Mars is the Earth's outer neighbor; it is the fourth planet from the Sun. The first historical study of Mars was credited to Francisco Fontana, who rendered telescopic observations in 1638. Around 1877, Giovanni Virginio Schiaparelli described the “canali” or channel formations in the planet. But it was not until 1965 that the first close-up picture of Mars revealed a world strangely familiar, yet different enough to challenge our perceptions of what makes a planet work. Although spacecraft have shown us that Mars is rocky, cold, and sterile beneath its hazy, pink sky, scientists presume there was an early warm period and that the heat from inside Mars probably caused the volcano’s formation as well as the continental drift.
In order to better understand the geological features, water resources, climate and biological aspects of the red planet several spacecrafts including orbiters, landers, and rovers have been sent by the Soviet Union, the United States, Europe, and Japan.
The Mars 2001 Odyssey is an orbiter that was launched from Kennedy Space Center on 7 April 2001, it carries three main science instruments: The Gamma Ray Spectrometer (GRS), the Mars Radiation Environment Experiment (MARIE) and the Thermal Emission Imaging System (THEMIS). This last one combines a 5-band visual (VIS) spectral resolution imaging system with a 10-band spectral resolution infrared (IR) imaging system. The IR subsystem has a 100m/pixel spatial resolution while the VIS subsystem has a 19m/pixel resolution. THEMIS began observing the surface and atmosphere of Mars in February 2002 with the use of thermal infrared multispectral imaging between 6.5 and 15 m and visible to near-IR images from 450 to 850 nm. Some researchers have exploited the unique aspect of investigating images of Martian temperatures at a resolution of 100 m. Not surprisingly, the temperatures are related to material properties, such as composition and structural integrity, climate and topography, which indicate the origin, evolution and history of the planet.
A major result of the Mars Global Surveyor (MGS) imaging was the discovery of rock units that are layered at a meter to ten meter scales. For THEMIS IR a major result from the camera is the discovery that the physical and compositional properties of these layers can vary, implying temporal changes in the processes or environments that formed the different units. A key element of this discovery is the exposure at the surface of materials that directly reflect the properties of the underlying rock unit from which they were derived.
Statement of the Problem
In order to better understand Mars, the sites of Olympus Mons, and Valles Marineris were studied by utilizing THEMIS remote sensing imaging:
It is believed that since Olympus Mons is an extinct volcano that measures 24 km (15 miles) high, and the top of it is 70 km (40 miles) wide; the base is about 600 km (375 miles) wide. Therefore, it was originally hypothesized that basalt, pyroxene, granite and possibly gneiss, schist, slalte, marble might be present at this site. These goals were modified from the initial proposal and are discussed below.
Valles Marineris is 4000 km (2500 mi) long and reaches depths of up to 7 km (4 mi) and widths of up to 200 km (125 mi). Most researchers agree that Valles Marineris is a large tectonic "crack" in the Martian crust, forming as the planet cooled, affected by the rising crust in the Tharsis region to the west, and subsequently widened by erosional forces. Therefore, the initial goal was to test for the presence of water and minerals such as basalt, pyroxene and granite. The goals were also modified and are discussed below.
Objectives & Methodology
For the purposes of this project, the data used was derived from THEMIS to analyze the surface temperatures of two Martian sites as well as mineral compositions. The spectral bands scrutinized for this study involved the visible, ultraviolet (UV), and infrared (IR) ranges of the spectrum. All of the data utilized ISIS to process the images in order to view all 10 spectral bands. ENVI was used to analyze the range of spectral bands, once processing was complete. By using ENVI’s spectral library, it was believed that Martian minerals could be identified from inherent spectral signatures and correlated to minerals found on Earth.
As stated in the initial proposal, the project’s goals were to assess the mineralogy, temperature, and to investigate the presence of water in a region of Valles Marineris (Figure-I) and to assess the same parameters in a region of Olympus Mons (Figure-II). By using images from THEMIS, it was believed that the goals outlined above were quite attainable. Unfortunately, Martian the data utilized from THEMIS was problematic and consequently only thermal assessment and image classification was undertaken.
Figure-I Valles Marineris
Figure II- Olympus Mons
All THEMIS derived data exist as visible apparent brightness records (VISABR) and infrared brightness temperature records (IRBTR). These two datasets are collected as QUBE objects, which comply with the Planetary Data System (PDS) standards. A QUBE is an array of sample values in two dimensions. The “core” of a THEMIS QUBE is three-dimensional, with two spatial dimensions (samples and lines) and one spectral dimension (bands). The QUBE format allows THEMIS data to simultaneously a set of images (at different wavelengths) of the same target area, and also a multi-point spectrum at each spatially registered pixel target area (Figure-III). Additional information may be stored in “suffix” planes (back, side, or bottom) as conceptually depicted in Figure-IV.
Figure-III THEMIS QUBE core structure
Figure-IV Exploded view of PDS QUBE
The QUBE object’s complexity proved to be a significant obstacle for evaluating any data derived from THEMIS. An ENVI software package was used for analyzing the Martian data. Regrettably, ENVI cannot directly read the complex QUBE files. Consequently, a separate software called ISIS 3.0 Beta, which is provided free pf charge through the USGS Astrogeology Research Program at was needed to process the images. Moreover, the obstacles regarding Martian data were compounded by the fact that ISIS is supported only by a Linux operating system. Though ISIS has many image processing applications available, only two were needed to complete the project. The first is a THEMIS to ISIS (thm2isis) application which converts a THEMIS QUBE file to a .cub file. Once converted to a .cub file, the data was further processed by using the ISIS to TIFF (isis2tif) application, which converts a .cub file to a .tiff file. However, one should note that each band for a given data file must be converted to a separate .tiff file. For example, if a 10-band RDR.QUB file has been chosen, the end result will be 10 separate .tiff files for that single image. The bands must then be stacked together as an ENVI standard image for analysis. Once converted to .tiff files, the header files required editing. The entire process proved to be tedious and time consuming. The steps involved in processing an IR RDR. QUB file are outlined below in Figure-V.
Figure-V (IR RDR.QUB data processing)
As demonstrated in Figure-V, simply reaching a point where the data is in a useable form is a complex procedure in itself. The data for the visible images utilized a similar method except for two differences. To convert from Kelvin to Celsius the following formula was used: (B1*1.8-459.637). The other formula used for the band math portion was: (B1*SF) + offset, where SF is the scaling factor.
Once the cumbersome image processing was completed, analysis of the images using ENVI was performed on the two sites. The initial site evaluated was Olympus Mons (Figure-IV). The volcano is nearly 27 km high; it is over 20 times wider than it is tall. Thus, most of the volcano has a fairly gentle surface slope. The image also shows the distinct cliff which marks the base of Olympus Mons. In places, this scarp is up to 6 km high. In other places, it is hidden under lava flows cascading out into the surrounding lava plains. This cliff is unique among the giant shield volcanoes on Mars. The rough, crinkly patches around Olympus Mons are also unusual and form the Olympus Mons Aureole.
Though is has been stated that the area of Olympus Mons was studied, only a very small area near the base of the volcano as depicted by the arrow in Figure-VI was the true study area. The actual image or swath (Figure-VIII) was comparatively much smaller yet abundant in data and information. Figure-VIII illustrates the difficulty associated with obtaining both visible and infrared spectral bands. The red swaths are the available IR sites whereas the purple boxes are the available visible sites. Therefore, with limited choices, the area of study was chosen solely on obtainable visible and infrared information and not as a matter of preference.
Figure-VI Enhanced area of study
Figure-VII Swath of area studied
Figure-VIII Available Vis and IR sites
Figure-IX below shows a screen capture of analysis performed in ENVI on a portion of the image swath collected from Olympus Mons. The first image is a density slice measuring a temperature range of -27F to 4F. The middle image is band 9 of the IR image and the last image on the right is the visible image. As a point of reference the bolder depicted by the arrows was used to ensure that the same area could be viewed in three different frames.
The temperature data gathered revealed that there was approximately a thirty-one degree difference among the various areas calculated in the density slice. Much of the density slice area shown is dominated by the colors magenta and cyan which had calculated temperature ranges of -7.5522 F to -3.6582 F and -11.4461 F to -7.5522 F respectively. Yellow was the next most abundant color. It ranged from -15.3401 F to -11.4461 F. It is difficult to verify the temperature consistency at this location. Though, since Olympus Mons possesses different heights, it was pressumed that the temperature variance is most likely attributed to elevation.
Further analysis of the temperature in this region was conducted by examining all 10 IR and 5 Vis bands individually in ISIS after conversion to tiff files. The tiff files were then individually edited, stacked together, and scrutinized by performing an IR z-profile analysis. The two plots generated by the z-profile analysis can be viewed in figures Xa and Xb. Figure-Xc, represents both spectral plots for the Volcano Site. Unfortunately, the main goal of the research project could not be derived in part due to the inability of obtaining the true kinetic temperature a Mars. Accordingly, emmissivity could not be calculated in order to make a comparison of Earth’s mineral library to that of Mars. Yet, if emmissivity could have been obtained, several obstacles remained such as the need to consider atmospheric correction and assessment of interference due to Martian dust.
Figure-Xa IR z-profile plot
Figure-Xb Vis z-profile plot
However, a region of interest was selected in which an image analysis was performed to compare similar physical or chemical characteristics (Figure- XI). In this case the region selected was one pixel within the red box illustrated in (Figure- XIb) and further clarified by the black arrow. As revealed in (Figure XIc) there was a significant similarity throughout the entire image. This effect is attributed to the fact that this particular image is a slope from a volcanic site. Therefore it is presumed that any ancient erupted material would have been evenly distributed along side the volcano or at base of the volcano. The image below (Figure XIc) appears to support this presumption.
Figure XI Radiance
Figure XIa Density Slice
Figure XIb IR image
Figure XIc Image classification
Canyon Region (Valles Marineris)
The Valles Marineris is a system of canyons located just south of the Martian equator. The system is about 4000 km long. Comparatively, the same canyon would extend all the way across the United States. The central individual troughs, generally 50 to 100 km wide, merge into a depression as much as 600 km wide. In places the canyon floor reaches a depth of 10 km, 6 to 7 times deeper than the Grand Canyon. As with the Olympus Mons site, only a small area of Valles Marineris was studied. Figure-XII is an enhanced image of the region while figure-XIII shows a swath of the actual study area. The same methodologies which were applied the Olympus Mons site were applied to the Valles Marineris site and are discussed below.
Figure-XII Enhanced area of study
Figure-XIII Swath of area studied
The temperature data gathered revealed approximately a thirty-seven degree difference among the various areas calculated in the density slice (Figure XI). Much of the area shown was dominated by the colors magenta and cyan which had calculated temperature ranges of -5.1908 F to -0.5448 F and -9.8368 F to -5.1908 F respectively. Yellow and maroon, were the next two most abundant colors. Yellow ranged from -14.4828 F to -9.8368 F, while maroon ranged from -0.5448 F to 4.1012 F. However, as with the Olympus Mons site it was difficult to verify the temperature consistency at this location aside for the approximation displayed by the data.
Analysis was performed on the Valles Marineris site consistent with the methodology applied to the Olympus Mons site. The radiance was calculated for the 10 infrared bands as well as for the visible and a final plot was obtained to depict their differences. These results are shown in Figure X. Figure Xa shows the z-profile analysis for the IR image while Figure Xb shows the z-profile analysis for the visible image. Figure Xc illustrates the combined analysis as one graph.
Figure-Xa IR radiance
Figure-Xb Vis radiance
Figure-Xc Combined IR and Vis
Contrary to the volcano site, which showed continuity throughout with regard to the image classification, the Valles Marineris site showed significantly different results. Only the immediate area surrounding the impact crater exhibited physical or chemical characteristics which were similar to the area selected. One possible theory for explaining this phenomenon would be to suggest that similar material surrounding the crater was scattered in the general vicinity of the crater upon impact from a terrestrial body. The similar material, of course, would be that of the foreign body which impacted the site. Therefore, only the general area around the crater would be expected to consist of similar mineralogy. The rest of the study area would be expectedly different, which is what was observed.
Figure XIa Density slice
Figure XIb IR image
Figure XIc Image Classification
Conclusion Due to limited data availability and difficult image processing, only analysis of radiance in relation to the individual wavelengths in each spectral category (IR and Vis) could be pursued. Though the results of this study yielded interesting results, the lack of ground-truthing temperature data prevented the calculation of emmissivity. As stated earlier, the kinetic temperature of Mars could not be derived. In order to make mineralogy comparisons between Earth and Mars, emmissivity which is calculated from Mars kinetic temperature, must be obtained. However, this does not imply that had emmissivity had been calculated that mineralogy could be confirmed. The presence of minerals such as basalt, pyroxene and granite is inherently difficult to evaluate due the influence of other factors. Because further work is needed to remove atmospheric interactions and other environmental interactions that could affect the results yielded, the information obtained is at best an approximation based on the technology available.
Despite the problems associated with deriving mineralogy results, this study did produce interesting findings. From the results obtained it was noted that the physical/chemical composition surrounding the volcano site was consistent throughout the image, while the crater area analyzed in the Valles Marineris site displayed similarity only with the impact area. In addition, it was confirmed by the image classification, that regardless of the temperature of an object, the radiance for different substances with different physical and chemical properties reveals subtle differences when compared as a whole. The significance of this research is that the calculation of radiance is the most appropriate parameter for remote sensing of terrestrial bodies.