averaged to form the value here." END_OBJECT = COLUMN OBJECT = COLUMN NAME = HISTOGRAM_UPPER_KNEE DATA_TYPE = MSB_UNSIGNED_INTEGER START_BYTE = 9 BYTES = 1 DESCRIPTION = "The average DN (we should change this to absolute dB, free of Muhleman normalization) of the upper knee of the histogram of all image framelets used in this row. For each framelet, the histogram of image pixel values was computed, and the abscissa to the right of 84.13% of the pixel values was recorded. For a normal distribution, this percentile is one standard deviation above the mean; note that the pixel distribution is rarely normal and is expressed on a logarithmic abscissa. The values found for each histogram were finally averaged to form the value here." END_OBJECT = COLUMN OBJECT = COLUMN NAME = HISTOGRAM_MODE DATA_TYPE = MSB_UNSIGNED_INTEGER START_BYTE = 10 BYTES = 1 DESCRIPTION = "The average DN (we should change this to absolute dB, free of Muhleman normalization) of the mode of the histogram of all image framelets used in this row. For each framelet, the histogram of image pixel values was computed, and the abscissa with the largest number of pixel values was recorded. The values for all framelets were averaged to form the value here." END_OBJECT = COLUMN OBJECT = COLUMN NAME = SCATTERING_LAW_CONSTANT_TERM DATA_TYPE = MSB_UNSIGNED_INTEGER START_BYTE = 11 BYTES = 1 OFFSET = -35 SCALING_FACTOR = 0.2 UNIT = DECIBEL VALID_MINIMUM = -35 VALID_MAXIMUM = 15 DESCRIPTION = "The average radar backscatter cross-section of all image framelets used in this row. For each framelet, a quadratic polynomial has been fitted to the backscatter cross-section as a function of incidence angle. The incidence angle over any framelet only varies by 1-2 degrees, so this fit is a local one only and may be dominated by geologic features within the framelets." END_OBJECT = COLUMN OBJECT = COLUMN NAME = SCATTERING_LAW_LINEAR_TERM DATA_TYPE = MSB_UNSIGNED_INTEGER START_BYTE = 12 BYTES = 1 OFFSET = -5 SCALING_FACTOR = 0.04 UNIT = DECIBEL_PER_DEGREE VALID_MINIMUM = -5 VALID_MAXIMUM = 5 DESCRIPTION = "The average radar backscatter cross-section derivative of all image framelets used in this row. For each framelet, a quadratic polynomial has been fitted to the backscatter cross-section as a function of incidence angle. The incidence angle over any framelet only varies by 1-2 degrees, so this fit is a local one only and may be dominated by geologic features within the framelets." END_OBJECT = COLUMN OBJECT = COLUMN NAME = SCATTERING_LAW_QUADRATIC_TERM DATA_TYPE = MSB_UNSIGNED_INTEGER START_BYTE = 13 BYTES = 1 OFFSET = -15 SCALING_FACTOR = 0.12 UNIT = DECIBEL_PER_DEGREE_SQUARED VALID_MINIMUM = -15 VALID_MAXIMUM = 15 DESCRIPTION = "The average radar backscatter cross-section second derivative of all image framelets used in this row. For each framelet, a quadratic polynomial has been fitted to the backscatter cross-section as a function of incidence angle. The incidence angle over any framelet only varies by 1-2 degrees, so this fit is a local one only and may be dominated by geologic features within the framelets." END_OBJECT = COLUMN