This product is a global thermal inertia mosaic of Mars generated from high-resolution Thermal Emission Imaging System (THEMIS) nighttime infrared images (Christensen et al., 2004; Fergason et al., 2006a) that allows the identification, assessment, and global correlation of martian surface materials. The higher spatial resolution of THEMIS data (100 m per pixel), relative to Thermal Emission Spectrometer (TES) globally binned data (~3 km per pixel (Mellon et al., 2000; Putzig et al., 2005; Putzig and Mellon, 2007)), enables one to more easily quantify the physical properties of morphologic features observed in high-resolution images, such as those from the High Resolution Imaging Science Experiment (HiRISE), Context Camera (CTX), and Mars Orbiter Camera (MOC), and to better understand geologic processes acting on local scales. Thermophysical variations often correspond to features identified in high-resolution images, and the integration of these data sets allows more robust scientific conclusions to be reached.
Thermal inertia is defined as I=(kρc)1/2 density of the surface material, and c is the specific heat. It represents the resistance to change in temperature of the upper few centimeters of the surface throughout the day, and is essentially independent of local time, latitude, and season. Darker areas in the map have a lower apparent thermal inertia and likely represent fine particles, such as dust, silt, or fine sand. Brighter regions have a higher apparent thermal inertia surfaces and may consist of coarser sand, surface crusts, rock fragments, bedrock, or a combination of these materials. For example, exposed bedrock typically has an apparent thermal inertia greater than ~1800 Jm-2K-1s-1/2 (Fergason et al., 2006b; Edwards et al., 2009), whereas rocky regolith has an apparent thermal inertia greater than ~400 Jm-2K-1s-1/2 that of exposed bedrock. The technique of Fergason et al. (2006a) was used to generate these mosaics. See Fergason et al. (2006a), Christensen and Fergason (2013), and the metadata associated with each product for additional information on how these mosaics were generated and the related uncertainties.
These mosaics are released in collaboration with the Mars Space Flight Facility at Arizona State University. These products can also be accessed at: http://jmars.mars.asu.edu/maps/ thermal_inertia.
In addition, these mosaics, and the individual THEMIS-derived thermal inertia images used to generate these mosaics, are available in JMars. You can download the latest version of JMars at http://jmars.asu.edu/.
Please contact Robin Fergason (firstname.lastname@example.org) with questions regarding any of these released data products.
We chose to generate two separate mosaic products because each deliverable has its own functionality and is aimed towards a different type of user. Each tile is 30° latitude by 60° longitude, and titles are designated by the coordinates of the lower right corner.
Qualitative (8-bit) Products: The 8-bit products are normalized to remove any image- to-image variations resulting in a seamless final product comparable to the 100 m/pixel nighttime IR THEMIS dataset currently available (see http://jmars.mars.asu.edu/maps/? layer=thm_nightir_100m_v8). In addition, the individual images and tiles have been normalized in such a way that multiple tiles can be downloaded and mosaicked together without noticeable seams between tiles. This product is particularly useful in geologic mapping, where a seamless product is needed to differentiate unit boundaries, textures, and morphologic features. This product will also be appropriate for generating publication-quality figures to be included in scientific publications and presentations at professional meetings.
Quantitative (32-bit) Products: We have also generated a 32-bit (quantitative) global thermal inertia mosaic to better meet the needs of the planetary science and engineering communities. Apparent thermal inertia values in quantitative (32-bit) mosaics are unmodified during the mosaicking process and are therefore suitable for scientific investigation. Thermal inertia is independent of season and local time, and thus image-to-image differences are typically less apparent in thermal inertia than temperature data. However, image-to-image differences can still be prominent in derived thermal inertia. Understanding the cause of these differences is complex due to many factors affecting the measured surface temperatures and resulting derived thermal inertia values. Examples of some potential factors include: 1) uncertainties in the calibration of the THEMIS instrument, particularly relating to knowledge of the image start time resulting in a random error standard deviation between images of ~4 K at 180 K (Christensen et al., 2004), which corresponds to errors of 50-120 Jm-2K-1s-1/2) transient atmospheric phenomena, such as water-ice clouds, that are not taken into account; 3) surface slopes not included in this derivation of apparent thermal inertia; and 4) the presence of ice or bedrock in the shallow subsurface (Putzig and Mellon, 2007; Bandfield, 2007; Bandfield and Feldman, 2008), which are not considered in the 1-D model necessary to generate global data in an automated manner. Considering all these uncertainties, the absolute accuracy of the THEMIS thermal inertia is ~20% (Fergason et al., 2006a). However, relative differences within a single image are quite accurate (~10%; Fergason et al., 2006a), and therefore the apparent thermal inertia of adjacent surfaces is effective in identifying differences between local materials. We have minimized these image-to-image variations by ordering images so that the warmest season available in on the top of the mosaic. These warmer images also typically include the highest quality data, as they have the highest signal to noise.