KILOMETER-SCALE ROUGHNESS MAPS OF MERCURY

Download maps of topographic roughness of the northern circumpolar (65ºN - 84ºN) region of Mercury described in the following paper:

Kreslavsky, M., J.W. Head, G.A. Neumann, D.E. Smith, M.T. Zuber (2014), Kilometer-Scale Topographic Roughness of Mercury: Correlation with Geologic Features and Units, Geophys. Res. Lett., in press.
Please, refer this paper if you use these data.

MLA_Roughness_composite.png is a color composite roughness map that combines roughness at 0.7 km, 2.8 km, and 11 km baseline coded as Blue, Green, and Red color channels, respectively; higher intensity in each channel denotes a higher roughness. Each channel is non-linearly stretched to optimize visual perception of the map. The map is suitable for qualitative analysis only. Actually, this is Fig. 3 from the paper; see the paper for further details. The map is in Lambert azimuthal equal-area projection centered at the north pole with sampling of 16 pixels per degree.

MLA_Roughness_composite.tif is a GeoTIFF version of the same map suitable for smooth import into GIS software.

MLA_RoughnessMaps.zip is a ZIP archive that contains the following raw roughness map data files:

N16r01.bsq     roughness at 0.7 km baseline
N16r04.bsq     roughness at 2.8 km baseline
N8r16.bsq       roughness at 11 km baseline

If you use ArcGIS, import of these raster files should be straightforward. In addition to these data files, this archive contains all auxiliary ArcGIS files needed for smooth import. Just unzip them into the same folder/directory in your computer.

If you do not use ArcGIS, you also need to know the following:

All *.bsq files are bare pixel data arrays, no headers, no any other information, pixel values only. Each pixel is two-byte (16 bit) unsigned integer, least significant byte first ("Intel" or "PC" byte order).

The maps are in Lambert azimuthal equal-area projection centered at the north pole.

Pixel array dimensions and sampling are the following:

N16r01.bsq     1440 x 1440    16 pixels per degree
N16r04.bsq     1440 x 1440    16 pixels per degree
N8r16.bsq       720 x 720        8 pixels per degree

You are advised to use the roughness data in relative sense, "rougher - smoother". Higher pixel values mean "rougher". However, if you do need absolute roughness values, below is the calibration information.

The mapped measure of roughness is the interquartile range of the 2nd derivative of elevation along profiles. See the paper for further details. To obtain this quantity you need to multiply the pixel number value by the following scaling coefficients:

10.0 ´ 10-6 m-1            for 0.7 km baseline      (file N16r01.bsq)
2.50 ´ 10-6 m-1            for 2.8 km baseline      (files N16r04.bsq)
0.625 ´ 10-6 m-1          for 11 km baseline       (files N8r16.bsq)

If you have any questions about these data, please contact Mikhail Kreslavsky: mkreslav@ucsc.edu