General Information
Day IR
Night IR
Visible
Thermal Inertia
Night IR over Day IR
Thermal Inertia over Day IR
Decorrelation Stretched Day IR (DCS)
References
Acknowledgements
Data Citation
General Information
Individual THEMIS images, both infrared (IR) and visible, are geometrically projected using ISIS software, and then mosaiced together. These mosaics are centered on the latitude and longitude value provided by the proposer and are 3° latitude x 3° longitude (~180 km square). To remove appearance of differences due to seasonal, time of day, and atmospheric variation during acquisition, these images are normalized to one another. Thus, these mosaics do not provide quantitative information, but are helpful for understanding relative differences between geologic materials and morphologic surface features, and for mapping surface units.For both daytime and nighttime IR mosaics, band 9 radiance values (centered at 12.57 µm) are used because this wavelength range has the highest signal to noise value and is relatively transparent to atmospheric dust. The brightness temperature of the surface is determined by fitting a Planck curve to band 9 calibrated radiance that has been corrected for instrumental effects, and best approximates the surface kinetic temperature [Christensen et al., 2004].
Day IR
Daytime surface temperatures in a single THEMIS image are primarily affected by albedo and local topography [e.g. Kieffer et al., 1973; Palluconi and Kieffer, 1981; Christensen, 1982]. Because of the dependence of topography on daytime temperature, day IR mosaics are useful for understanding the morphology of a region, and provide context for both THEMIS and MOC NA visible images. Brighter regions have a relatively high temperature and are typically caused by either Sun-facing or low albedo surfaces. Darker regions have relatively low temperatures and are generally surfaces with a high albedo or are in shadow.
Night IR
At night the effects of topography and albedo have largely dissipated, and differences in surface temperature are primarily due to its thermal inertia, which is a function of the physical nature of the surface [e.g. Kieffer et al., 1973; Palluconi and Kieffer, 1981; Christensen, 1982]. Because of this relationship, relative differences in nighttime temperature can be used to understand qualitative differences in thermal inertia. Bright regions have a higher temperature, corresponding to a relatively high thermal inertia (such as exposed bedrock, rockier, more indurated, or coarser particle sizes on the surface). Dark regions have a lower temperature and have a relatively low thermal inertia (such as dustier, finer particle sized material, or unconsolidated grains). A delta-temperature (K) is provided and characterizes the difference between the 2 sigma minimum and maximum nighttime temperature values, and provides a sense of the range in temperature within the mosaic.
Visible
For visible mosaics, band 3 images (centered at 654 nm) are mosaiced together. Current THEMIS coverage of most landing sites is limited, but complete coverage of these regions is a high priority, and gaps in visible image coverage are currently being targeted by the THEMIS mission planners. As increased coverage becomes available, these mosaics will be improved and updated.
Thermal Inertia
THEMIS infrared data have an improved spatial resolution (100 m per pixel) over previous datasets, such as MGS-TES or Viking IRTM. This dataset enables the quantification of surface physical characteristics to determine particle size information and identify surface exposures of bedrock, and allows these physical properties to be correlated to morphologic features. This dataset can also facilitate an improved understanding of geologic processes that have influenced the Martian surface. During the mosaicing processes, thermal inertia images have been normalized to one another to remove any image to image variations caused by atmosphere, seasonal, or time of day variations that are not adequately accounted for by the thermal model. The scale for thermal inertia mosaics, however, is determined from individual images before normalization, and is representative of the mosaiced region.
The thermal model used to calculate THEMIS thermal inertia values is derived from the Viking IRTM thermal model [Appendix 1 of Kieffer et al., 1977], with the primary modification being the replacement of a constant atmospheric thermal radiation with a one-layer atmosphere that is spectrally gray at solar wavelengths, and the direct and diffuse illuminations are computed using a 2-stream delta-Eddington model. The effects of 3-dimensional blocks on the surface, condensate clouds, and the latent heat of water ice are not considered (H. H. Kieffer, personal communication, 2006). This model can incorporate the effects of a radiatively-coupled sloping surface at any azimuth, but for the nominal thermal inertia calculations, slopes are not considered. Generally, slopes below 10° at all azimuths have a small effect on the nighttime surface temperature, and therefore the thermal inertia. Higher slope angles may be problematic, but this conclusion is dependent on the slope azimuth and the season. Due to the potential for slopes to be a factor, surfaces with slopes greater than ~10° should be interpreted with caution [Fergason et al., in press]. Work is currently being done to incorporate slopes into these mosaics, and the thermal inertia mosaics will be improved and updated as these data become available.
The THEMIS band 9 nighttime temperatures are converted to a thermal inertia by interpolation within a 7-dimensional look-up table. This table was created by first selecting a range of values appropriate for the Martian surface (nodes) for seven input parameters: latitude, season, local solar time, atmospheric dust opacity, thermal inertia, elevation (atmospheric pressure), and albedo. These nodes were then used to calculate the model-derived surface kinetic temperature for the specified conditions using the thermal model described above. Model parameters appropriate for an individual THEMIS image and the measured band 9 nighttime brightness temperatures are then used to interpolate the thermal inertia between these calculated node values [Fergason et al., 2006].
The look-up table includes a thermal inertia range of 24 to 3000, and values exceeding 1800 have been observed thus far in the mission [e.g. Edwards et al., 2005]. This thermal inertia range is significantly larger than that used in the TES standard model (maximum of 800), and allows the detection of exposures of consolidated materials or bedrock on the surface. This extended thermal inertia range was required because of: (1) the higher resolution of THEMIS; (2) initial results from THEMIS nighttime temperatures suggesting the presence of bedrock [e.g. Christensen et al., 2003]; and (3) the fact that many regions on Mars were saturated at the maximum value of thermal inertia in the TES model [Fergason et al., 2006].
Night IR over Day IR
The night IR mosaic is colorized and overlaid onto the day IR mosaic (see above for details on the individual datasets) to allow one to more easily relate variations in the physical properties of the surface to morphological features. Blue represents lower temperatures and red corresponds to higher temperatures. A delta-temperature (K) is provided and characterizes the difference between the 2 sigma minimum and maximum nighttime temperature values, and provides a sense of the range in temperature within the mosaic.
Thermal Inertia over Day IR
The thermal inertia mosaic is colorized and overlaid onto the day IR mosaic (see above for details on the individual datasets) to allow one to more easily relate variations in thermal inertia to morphological features. Blue corresponds to lower thermal inertia values and red represents higher thermal inertia values.
Decorrelation Stretched Day IR (DCS)
A decorrelation stretch (DCS) enhances color differences between channels in an image by removing highly correlated information. For THEMIS data, three IR channels are displayed in red, green, and blue (RGB). Because the color scheme is additive, mixtures of these channels produce additional colors such as yellow (red & green), cyan (green & blue), and magenta (red & blue).
The THEMIS DCS images produced in support of MSL landing site selection display calibrated radiance bands 8 (11.79 µm), 7 (11.04 µm), and 5 (9.35 µm) in red, green, and blue, respectively [Hamilton et al., 2007]. Colors are blended across the mosaic laterally in an attempt to create a uniformly colored image (with varying degrees of success). Radiance images have not had the effects of temperature removed, so color intensity/brightness variations (e.g., light blue vs. dark blue) may indicate differences in temperature (resulting from albedo and/or particle size variations) rather than composition. Differences in hue (e.g, green vs. magenta) are more likely related to compositional variation.
THEMIS DCS images are a visualization tool intended primarily for identifying compositional variations within a geographic region. Because these images contain highly manipulated data, no quantitative information can be extracted from them. However, they can be used for rudimentary compositional distinctions, to identify locations where multi- or hyperspectral compositional data should be retrieved for quantitative analysis, and they can be used to map out the coincidence of compositional variations with variations in other data sets, such as thermal inertia or geomorphology. The 8-7-5 DCS tends to show the most mafic (usually olivine-bearing) materials as purple/magenta in color, whereas relatively more felsic materials tend to be yellowish. (Sulfate band positions commonly coincide with those of more felsic silicates and may be expected to appear as yellowish in an 8-7-5 DCS image.) In general, compositional variations tend to correlate with geomorphology.
Artifacts and atmospheric effects are present in many of the DCS mosaics; several typical examples are described here, but if you have any question at all as to what you are looking at and whether it is on the surface, in the atmosphere, or an artifact, please ask for assistance (see Hamilton et al. [2007] for contact information). We have endeavored to avoid using low-quality data or data with significant atmospheric features, but these may be used if no other data are available; in such cases, the mosaics will be updated as higher-quality data become available. Artifacts commonly result from corrupted or missing data within an image or attempts to mosaic adjacent images acquired at times of day that are several hours apart when surface temperatures are significantly different (and shadowing varies visibly). Many artifacts are linear strips of discoloration, "seams", and color variations oriented along-track, ~NNE-SSW (e.g., the center of the Becquerel crater mosaic). Data gores within an image commonly result in artifacts that are oriented across-track and may be identified by an oddly-colored, noticeably artificial line (e.g., the Nili Fossae crater mosaic). A rarer class of artifacts occurs in cases where extreme variations in surface temperature are not corrected completely by the standard THEMIS calibration. These artifacts appear as very intense blotches of color that do not correlate with geomorphology (e.g., in the left half of the Juventae Chasma mosaic). The atmospheric effects that can be mistaken for compositional variations are water ice clouds and changes in dust opacity associated with topography. In the 8-7-5 DCS, clouds are blue, irregular in shape, do not correlate with geomorphology, and generally do not cross image boundaries (i.e., from left to right in the mosaic, e.g., the center of the Bequerel crater mosaic). Clouds commonly are present near the rims of the canyons of the Valles Marineris, but may occur anywhere. Large topographic variations within a THEMIS image (most typically around the Valles Marineris) have correlated atmospheric dust opacity variations, which are sufficient to produce differences in the coloration of a DCS image. Differences in color between plateau materials and materials within the chasmata in particular should be assessed with skepticism; images crossing large topographic boundaries should be atmospherically corrected for the most robust evaluation.
Individual decorrelation stretched THEMIS images are available through the THEMIS PDS website and offer additional band combinations, and in some cases, data acquired at different times (which may be useful in identifying ephemeral atmospheric features).
Acknowledgements
Ryan Luk (Arizona State University) made the Day IR, Night IR, and Visible mosaics and the Night IR over Day IR overlay images. Robin Fergason (Arizona State University) made the Thermal Inertia mosaics and the Thermal Inertia over Day IR overlay images. Mikki Osterloo and Brian McGrane (University of Hawai'i) made the DCS images.
References
Christensen, P. R. (1982), Martian dust mantling and surface composition: Interpretation of thermophysical properties, J. Geophys. Res., 87(B12), 9985-9998.
Christensen, P. R., J. L. Bandfield, J. F. Bell III, N. Gorelick, V. E. Hamilton, A. Ivanov, B. M. Jakosky, H. H. Kieffer, M. D. Lane, M. C. Malin, T. McConnochie, A. S. McEwen, H. Y. McSween Jr., G. L. Mehall, J. E. Moersch, K. H. Nealson, J. W. Rice Jr., M. I. Richardson, S. W. Ruff, M. D. Smith, T. N. Titus, and M. B. Wyatt (2003), Morphology and composition of the surface of Mars: Mars Odyssey THEMIS results, Science, 300(5628), 2056-2061.
Christensen, P.R., B.M. Jakosky, H.H. Kieffer, M.C. Malin, H.Y. McSween, Jr., K. Nealson, G.L. Mehall, S.H. Silverman, S. Ferry, M. Caplinger, and M. Ravine, The Thermal Emission Imaging System (THEMIS) for the Mars 2001 Odyssey Mission (2004), Space Science Reviews, 110, 85-130.
Edwards, C. S., J. L. Bandfield, P. R. Christensen, and R. L. Fergason (2005), Global distribution of bedrock on Mars using THEMIS high resolution thermal inertia, Eos Trans. AGU, 86(52), Fall Meet. Suppl., Abstract P21C-0158.
Fergason, R.L., P.R. Christensen, and H.H. Kieffer (2006), High resolution thermal inertia derived from THEMIS: Thermal model and applications, J. Geophys. Res., 111, E12004, doi:10.1029/2006JE002735, 2006.
Hamilton, V. E., M. M. Osterloo, and B. S. McGrane, THEMIS decorrelation stretched infrared mosaics for compositional evaluation of candidate 2009 Mars Science Laboratory landing sites, Lunar Planet. Sci., XXXVIII, 2007.
Kieffer, H. H., S. C. Chase, Jr., E. Miner, G. Münch, and G. Neugebauer (1973), Preliminary report on infrared radiometric measurements from the Mariner 9 spacecraft, J. Geophys. Res., 78(20), 4291-4312.
Kieffer, H. H., T. Z. Martin, A. R. Peterfreund, B. M. Jakosky, E. D. Miner, and F. D. Palluconi (1977), Thermal and albedo mapping of Mars during the Viking primary mission, J. Geophys. Res., 82(28), 4249-4291.
Palluconi, F. D., and H. H. Kieffer (1981), Thermal inertia mapping of Mars from 60° S to 60° N, Icarus, 45, 415-426.
Data Citation
Acknowledgement of the THEMIS site and its resources will help ensure the continuing support necessary for the validation, archiving, and distribution of THEMIS images.
For news media or educational purposes, please credit THEMIS images as: NASA/JPL/Arizona State University.
If you use THEMIS images or mosaics in a research work, please cite them by image ID or mosaic name in the caption and put the following in your references:
Christensen, P.R., N.S. Gorelick, G.L. Mehall, and K.C. Murray, THEMIS Public Data Releases, Planetary Data System node, Arizona State University, <http://themis-data.asu.edu>.
If you mention the THEMIS instrument in a research work, please cite it as follows in your references:
Christensen, P.R., B.M. Jakosky, H.H. Kieffer, M.C. Malin, H.Y. McSween, Jr., K. Nealson, G.L. Mehall, S.H. Silverman, S. Ferry, M. Caplinger, and M. Ravine, The Thermal Emission Imaging System (THEMIS) for the Mars 2001 Odyssey Mission, Space Science Reviews, 110, 85-130, 2004.
If you mention the THEMIS-derived thermal inertia in a research work, please cite it as follows in your references:
Fergason, R.L., P.R. Christensen, and H.H. Kieffer, High Resolution Thermal Inertia Derived from THEMIS: Thermal Model and Applications, J. Geophys. Res., 111, E12004, doi:10.1029/2006JE002735, 2006.
If you mention the THEMIS-derived decorrelation stretch (DCS) images in a research work, please cite it as follows in your references:
Hamilton, V. E., M. M. Osterloo, and B. S. McGrane, THEMIS decorrelation stretched infrared mosaics for compositional evaluation of candidate 2009 Mars Science Laboratory landing sites, Lunar Planet. Sci., XXXVIII, 2007.