SEARCH

SEARCH BY CITATION

Keywords:

  • M3;
  • lunar;
  • maturity;
  • space weathering;
  • spectra

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[1] High spectral and spatial resolution data from the Moon Mineralogy Mapper (M3) instrument on Chandrayaan-1 are used to investigate in detail changes in the optical properties of lunar materials accompanying space weathering. Three spectral parameters were developed and used to quantify spectral effects commonly thought to be associated with increasing optical maturity: an increase in spectral slope (“reddening”), a decrease in albedo (“darkening”), and loss of spectral contrast (decrease in absorption band depth). Small regions of study were defined that sample the ejecta deposits of small fresh craters that contain relatively crystalline (immature) material that grade into local background (mature) soils. Selected craters are small enough that they can be assumed to be of constant composition and thus are useful for evaluating trends in optical maturity. Color composites were also used to identify the most immature material in a region and show that maturity trends can also be identified using regional soil trends. The high resolution M3 data are well suited to quantifying the spectral changes that accompany space weathering and are able to capture subtle spectral variations in maturity trends. However, the spectral changes that occur as a function of maturity were observed to be dependent on local composition. Given the complexity of space weathering processes, this was not unexpected but poses challenges for absolute measures of optical maturity across diverse lunar terrains.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[2] Space weathering has a significant effect on the optical properties of lunar regolith. Its effects on lunar reflectance spectra are both a boon to those who wish to study space weathering itself and a bane to those who wish to study, with remote compositional measurements, other physical processes on the lunar surface. As a result, it has been the subject of considerable study both in petrologic and remote sensing literature [e.g., Housley et al., 1973; Hapke et al., 1975; Morris, 1976, 1978; Housley, 1979; Pieters et al., 2000; Hapke, 2001; Noble et al., 2007; Taylor et al., 2010]. It has been shown that the physical process largely responsible for the optical changes associated with space weathering is the production of nanophase iron (npFe0), nanometer-sized particles of native metallic iron. This has been produced by micrometeorite, impact-produced vaporization and subsequent deposition of silica-rich glass, containing myriads of these npFe0 grains, as rims on virtually all particles in a mature soil. Controversy exists as to the contributions made to the formation of this npFe0 from solar wind sputtering [Pieters et al., 2000; Hapke, 2001; Taylor et al., 2001; Noble et al., 2007; Taylor et al., 2010]. These npFe0 particles occur either as extremely fine grained rims on mineral grains deposited during micrometeorite impact and sputtering or as slightly larger particles in welded glassy aggregates of rock and mineral fragments called agglutinates.

[3] The optical effects that accompany npFe0 production have been shown to commonly include three interrelated effects: reduction in contrast of mafic absorption bands; increase in spectral slope (“reddening”); and a darkening of visible to near-infrared albedo [Fischer and Pieters, 1994, 1996; Pieters et al., 2000; Hapke, 2001; Noble et al., 2007]. The fact that these changes in spectral properties correlate with other measures of regolith maturity (e.g., Is/FeO) [Morris, 1976] has led to attempts at developing indices based on spectral properties that allow remote mapping of maturity of the lunar surface. Lucey et al. [1995] used the radial distance between a pixel of interest and a hypermature end-member in a scatterplot of 0.75 μm reflectance versus 0.95 μm/0.75 μm reflectance ratio as a maturity index for Clementine UVVIS data. This approach provided a practical tool for separating the first-order compositional (FeO) differences from maturity for many regions of the Moon [Grier et al., 2001; Lucey et al., 2000a, 2000b]. However, some regional differences in mineralogy do not appear to be fully captured by the Clementine parameters, resulting in potential errors in estimating absolute maturity or FeO contents across different regions of the Moon [Clark and McFadden, 2000; Elphic et al., 2000; Staid and Pieters, 2000]. For example, Staid and Pieters [2000] showed that the position of absolute values of this index, when applied to mare soils, is affected by the types of minerals in the regolith, particularly in the abundance of pyroxene and opaque phases (ilmenite and Fe spinel). Wilcox et al. [2005] confirmed the results of Staid and Pieters and improved the Lucey et al. [2000a, 2000b] index for specific applications to mare basalts by considering the average slope of mare basalt soil weathering trends in the Clementine two-parameter space.

[4] In this work, we assess the ability of new, high-resolution spectral data provided by the Moon Mineralogy Mapper (M3) mission to describe optical maturity trends of various geologic units in terms of the reddening, darkening, and contrast-reducing effects of npFe0. We have used scatterplots of spectral data and associated parameters from study regions around fresh craters, with relatively crystalline material in their ejecta, to show that simple parameterizations can reveal the maturity trends expected for mare and for highlands soils. We then assess the applicability of the currently available high-resolution imaging spectrometer data to global maturity indices.

2. Data and Methods

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[5] The Moon Mineralogy Mapper is a pushbroom imaging spectrometer aboard India's Chandrayaan-1 that acquired data in the 0.43–3.0 μm wavelength range. The data presented here were acquired in the lower resolution mode that has 85 usable spectral bands. We used data from part B of M3's first optical period, which operated at a 100 km elevation orbit giving a 140 m/pixel spatial resolution [Boardman et al., 2011]. For a detailed description of the M3 and Chandrayaan-1 operations and data acquisition strategies, including descriptions of M3's optical periods, the reader is referred to Boardman et al. [2011]. The M3 data presented here were calibrated to radiance using prelaunch and in-flight calibration coefficients for M3's radiance Version K (R. O. Green et al., The Moon Mineralogy Mapper (M3) imaging spectrometer for lunar science: Instrument, calibration, and on-orbit measurement performance, submitted to Journal of Geophysical Research, 2011). Table 1 gives a complete list of all M3 files used in this study, along with the latitude and longitude range of the data used and their average phase angle.

Table 1. List of M3 Observations Used in This Study
CraterM3 ObservationLongitudeLatitudePhase Angle (deg)
South RayM3G20090204T11344415.39−9.1846.26
Mare SerenitatisM3G20090107T13022524.2919.7627.40
Mare Tranquilitatis 1M3G20090203T16045226.463.2444.70
Mare Tranquilitatis 2M3G20090203T16045226.131.0740.24
Mare Humorum 1M3G20090208T080838−35.37−20.4761.84
Mare Humorum 2M3G20090208T160125−40.06−24.9657.58
Feldspathic Highlands 1M3G20090426T061945−148.23−0.4252.45
Feldspathic Highlands 2M3G20090426T061945−147.87−2.2546.26
Feldspathic Highlands 3M3G20090426T142319−152.03−14.1749.02
Mafic Highlands 1M3G20090426T061945−148.1537.5551.30
Mafic Highlands 2M3G20090426T061945−148.4035.1253.89
Mafic Highlands 3M3G20090426T061945−147.65−2.9242.37

[6] Because several sets of M3 data taken at various viewing geometries are compared in this study, Version K radiance data were converted to reflectance, using M3's initial photometric corrections, which normalizes all observations to an effective standard viewing geometry of i = 30° and e = 0°. This correction starts with an intermediate conversion of radiance to I/F, which is calculated as radiance multiplied by pi and divided by a solar spectrum that has been corrected for Moon-Sun distance. A Lommel-Seeliger scattering term is then removed from the I/F data, leaving a quantity that depends primarily on phase [Hillier et al., 1999]. This quantity is applied to a model-derived look up table that completes the conversion to reflectance and normalizes to the standard viewing geometry. Hicks et al. [2011] discussed M3's initial photometric correction strategy in detail. Currently, the M3 photometric corrections are able to normalize observations to a standard viewing geometry in terms of absolute albedo, but do not address sufficiently the wavelength-dependent effects of viewing geometry on spectral data. Specifically, the reddening of lunar spectra that is known to occur with increased phase angle [Hillier et al., 1999] has not been addressed. This means that a systematic, but as yet unquantified, error is present in our spectral slope and albedo calculations. However, because measurements are only used for comparison within single geologic units, rather than a comparison across multiple units, the net effect of this error should be small. However, while we believe that the observations reported here are reliable, they should be considered preliminary until a complete photometric correction can be applied.

[7] A correction factor was applied to suppress residual band-to-band artifacts in the spectra. This correction factor, referred to internally by the M3 team as “R4RC1,” is a mild gain factor, with values very close to unity (and in fact are set to unity at wavelengths greater than 2700 nm), which is divided out of the reflectance data [Clark et al., 2011]. This correction factor is derived from the mean of a set of featureless M3 spectra with removal of a straight line continuum and residual curvature. The data presented here have not had an estimate of thermal emission removed, so the work presented here avoids long wavelengths (>2.2 μm), particularly the 2 μm mafic absorption band.

[8] The primary goal of data selection for maturity analysis was the separation of the variance in reflectance, due to maturity, from the rest of the reflectance “signal.” We approached this daunting task by looking at small, fresh craters in an extended unit. The ejecta surrounding recent, small-impact craters are relatively crystalline and, because of the size of the crater, the crater and the ejecta derived from it can be assumed to be the same composition. Therefore, the dominant signal in the spectral variation in these materials should be due to maturity. We used a modified version of the Small Crater Rim and Ejecta Probing (SCREP) [cf. Kramer, 2010] procedure to create regions of interest (ROIs) around the ejecta blankets of craters, particularly those that were identified as being small enough that they were unlikely to have penetrated to any different underlying bedrock layer and had identifiable changes in maturity parameters. The SCREP procedure uses crater depth-to-ejecta thickness relationships to define, in terms of crater radii, an ROI containing the rim and proximal ejecta of an impact crater, which presumably contains the most immature material. We modified the SCREP procedure so that the modeled ROI extends to 5 crater radii from the rim in order to better capture a range of maturity variation, rather than the most immature end-members.

[9] An important aspect of this method of selecting immature material is the choice of crater size. The selected craters have to be large enough to have ejecta that fills several pixels, but also small enough that it can be reasonably inferred that the ejecta samples a single geologic unit. For this study, craters were chosen between 0.3 and 1.2 km in diameter, with a mean of approximately 450 m. However, more important than the absolute size of the craters being included in a maturity analysis is that care be taken to look for spectral evidence that the assumption of homogenous ejecta is valid. Several craters identified for possible inclusion in our study that were within the 0.3–1.2 km range were ultimately excluded, because we thought their ejecta may have sampled more than one unit. This was based on scatterplots of spectral indices (discussed below). If the trends showed “kinks” that suggested mixing of two maturity trends (for two different geologic units), or if the trend was very different from that of other craters in the vicinity (again suggesting a different unit was sampled), we excluded them. Ultimately, the successful identification of immature material depends on careful sample selection on the part of the analyst.

[10] As discussed earlier, exposure to space weathering is known to affect spectra in three ways. We looked for simple parameter formulations that best capture these three variations in M3 data. After considering several parameters, we selected an integrated 1 μm band depth to estimate spectral contrast in a ferrous absorption band, a near-infrared continuum ratio to estimate reddening, and an albedo parameter to estimate darkening. The albedo parameter we chose is simply the reflectance at 1.58 μm to avoid any spectral region associated with ferrous absorptions in pyroxene and olivine. For the continuum parameter, we chose a simple ratio of two points on either side of the 1 μm ferrous absorption, rather than a true slope, which would also be dependent on absolute albedo. To distinguish this formulation from other continuum measures that are based on true slope (ΔY/ΔX) measurements, we will refer to this parameter as the “continuum ratio.”

[11] We chose an integrated 1 μm band depth measurement rather for our estimate of spectral contrast. This is calculated as the sum of the continuum-removed band depths of all channels within the 1 μm ferrous absorption band. Because this measures the band strength of the entire absorption band relative to a continuum, this is a more robust measure of absorption band depth than simple ratios often used in multispectral data such as Clementine UVVIS data. However, it should be noted that because our integrated band depth estimation is a measure of actual band strength, stronger absorptions are indicated by higher values of our band depth parameter. The exact formulations of all three parameters are given in Table 2. While we are confident that each parameter gives dependable results, we note that ongoing calibration issues in the M3 data related to scattered light in the detector (Green et al., submitted manuscript, 2011) make the results presented here preliminary.

Table 2. Descriptions of Parameters Used in This Worka
Short NameParameterFormulation
  • a

    R#### notation refers to the reflectance at a particular wavelength given in nanometers. For example, R1579 refers to the reflectance at 1579 nm. Numbers (without the R#### notation) refer to the wavelength in nanometers.

Albedo1.6 μm reflectanceR1579
Band Depth1 μm integrated band depthequation image RC = 1 μm continuum reflectance at the given wavelength; calculated as a straight line at all M3 wavelengths with slope [(R1508 − R730)/(1508 − 730)] (sums band depths from 730 to 1508 nm)
Continuum RatioContinuum RatioR1508/R730

3. Results

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

3.1. Fresh Craters

[12] Figure 1 plots examples of maturity trends for fresh highland and mare craters. The mare trend is from Mare Serenitatis, and data for South Ray crater (near the Apollo 16 landing site) represents a highlands trend. The bottom, left, and rear “sides” of Figure 1 form Figures 2a, 2b, and 2c (see notations in Figure 1). Because of the difficulties associated with clearly depicting a three-dimensional data space in a two-dimensional plot, the combination of Figures 2a2d will be used throughout the paper as a standard way to show maturity variations. Figure 2a plots band depth versus continuum ratio. Le Mouélic et al. [2002] used a plot similar to this to determine maturity trends for a region surrounding Tycho Crater. Mare soils should, in general, show significant variations in this type plot. Highland soils, however, should show significant variations in Figure 2b, which plots the continuum ratio versus albedo. Figure 2c is a plot of albedo versus band depth. Scatterplots of these two spectral properties are perhaps most familiar since they form the basis for the OMAT parameter defined by Lucey et al. [1995]. However, as previously noted, our band depth parameter is formulated differently than typical band depth ratios such as those often applied to Clementine data UVVIS (an integrated band depth for M3 data versus an 0.95 μm/0.75 μm reflectance ratio for Clementine data). Because our formulation is a summation of band strength relative to a continuum and the Clementine formulation is a ratio, higher numbers in our parameter indicate greater band depths, while in the Clementine ratio, lower numbers suggest stronger band depths in the Clementine ratio. Figures 2d is the same as Figure 2c but with the axes exchanged and the data in the y axis reversed to allow a more ready comparison to previous results such as those from analyses of Clementine UVVIS data [e.g., Lucey et al., 2000b].

image

Figure 1. Three-dimensional scatterplot of optical maturity variations for a fresh crater in Mare Serenitatis (red) and South Ray Crater (blue) located near the Apollo 16 landing site. The x axis is integrated band depth, the y axis is albedo, and the z axis 9 s continuum ratio. The bottom, left, and rear sides form Figures 2a2c (see labels). Arrows indicate direction of increasing maturity.

Download figure to PowerPoint

image

Figure 2. A set of four two-dimensional plots containing the same data as in Figure 1: (a) band depth versus continuum ratio, (b) continuum ratio versus albedo, (c) albedo versus band depth, and (d) same as Figure 2c but with axes interchanged and data in the y axis reversed. This set of four plots will be the standard way maturity variations are displayed in this paper.

Download figure to PowerPoint

[13] Figure 3 shows the location of a Mare Serenitatis Crater plotted in Figure 2, along with a radial transect of representative spectra. A similar transect for South Ray Crater, the highlands crater whose maturity trends are illustrated in Figures 1 and 2, is shown in Figure 4. From inspection of Figures 3 and 4, the relation of trends shown in Figure 2 becomes clear. The trend for Mare Serenitatis, based on the behavior of this small, fresh crater, is that much of the variation in the three-parameter space is in the 1 μm band depth and the continuum, while the albedo variation is less significant. For the typical highlands, such as South Ray Crater, the maturity trend is different in that it has little variation in band depth, but does vary to some degree in its continuum ratio. Most of the variation for this crater, however, is in the albedo. These are typical trends for many mare and highland surfaces.

image

Figure 3. (a–c) Location of the Mare Serenitatis crater plotted in Figures 1 and 2. (d) A transect of reflectance spectra for this crater is also shown. (e) Same as Figure 3d but with all spectra scaled to unity at 750 nm. Dashed vertical lines are the wavelengths used for continuum ratio and limits of summation for integrated band depth. Most of the spectral variation is in the band depth and continuum ratio. Dashed box in Figure 3c is the location of the mosaic shown in Figure 9.

Download figure to PowerPoint

image

Figure 4. (a–c) Location of South Ray Crater plotted in Figures 1 and 2. (d) A transect of reflectance spectra for this crater is also shown. (e) Same as Figure 4d but with all spectra scaled to unity at 750 nm. Dashed vertical lines are the wavelengths used for continuum ratio and limits of summation for integrated band depth. Most of the spectral variation is in the band depth and continuum ratio.

Download figure to PowerPoint

[14] Maturity trends for craters in two additional mare regions, Mare Tranquilitatis and the Mare Humorum, are plotted in Figures 5 and 6, respectively. In Figures 5 and 6, data from two fresh craters are plotted to show that the data from different craters in the same region are consistent. The general similarity between trends in Figures 5 and 6 and those shown for Mare Serenitatis in Figure 2 confirms that optical maturity systematics expected of mare units are being captured by the M3 data with the three parameters employed.

image

Figure 5. Fresh crater data for Mare Tranquilitatis. Two craters are shown to show that results are consistent. Plots are the same configuration as in Figure 2.

Download figure to PowerPoint

image

Figure 6. Fresh crater data for Mare Humorum. Two craters are shown to show that results are consistent. Plots are the same configuration as in Figure 2.

Download figure to PowerPoint

[15] Maturity trends for additional highlands craters are shown in Figures 7 and 8. All craters plotted in Figures 7 and 8 were taken from an area in the western farside highlands surrounding Crater Timiryazev (147°W, 5°S). Craters containing relatively feldspathic material are plotted in Figure 7, whereas craters with more mafic regolith are plotted in Figure 8. Highlands materials with relatively feldspathic compositions, similar to those of South Ray Crater ejecta shown in Figure 2, exhibit little variation in band depth but significant variation in albedo. Continuum ratio variation is small in these craters but larger in mafic highlands craters (Figure 8). The relatively mafic highlands craters, which of course exhibit greater band depths than the feldspathic highlands craters, form a nearly linear trend in the band depth versus continuum ratio plot (Figure 8a). However, the trends in Figures 8b8d are quite different. The three trends hardly overlap in these other parameter spaces, and one crater (light gray circles) shows a trend that mimics a mare trend. All three parameter spaces are required to capture these important differences in maturity trends.

image

Figure 7. Fresh crater data for three fresh craters in the western farside north of Crater Timiryazev. The crater data plotted are for normal highlands materials that do not have mafic absorption bands. Plots are the same configuration as in Figure 2.

Download figure to PowerPoint

image

Figure 8. Fresh crater data for three fresh craters in same region as in Figure 6 but for craters that have mafic absorption bands (i.e., are noritic). Plots are the same configuration as in Figure 2.

Download figure to PowerPoint

3.2. Regional Soils

[16] We chose Mare Serenitatis for a more regional-scale analysis of mare soils in addition to local analyses of fresh craters within it. It was chosen because the fresh crater data reveal relatively simple trends, and the basalt has low-titanium content [Prettyman et al., 2006] (ilmenite may introduce a bias in the maturity “signal”). We created a mosaic of M3 data at full 140 m/pixel (global mode) resolution to investigate soil variation in this region. Figure 9a shows the continuum ratio image for these soils, while the Figure 9b shows the albedo image and Figure 9c shows the band depth image. Figure 9d is a color composite of Figures 9a9c. In this composite, the continuum ratio has been mapped to the red channel, the green channel contains the albedo parameter, and the band depth parameter has been mapped to the blue channel. For both the albedo and band depth parameters in the green and blue channels, respectively, the color planes have been inverted, giving a color composite where low values in each channel are indicative of immature soils. Thus, dark pixels in this color composite represent the least mature material in the region. In most cases, these dark pixels are found around fresh craters. However, there are several craters, particularly in the middle and bottom right areas in the composite, where there is a large albedo change but little change in band depth or continuum ratio. These are the craters that are likely to be sampling a different subsurface unit rather than showing a difference in degree of space weathering. These tend to be larger craters, and they do not exhibit the variation in band strength associated with maturity trends for this region.

image

Figure 9. Mare Serenitatis mosaic. (a) Continuum ratio image, (b) albedo image, (c) integrated 1 μm band depth image, and (d) maturity color composite: R, continuum ratio; G, albedo; B, band depth. Green and blue channels are inverted so that dark material is the most immature. North is to the top. Red circle in Figure 9b denotes fresh crater used in Figure 10.

Download figure to PowerPoint

4. Discussion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[17] Because of their obvious usefulness, many maturity indices based on spectral properties of lunar surface material have been used in the literature. The OMAT index, defined by Lucey et al. [1995], when applied to Clementine UVVIS data, successfully identified relative maturity variations for regional soils using a two-parameter (albedo and band ratio) space. Other two parameter spaces have been explored in the literature as well, such as the continuum versus band depth space used by Le Mouélic et al. [2002], to show maturity trends in highlands materials surrounding Tycho Crater. While these parameter spaces are appropriate for defining overall differences in maturity across a given terrain, our analysis of M3 data shows that the use of simple parameter spaces often masks subtleties in maturity variation that are revealed when three parameters are used to describe maturity variations. In Figure 2, trends for Mare Serenitatis appear essentially linear in albedo versus band depth space (Figure 2c), but the trend is shown to have a slightly curvilinear relationship in band depth versus continuum ratio space (Figure 2a). Maturity trends for the Mare Humorum Crater ejecta show similar behavior in Figure 6, though to a lesser degree. In Figure 8, crater ejecta for highlands materials with mafic absorption bands show simple linear behavior in band depth versus continuum ratio space (Figure 8c), but the other two parameter spaces in Figures 8b and 8c show much more complex behavior. These diverse variations would need to be accounted for by a global maturity index that seeks to rigorously quantify maturity variation for all lunar regions.

[18] Another challenge that absolute maturity indices must overcome is that the spectral properties they are based on are often affected by factors other than maturity differences, most notably differences in composition. This most significantly affects the application of a maturity index across regional boundaries. For example, with Clementine data, Staid and Pieters [2000] showed that in a simple albedo versus band ratio parameter space, the mature mare soils for Mare Frigoris have OMAT values similar to values for immature material in central Mare Serenitatis. This indicates that mare maturity variation is not completely separated from compositional variation in this parameter space. Our data demonstrate this as well. Figure 10 shows the combined band depth versus continuum ratio trends for the crater data shown in Figures 1 and 48. The trend for Mare Tranquilitatis is both rotated and offset relative to the Mare Humorum and Mare Serenitatis trends. Such differences in the observed trends must be due to differences in compositions between the regions. For example, some of the most mature material for the mafic highlands crater plots in the same region as the least mature material for a feldspathic highlands crater trend. For an index to be meaningful in an absolute sense, the most mature surface material in both cases would have to be given the same parameter value. Variations in soil chemistry preclude convergence with simple indices.

image

Figure 10. (a) Integrated 1 μm band depth versus continuum ratio trends for crater data from Figures 1 and 48. (b) Convex hulls (outlines) of trends shown in Figure 10a.

Download figure to PowerPoint

[19] Figure 11 further illustrates this point. Because the majority of the lunar surface is comprised of mature soils, a global map of variations in the three parameters used in this paper is an indication of the final products of the space weathering process. Figure 11 is an RGB composite of these three parameters. The color mappings in this composite are the same as in Figure 9: the red channel contains the continuum ratio; the green channel contains the albedo parameter; and the blue channel is a map of the band depth parameter. Again, the green and blue channels have their gray scales inverted, so that dark tones indicate immature material. In Figure 11, most highlands areas have blue tones, indicating an unsurprising dominance by the albedo channel (because the green channel has been reversed), or magenta tones, indicating relatively high-albedo and continuum ratio changes relative to band depth. Yellow tones in Figure 11, occurring mostly in the maria, are indicative of high band depths, relative to albedo and continuum ratio, and green tones indicating equal contributions from both continuum ratio and band depth. It is clear that lunar materials do not mature to a common endpoint, which would have to be the case for a robust single absolute index. Because of these problems, we avoided the temptation to attempt an M3 global maturity parameter that reflected absolute maturity variations across the Moon. We advocate using the three parameters explored in this paper to define regional maturity trends, from which the most immature soils can be determined through careful spectral analysis. Within a homogenous area, optical maturity trends as defined by three simple spectral parameters are quite regular, making the task of separating mature from immature material relatively straightforward.

image

Figure 11. Maturity color composite for the lunar surface (based on optical period 1B data only): R, continuum ratio; G, albedo; B, band depth. Green and blue channels are inverted so that dark material is the most immature. North is to the top in these images.

Download figure to PowerPoint

5. Conclusions

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[20] The Moon Mineralogy Mapper has provided spectral data with high spatial and spectral resolution that enables many new avenues of research related to the optical effects of space weathering. This paper shows that the new M3 data can be used to show maturity trends in a consistent way for many areas of the Moon. However, it also shows that the maturation process is very complex, and it is the nature of the maturation process, not the capabilities of a sensor, that makes unlikely a single absolute maturity index that can be used across lunar terrains. Because of compositional differences, each type of lunar soil can have its own starting and ending point in the maturity parameter space, and can even have a unique pathway between those points. The strength of high-resolution imaging spectrometer data, such as those provided by M3, is that it is can reveal much more subtle detail in maturity trends. It is expected that this will enable many new avenues of research leading to much better understandings of the complicated process known as space weathering and its effects on spectral data.

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[21] M3 is supported as a NASA Discovery Program mission of opportunity. The science results are supported through NASA contract NNM05AB26C. The M3 instrument was flown as a guest instrument on Chandrayaan-1. The M3 team is honored to have the opportunity to participate in this mission and is grateful for the work of the ISRO team that enabled M3 data return.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
jgre2896-sup-0001-t01.txtplain text document1KTab-delimited Table 1.
jgre2896-sup-0002-t02.txtplain text document1KTab-delimited Table 2.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.