9512.net
甜梦文库
当前位置:首页 >> 林学 >>

Temperature-Dependent Near-Infrared Spectral Properties of Minerals, Meteorites, and Lunar Soil


Icarus 155, 169–180 (2002) doi:10.1006/icar.2001.6754, available online at http://www.idealibrary.com on

Temperature-Dependent Near-Infrared Spectral Properties of Minerals, Meteorites, and Lunar Soil
John L. Hinrichs and Paul G. Lucey
Hawaii Institute of Geophysics and Planetology, University of Hawaii at Manoa, 2525 Correa Road, Honolulu, Hawaii 96822 E-mail: lucey@higp.hawaii.edu Received April 16, 2001; revised June 20, 2001

The near-IR spectral properties of minerals, meteorites, and lunar soil vary with temperature. The manner in which these materials vary is diagnostic of aspects of their composition. We quantify the spectral dependence on temperature by reporting the change in relative re?ectance with temperature as a function of wavelength. We call this quantity, R/ T (in units of K?1 ), as a function of temperature the “thermo-re?ectance spectrum.” The thermo-re?ectance spectra of olivine and pyroxene are distinct, and most of the observable structure in thermo-re?ectance spectra of the ordinary and carbonaceous chondrites can be understood in terms of a mixture of the thermo-re?ectance spectra of olivine and pyroxene. The magnitude of thermo-re?ectance spectra of meteorites and lunar soils is much less than that of pure minerals. Lunar soils are particularly subdued. While conventional analysis of remotely obtained spectra of the Moon can neglect temperature effects, spatially resolved measurements of the surface of the asteroid Vesta will likely have a strong temperature-dependent component based on measurements of a eucrite and a howardite. c 2002 Elsevier Science (USA) Key Words: infrared observations; spectroscopy; asteroids; mineralogy; meteorites.

but were consistent with spectra of olivines at low temperature. Lucey et al. (2002) reported that spectral variation with temperature is present on Eros, and the nature of this variation suggests an LL chondrite-like surface composition. In this paper we report new laboratory measurements of the near-infrared (.5–2.3 ?m) spectra of minerals, meteorites, and lunar soils over a range of temperature. We show that the sensitivity of the re?ectance of olivine and pyroxene to temperature vary with wavelength in a diagnostic way, such that information on the relative abundance of olivine and pyroxene in rocks containing these minerals can be discerned. We also present the magnitude of the effect for these materials and show the effect of lunar space weathering on temperature effects in lunar soils.
2. EXPERIMENT DESIGN

1. INTRODUCTION

Roush and Singer conducted a series of laboratory spectral observations using an evacuated environment chamber with controllable temperature to show that the re?ectance spectra of several ma?c silicates vary with temperature (Roush 1984, Singer and Roush 1985, Roush and Singer 1986). Roush and Singer (1987) pointed out that this effect might be important for spatially resolved observations of asteroids where surface temperatures could vary by as much as 100 K for illuminated portions of the asteroid. Later, with receipt of spatially resolved measurements by Galileo and in anticipation of much more data from NEAR, Hinrichs et al. (1999), Schade and Wasche (1999), and Moroz et al. (2000) expanded on this theme with new measurements and models and reiterated the issues raised by Roush and Singer. Observation of a temperature-induced spectral effect was reported by Lucey et al. (1998), who showed that the spectra of the olivine-rich A-asteroids were not consistent with those of olivines of any composition under terrestrial ambient conditions
169

Measurement of temperature effects on optical properties of powders is complicated by the fact that powders are highly insulating, which promotes the development of thermal gradients within the sample (Henderson and Jakosky 1994). Thermal gradients cause the temperature of the optical surface to differ from that of the sample interior or base where in situ temperature measurements might be made. Thermal gradients are promoted by measurements in vacuum, which are often done to prevent deposition of frost or oxidation of samples, because gas conduction is minimized. Moroz et al. (2000) addressed this problem by measuring their samples in a dry purge gas to promote gas conduction. In the experiments described in this paper we took a different approach, measuring samples in vacuum but carefully controlling the radiative environment of the sample to minimize temperature gradients. In their experiments Roush and Singer placed a thermocouple inside the sample near the surface, but near-surface thermal gradients can be large with skin depths on the order of a few tens of micrometers. One aspect of the Roush and Singer experiment was that the radiative environment was not controlled nor documented suf?ciently to model the in?uence of radiation. Because the thermal conductivity of powders is notoriously low and the radiation effects large, we designed a chamber that allowed control of the radiation environment both to satisfy our concern

0019-1035/02 $35.00 c 2002 Elsevier Science (USA) All rights reserved.

170

HINRICHS AND LUCEY

that radiative effects might in?uence the temperatures reported and to characterize the importance of radiative effects. Our approach to controlling the environment is to completely surround the particulate samples in a chamber of uniform temperature and to illuminate and observe these samples using windows made of BK7 optical glass, which is opaque to thermal radiation but admits NIR light for our measurements (Fig. 1). In this way the samples are effectively in a blackbody cavity and the optical surface will approach the temperature of the interior. More importantly, the optical surface will approach the temperature of the walls of the chamber where we measure the temperature. The major violation of this near-blackbody condition is the radiation of our illuminating light source. We will show that radiative heating by our light source probably limited our measurements to a minimum temperature of about 150 K. The temperature-controlled chamber is itself contained within an outer chamber, that acts as a vacuum jacket and has its own radiation shield, and can be cooled near liquid nitrogen temperatures or left at ambient temperature. The inner chamber is heated (or cooled) through its conductive connection to a temperature-controlled plate upon which the sample sits. In initial experiments we sometimes operated with the inner chamber removed. Spectra were obtained using an Analytical Spectral Devices FR ?eld re?ectance spectrometer. The instrument covers .35 to 2.5 ?m at 3-nm resolution shortward of 1 ?m and 10-nm resolution longward of 1 ?m. Below 1 ?m a photo-diode array spectrometer is used; two scanning grating spectrometers, each using a thermoelectrically cooled, single-element InGaAs detector, cover wavelengths beyond 1 ?m. The ?ber optic input to the spectrometer is transferred to the sample via an in-house built lens assembly. Data are referenced to a Spectralon re?ectance standard and a broadband collimated light source. All measurements were made at a photometric geometry of i = 10.5? , e = 10.5? and phase = 21.0? . No correction is made for the weakly non-Lambertian behavior of Spectralon.
3. RESULTS

of 100 to 400 K. In the third experiment we again did not use the inner chamber, and this time the outer radiation shield temperature was close to ambient and we again varied the baseplate temperature on which the sample rests over the same range of 100 to 400 K. The results are shown in Fig. 2. The green lines are spectra obtained with the temperature of the background equal to the sample base temperature. They show the largest variation in spectral properties of the three experiments. The blue lines are for the room temperature background and the red lines are for the low temperature background. In the (near) constant background cases the spectral variation is much less than in the case of the variable background. Clearly there is coupling of the spectral properties with the radiative environment in these experiments. This is illustrated further by Fig. 3. There is a strong dependence of the 1.4-?m re?ectance of this sample on baseplate temperature in the experiment where the background temperature equaled the baseplate temperature, while in the other experiments only a weak dependence of re?ectance on baseplate temperature is observed. In the “constant” background cases there was some coupling between the outer radiation shield and the sample temperatures. For example, in the cold background case the outer radiation shield temperature increased from 87 to 103 K while the sample was increased in temperature from 100 to 400 K. In the warm background case the radiation shield increased from 286 to 294 K while the sample temperature increased from 100 to 400 K. Therefore, some of the re?ectance variation observed in the constant background cases is probably due to the small variation in background temperature. While we have not performed formal thermal modeling of these samples, these experiments clearly demonstrate that radiation is dominant in controlling optical surface temperature for this particulate sample and that good control of the radiative environment is essential to make reliable measurements of temperature effects of spectra of particulates in a vacuum. The second aspect of the radiative environment is heating by the illuminating light source. To study this effect we modeled the sample surface as one with only radiative coupling to the environment with the additional load of illumination heating by
4 4 σ Tbackground + R(I ) = σ Tsample ,

3.1. Effect of the Radiative Environment on Sample Spectra We assessed the importance of the radiation environment by measuring a single sample in three different radiative conditions: 1. radiative background equal to sample base temperature, 2. radiative background roughly constant at ?90 K, and 3. radiative background temperature roughly constant at ?300 K. The sample was an Fo86 olivine, ground and sieved to 20–63 ?m. In the ?rst experiment we collected spectra with the inner chamber installed and varied the temperature of the baseplate from 100 to 400 K. The inner chamber walls were in good conductive contact with the temperature-controlled baseplate and were always within a few degrees of the baseplate temperature. In the second experiment we removed the inner chamber and cooled the outer radiation shield to a temperature of ?90 K but varied the baseplate temperature on which the sample rests over the range

where σ is the Stefan–Bolztmann constant, R is the re?ectance of the sample, I is the irradiance of the light source, Tbackground is the temperature of the background, and Tsample is the temperature of the sample. We used a vendor-provided absolute radiance calibration for our spectrometer to determine the irradiance input by our light source, which was 44 W/m2 . Under the assumption of no conductive heat transfer and instantaneous equilibrium, Fig. 4 plots the sample chamber (background) temperature versus the model sample surface temperature for the irradiance of our light source and a range of re?ectances. The effect of light source heating is marked at the lower temperatures, even though our illumination source is only 1/30 the intensity of sunlight at one A.U.

TEMPERATURE-DEPENDENT SPECTRAL PROPERTIES

171

FIG. 1.

Schematic of the radiation controlled environment chamber and photograph with cover removed showing the inner chamber.

FIG. 2. Spectra of powdered (20–63 ?m grain size) forsteritic (Fo86 ) olivine under various radiative conditions. Green lines represent the spectra obtained within the constant-temperature cavity. Blue lines are spectra obtained with a near ambient radiation shield temperature, and red lines are spectra obtained using a background temperature near 90 K. In all cases the sample base temperature was varied from 100 to 400 K.

172

HINRICHS AND LUCEY

FIG. 3. 1.4-?m re?ectance of the forsteritic olivine as a function of sample base temperature for three background conditions. Dots represent measurements with sample based and background temperature equal. Triangles are measurements with background temperature near 90 K, and crosses are measurements with background temperature near 290 K. When the background is held equal to sample base temperature the re?ectance depends strongly on temperature. When the background temperature is nearly constant, the dependence of re?ectance on the sample base temperature is weak.

FIG. 5. Re?ectance of pyroxene at 1.05 ?m versus temperature. The crosses are sample base temperatures. The dots are the same data with temperatures adjusted to represent sample surface temperature given 30% albedo and sample illumination of 44 W/m2 and zero sample thermal conductivity. Note that the break in slope remains even after the correction.

Measurements of re?ectance at low temperatures in our apparatus is consistent with this effect. Figure 5 shows that the re?ectance of a pyroxene rises with decreasing background temperature but begins to ?atten out with decreasing temperature. Applying our simple model does not completely linearize the relationship between temperature and re?ectance for this sample at this wavelength. In part this is probably due to the shifting of the ferrous band away from this wavelength. In the absence of more experiments, we simply note that radiative heating by

the light source is an important effect, especially for measurements of powders in vacuum, but not necessarily to be ignored in measurements at higher pressures. For the rest of this paper, we con?ne our analysis to portions of the temperature–re?ectance curves which are relatively linear and do not show obvious evidence of rollover due to heating. The low temperature cutoff varies with sample albedo. 3.2. Spectral Properties of Selected Minerals as a Function of Temperature We measured four minerals in these series of experiments, three olivines and one orthopyroxene. In this section we introduce a new concept, the spectral sensitivity of minerals to temperature variation, which we term the “thermo-re?ectance spectrum.” The thermo-re?ectance spectrum is a plot of the change in re?ectance of a sample, relative to the sample measured at 300 K, per unit temperature change, as a function of wavelength. This curve quanti?es the degree to which a mineral or other sample spectrum responds to temperature. We show that the shape of the thermo-re?ectance spectra of olivine and pyroxene are distinct and diagnostic of the presence of these minerals. 3.2.1. Olivines. The three olivines measured are similar in composition (Fo86 , Fo88 , and Fo89 ). The samples were ground and wet-sieved in methanol to size ranges of 20–63 32–63, and <45 ?m respectively. The spectrum of a typical temperature series is shown in Fig. 6. In all cases the band due to ferrous iron in the olivine broadens with increasing temperature, and our spectra are qualitatively consistent with the spectra of Roush (1984). See Singer and Roush (1985) for a discussion of the physical mechanisms behind these effects. Figure 7 shows the relationship between re?ectance and temperature for selected wavelengths for the Fo89 sample to illustrate the change in behavior with wavelength. It is clear that there are strong differences with wavelength (as is evident from

FIG. 4. The increase of sample surface temperature due to sample heating by the illumination at 44 W/m2 may be approximated using this family of curves generated from a worst-case scenario of zero sample thermal conductivity. The curves range from 0% albedo at top to 100% albedo at the bottom in 20% albedo increments.

TEMPERATURE-DEPENDENT SPECTRAL PROPERTIES

173

FIG. 8. Thermo-re?ectance spectra of three olivines: (a) Fo89 of <45 ?m, (b) Fo86 of 20–63 ?m, (c) Fo88 of 32–63 ?m. FIG. 6. Spectra of olivine (Fo86 ) with 20–63 ?m grain size obtained at sample base temperatures from 80 to 400 K in 20 K increments.

the temperature series spectra themselves). At 1.4 ?m there is a strong anticorrelation between temperature and re?ectance, with almost a 10% difference in re?ectance over 100 K. Other wavelengths show lesser dependence, with some showing no correlation at all. Below 150 K all wavelengths show rolloff in re?ectance that we attribute to light source radiative heating. We parameterize these data as a function of wavelength by ?tting the temperature–re?ectance curves at each wavelength with straight lines and report the slope of the curve as a function of wavelength. The curves are the thermo-re?ectance spectra, in units of R (relative to 300 K)/ T . The thermo-re?ectance spectra of the three samples are shown in Fig. 8. In all cases there

is a strong wavelength-dependent variation in the sensitivity of re?ectance to temperature, varying from no sensitivity, to slight positive correlation, to strong negative correlation. 3.2.2. Pyroxene. We measured an orthopyroxene (En86 ) wet-sieved to 45–90 ?m grain size. The sample was split from a larger sample also measured by Roush (1984). Detailed chemical analysis can be found in Singer (1981). Figure 9 shows the temperature series. As reported by Singer and Roush (1985), the 1-?m band broadens on the long-wavelength wing, and the 2-?m band both broadens and distinctly shifts to longer wavelengths. The thermo-re?ectance spectrum is shown in Fig. 10 and displays a sharp negative peak at 1 ?m, a relatively broad maximum near 1.7 ?m (indicating that this sample increases in re?ectance with increasing temperature at these wavelengths),

FIG. 7. Fo89 green sand beach olivine re?ectance relative to 300 K spectra. Diamonds are 0.8 ?m, asterisks are 1.0 ?m, triangles are 1.4 ?m, squares are 1.5 ?m, and plus signs are 2.0 ?m.

FIG. 9. Spectra of orthopyroxene (bronzite, En86 ) from 100 to 400 K. Note the broadening of the 1-?m band and the broadening and shift to longer wavelengths of the 2-?m band.

174

HINRICHS AND LUCEY

TABLE I Meteorites Measured
Meteorite El Hammammi Allegan Manbhoom Warrenton Murchison EET83251 EET87503 Type H5 ordinary chondrite H5 ordinary chondrite LL6 ordinary chondrite CO carbonaceous chondrite CM2 carbonaceous chondrite Eucrite Howardite

FIG. 10. grain size.

Thermo-re?ectance spectrum of orthopyroxene (En86 ); 45–90 ?m

on asteroids. The two major issues are whether this effect can in?uence interpretation of asteroid spectra, as was addressed by Moroz et al. (2000), and whether the effect, if detected, might be used as a compositional analysis tool. To this end we measured the thermo-re?ectance spectra of several chondrites and howardite/eucrite/diogenite (HED) meteorites (Table I). 3.3.1. Ordinary chondrites. As potential analogs for many asteroids, ordinary chondrites are obvious candidates for analysis. We measured the thermo-re?ectance spectra of one LL and two H type ordinary chondrites. The spectral temperature series are shown in Figs. 12 and 13. The temperature series of the LL6 chondrite Manbhoom is presented in Fig. 14. The ordinary chondrite results are summarized in Fig. 15, where the thermo-re?ectance spectra of the meteorites are shown. Both El Hammammi and Allegan (H5) have thermo-re?ectance spectra with smaller excursions than the pure minerals. This is not surprising given that these rocks are mixtures of components plus neutral metal, which does not contribute a temperature-varying

and a second negative feature with its minimum just beyond 2 ?m. 3.2.3. Mineral summary. Both olivine and pyroxene have spectra that are sensitive to temperature, consistent with earlier studies. Figure 11 shows the thermo-re?ectance spectra of olivine and pyroxene superimposed, illustrating the large difference between the two in terms of the wavelengths of greatest sensitivity. In subsequent sections, the presence of these minerals can be discerned in thermo-re?ectance spectra of rocks which contain them. 3.3. Meteorites As meteorites are samples ultimately derived from the asteroid belt, measurement of the thermo-re?ectance properties of these rocks can constrain the expected magnitude of this effect

FIG. 11. Thermo-re?ectance spectra of pyroxene (solid line) and olivine (dashed line) illustrating their marked differences. Olivine re?ectance shows little dependence on temperature at wavelengths where pyroxene is most sensitive.

FIG. 12. Temperature series for ordinary chondrite El Hammammi. The top spectrum was collected with the sample at 400 K. Spectra below are for sample temperatures decreasing at 20 K intervals. Each temperature spectrum is offset by .01 re?ectance units from the one above it (400 K spectrum is not offset). Note that this offset display obscures the fact that some wavelengths are completely inactive with respect to temperature. The corresponding thermore?ectance spectra quantify how the spectrum of the sample varies with temperature (multiple measurements at 300 K and 200 K).

TEMPERATURE-DEPENDENT SPECTRAL PROPERTIES

175

FIG. 13. Temperature series for H5 ordinary chondrite Allegan. The top spectrum was collected with the sample at 400 K. Spectra below are for sample temperatures decreasing at 20 K intervals. Each spectrum is offset by .01 re?ectance units from the one above it (top spectrum is not offset). Note that this offset display obscures the fact that some wavelengths are completely inactive with respect to temperature. The corresponding thermo-re?ectance spectra quantify how the spectrum of the sample varies with temperature.

FIG. 15. Thermo-re?ectance spectra of ordinary chondrites H5 Allegan (a), H5 E1 Hammammi (b), and LL6 Manbhoom (c). Spectrum (a) has been offset 2 units upward and spectrum (b) has been offset 1 unit upward. The contrast on these spectra is 3–4 times less than that of the thermo-re?ectance spectra of minerals, but they exhibit similar structure. The presence of pyroxene in the spectra of the H5 chondrites is indicated by the strong and narrow negative feature near 1 ?m whereas olivine is indicated in the LL6 meteorite by the strong positive feature just beyond 1 ?m.

spectral signal. Both meteorites have thermo-re?ectance spectra reminiscent of pyroxene, with minima near 1 and 2 ?m, and maxima near 1.7 ?m. There is a suggestion of distortion of the 1-?m micron minimum which may be due to the presence of olivine. In contrast, the thermo-re?ectance spectrum of the LL6 chondrite Manbhoom is similar to that of olivine, with a maximum, rather than a minimum, near 1 ?m, though a minimum attributable to pyroxene is still evident near 2 ?m. More evidence of olivine in Manbhoom is consistent with the higher ratios of

olivine to pyroxene in LL chondrites relative to H chondrites (McSween and Bennett 1991). 3.3.2. Carbonaceous chondrites. Warrenton (CO3) and Murchison (CM2) were measured and both show some dependence of their spectra on temperature (Figs. 16 and 17). Because they were dark, signal-to-noise ratios are relatively low. In the case of Warrenton, the presence of the olivine, known to occur

FIG. 14. Temperature series for LL6 ordinary chondrite Manhboom. The top spectrum was collected with the sample at 200 K. Spectra below are for sample temperatures decreasing at 20 K intervals. Each spectrum is offset by .01 re?ectance units from the one above it (top spectrum is not offset). Note that this offset display obscures the fact that some wavelengths are completely inactive with respect to temperature. The corresponding thermo-re?ectance spectra quantify how the spectrum of the sample varies with temperature.

FIG. 16. Temperature series for carbonaceous chondrite Warrenton. The top spectrum was collected with the sample at 400 K. Spectra below are for sample temperatures decreasing at 20 K intervals. Each spectrum is offset by .01 re?ectance units from the one above it (top spectrum is not offset). Note that this offset display obscures the fact that some wavelengths are completely inactive with respect to temperature. The corresponding thermo-re?ectance spectra quantify how the spectrum of the sample varies with temperature (dual measurements at 300 K and 220 K).

176

HINRICHS AND LUCEY

FIG. 17. Temperature series for carbonaceous chondrite Murchison. The top spectrum was collected with the sample at 400 K. Spectra below are for sample temperatures decreasing at 40 K intervals. Each temperature spectrum is offset by .01 re?ectance units from the one above it (top spectrum is not offset). Note that this offset display obscures the fact that some wavelengths are completely inactive with respect to temperature. The corresponding thermo-re?ectance spectra quantify how the spectrum of the sample varies with temperature.

FIG. 19. Temperature series for eucrite EET83251. The top spectrum was collected with the sample at 400 K. Spectra below are for sample temperatures decreasing at 20 K intervals. Each spectrum is offset by .01 re?ectance units from the one above it (top spectrum is not offset). Note that this offset display obscures the fact that some wavelengths are completely inactive with respect to temperature. The corresponding thermo-re?ectance spectra quantify how the spectrum of the sample varies with temperature.

(McSween 1977), is evident by the presence of a peak in the thermo-re?ectance spectrum near 1 ?m (Fig. 18); no evidence of pyroxene is present. Murchison shows some weak structure in this thermo-re?ectance spectrum that may re?ect its mineralogy, but higher quality data are required to con?rm this. 3.3.3. HED meteorites. The HED association meteorites are likely samples from the asteroid Vesta (Binzel and Xu 1993), an asteroid which is a common target for proposed asteroid missions. We measured the eucrite EET83251 and the howardite EET87503 and their temperature series spectra are shown in Figs. 19 and 20. The thermo-re?ectance spectra of both meteorites are shown in Fig. 21. The magnitude of the thermo-

re?ectance spectra (i.e., the sensitivity of the spectra to temperature) is relatively large and is similar to that of the H chondrites. The shapes of the thermo-re?ectance spectra are similar to that of pyroxene; no evidence of olivine is present, consistent with the mineralogy of these meteorites. Because the surface of Vesta appears to be covered with unaltered HED or HED-like material, spatially resolved spectral measurements of Vesta should easily detect this effect as the re?ectance of this material changes by

FIG. 18. Thermo-re?ectance spectra of carbonaceous chondrites Murchison (a) and Warrenton (b). The structure is very weak compared to those of OCs and minerals. Spectrum (a) has been offset 1 unit upward.

FIG. 20. Temperature series for howardite EET87503. The top spectrum was collected with the sample at 400 K. Spectra below are for sample temperatures decreasing at 20 K intervals. Each temperature step is offset by .01 re?ectance units from the one above it (top spectrum is not offset). Note that this offset display obscures the fact that some wavelengths are completely inactive with respect to temperature. The corresponding thermo-re?ectance spectra quantify how the spectrum of the sample varies with temperature (dual measurements at 240 K and 260 K).

TEMPERATURE-DEPENDENT SPECTRAL PROPERTIES

177

TABLE II Measured Lunar Soils a
Sample 10084 12001 12023 12032 14163 60501 67711 72131 72501 73241 75061
a

Is /FeO 78 56 60 12 57 80 2.8 60 81 18 33

FeO 16.2 17.2 15.4 14.1 10.4 5.4 3.0
b

TiO2 7.3 2.8 2.8 2.9 1.8 0.6 0.26
b

8.8 8.5 18.2

1.56 1.73 10.3

FIG. 21. Thermo-re?ectance spectra of EET83251 (a) and howardite EET87503 (b). Spectrum (a) has been offset 1 unit upward.

b

Analysis from Morris et al. 1983. No data.

1% in portions of the spectrum over a 10 K temperature difference. For example, images of Vesta obtained near 1 ?m should show strong brightening toward the terminator if Vesta has a ?ne-grained, low thermal inertia surface. 3.3.4. Meteorite summary. The spectra of the meteorites measured all react to sample temperature (with the exception of Murchison for which we have insuf?cient sensitivity), and in most cases they do so in a way interpretable in terms of the mineralogical composition of the samples. In the 1-?m region, thermo-re?ectance spectra appear to switch from pyroxenedominated, to olivine-dominated (negative to positive feature) at abundances of olivine relative to pyroxene plus olivine between 0.5 and 0.65 judging by the thermo-re?ectance spectra of the H and LL chondrites. The magnitude of the thermo-re?ectance spectra (the total constrast from maxima to minima) is affected by mixing of temperature-active phases (e.g., ordinary chondrites) and dilution of the effect by the presence of inactive phases (carbonaceous chondrites and HEDs). The similarity of the spectra of Vesta to those of powdered HEDs suggests that this temperature effect will be strong on this object and ought to be taken into account in planning for missions to Vesta. 3.4. Lunar Soils The effect of temperature on lunar soils is of obvious relevance to lunar remote sensing, particularly since lunar surface temperatures can vary by more than 100 K on the illuminated surface, and also because lunar soils are the only materials for which space weathering processes are well documented. Pieters et al. (2000) showed that theoretical objections against operation of lunar-like space weathering on asteroids are probably invalid, so the optical properties of lunar soils are relevant to studies of asteroids. We selected 11 lunar soils to span a wide range of soil maturity and composition (Table II). Maturity ranges from immature to mature (Is /FeO 2.8 to 81), FeO content ranges from

3 to 18.2 wt%, and TiO2 content ranges from 0.26 to 10.3 wt%. Overall, the effect of temperature on spectral properties of these soils is minute, though measurable, and all spectral series show some inverse correlation of temperature and re?ectance. However, the effect is so weak that no structure can be discerned for most soils by inspection of spectra overplotted as is so evident in most of the meteorite spectra shown here. Two examples of temperature series spectra of lunar soils can be seen in Figs. 22 and 23, which are based on measurements of lunar soils 67711 and 12023 respectively. We show the derived thermo-re?ectance spectra of all 11 lunar soils in Fig. 24. Some of these thermore?ectance spectra show structure consistent with the presence of pyroxene or olivine (see Section 4), though with the exception

FIG. 22. Temperature series for lunar soil 67711. The top spectrum was collected with the sample at 400 K. Spectra below are for sample temperatures decreasing at 20 K intervals. Each temperature step is offset by .01 re?ectance units from the one above it (top spectrum is not offset). Note that this offset display obscures the fact that some wavelengths are completely inactive with respect to temperature. The corresponding thermo-re?ectance spectra quantify how the spectrum of the sample varies with temperature (dual measurements at 300 K).

178

HINRICHS AND LUCEY

FIG. 23. Temperature series for lunar soil 12023. The top spectrum was collected with the sample at 300 K. Spectra below are for sample temperatures decreasing at 20 K intervals. Each temperature step is offset by .01 re?ectance units from the one above it (top spectrum is not offset). Note that this offset display obscures the fact that some wavelengths are completely inactive with respect to temperature. The corresponding thermo-re?ectance spectra quantify how the spectrum of the sample varies with temperature (dual measurements at 280 K).

are good to ?rst order, indicating that most of the structure in the thermo-re?ectance spectra of the rocks can be accounted for using mixtures of the two minerals. The differences are likely due to differences between the mineral compositions in the rocks and those used for the model mixtures. This shows that thermo-re?ectance spectroscopy contains interpretable information regarding at least the olivine and orthopyroxene mineralogy of rock containing these minerals. It is noteworthy that the very plagioclase-rich lunar soil is well matched by an olivinepyroxene mixture. We have not measured the thermo-re?ectance spectrum of plagioclase, though the data of Roush (1984) suggest that the spectrum of this mineral is relatively inactive with respect to temperature. The thermo-re?ectance spectrum of this soil, which probably contains more than 50% plagioclase based on its very low iron content (3 wt% FeO), shows no effect of this component unless a small feature near 1.1 ?m is due to plagioclase. The presence of an abundant neutral component

of the spectrum of the bright soil 67711 most spectra are noisy. The maximum change in relative re?ectance with temperature is on the order of 1% or less per 100 K, compared to about 5–10% for the ordinary chondrites and HEDs and up to 35% for pure minerals. The weakness of the thermo-re?ectance effects results from two aspects of lunar soils. First, the more mature soils have inherently low contrast owing to the masking effect of abundant submicroscopic iron (Hapke 2000). The relatively subtle changes discussed here are greatly attenuated by this material. Second, many of the soils are low in iron, which is responsible for the temperature-dependent absorptions. Finally, these soils are dark and the sensitivity of our experiment is relatively low for dark soils. These results indicate that temperature effects can be safely ignored for standard spectral analysis of lunar surface materials, though extremely fresh ma?c material, which is not represented in our suite of samples, would probably show temperature behavior similar to that of the ordinary chondrites.
4. COMPOSITIONAL CONTENT OF THERMO-REFLECTANCE SPECTRA

Many of the thermo-re?ectance spectra presented have signi?cant structure, much of which can be explained to ?rst order by the thermo-re?ectance spectral properties of olivine and pyroxene. To illustrate this we compare thermo-re?ectance spectra of several samples to appropriately scaled additive mixtures of thermo-re?ectance spectra of olivine and pyroxene. Figures 25 and 26 show thermo-re?ectance spectra of a lunar soil and several meteorites with additive mixtures of olivine and pyroxene thermo-re?ectance spectra. In all cases the matches

FIG. 24. Thermo-re?ectance spectra of lunar soils. From top to bottom spectra are in order of decreasing re?ectance. The magnitudes of these thermore?ectance spectra (peak to peak variation) are very small, on the order of 1% change in relative re?ectance per 100 K. The structures just short of 1 ?m are artifacts due to low signal-to-noise ratios.

TEMPERATURE-DEPENDENT SPECTRAL PROPERTIES

179

5. CONCLUSIONS

FIG. 25. Thermo-re?ectance spectra of the CO chondrite Warrenton and lunar soil 67711 (crosses) with additive mixtures of the thermo-re?ectance spectra of Fo86 olivine and En86 orthopyroxene (solid lines). The lunar soil is very well represented by this mixture with the exception of a narrow in?ection of the soil spectrum near 1.1 ?m. This small in?ection might be due to plagioclase, which is abundant in this soil, or another component. The thermo-re?ectance spectrum of the CO carbonaceous chondrite is mismatched at the shortest and longest wavelengths. Near 0.5 ?m the mismatch is due to the presence of ferric oxide caused by terrestrial weathering in the meteorite. Ferric oxide spectral properties are also temperature sensitive (Morris et al. 1997). The mismatch at longer wavelengths is unexplained.

is indicated by the weakness of the thermo-re?ectance signal, which is similar in strength to mature soils. The neutral component dilutes the strength of the temperature sensitivity signal.

FIG. 26. Thermo-re?ectance spectra of H5 ordinary chondrites El Hammammi (a), Allegan (b), and eucrite EET83251 (c) (crosses) and additive mixtures of the thermo-re?ectance spectra of forsteritic olivine (Fo86 ) and orthopyroxene (En86 ) (lines). There is general qualitative agreement in all cases. The mismatch near 0.5 ?m in El Hammammi is due to temperature-sensitive ferric oxide; other differences are probably due to differences between the compositions of the olivine and pyroxene used in the model and those in the meteorites.

The temperature-dependent spectral properties of minerals are more than a complication superimposed on analysis of re?ectance spectra of planetary materials; they contain compositional information that is complementary to the re?ectance spectra. We coined the term “thermo-re?ectance” to denote the quantity R/ T and “thermo-re?ectance spectra” for the wavelength dependence of this quantity. The thermo-re?ectance spectra of olivine- and pyroxene are substantially different, and we showed that for those olivine- and pyroxene-bearing samples which exhibit structure in their thermo-re?ectance spectra, the structure is consistent with a mixture of the thermo-re?ectance properties of olivine and pyroxene. The magnitude of the thermo-re?ectance of lunar soils is very small and indicates that in most cases this effect can be ignored in conventional analysis, although certain lunar locations might express this effect more strongly. This indicates that space weathering effectively masks the temperature effect as it does for the more familiar spectral properties. Compositional interpretation of thermo-re?ectance spectra does not require absolute calibration of data for derivation as the method relies upon detecting changes in the spectrum of a surface element or sample with temperature. Further, the temperature-dependent spectral properties of a remotely sensed surface directly couples compositional analysis with measurement of the temperature of a surface, for example with a radiometer. Radiometric estimates of temperature and temperature variation must be consistent with the observed spectral variation with temperature. This requirement may prove to be a powerful constraint on asteroid surface properties. The temperature-dependent spectral properties of HED meteorites suggest that the asteroid Vesta might exhibit temperaturedependent spectral effects strongly. An important issue to raise is that photometric analysis of the surface of Vesta must include temperature-dependent terms. For example, if Vesta has a low thermal inertia surface, at 1 ?m the re?ectance of the asteroid may drop by as much as 10% from subsolar point to terminator. At present, photometric models do not contain this dependence. Measurement of temperature-dependent spectral properties for powders in the laboratory must account for the radiative environment of the sample, speci?cally the temperature difference between background and sample (relative to conductivity) and radiative heating of the sample by the light source. Qualitatively, the trends measured by Roush and Singer are consistent those measured by us for the same samples, indicating that radiative effects did not dominate their experiments. This suggests that the Roush and Singer measurements of the spectral behavior with temperature of minerals not measured by us (plagiclase and clinopyroxene) are probably at least qualitatively correct, although the absence of documentation of the radiative environment leaves a residual uncertainty. Our experiments were inhibited by the intensity of our light source. Use of a spectrally

180

HINRICHS AND LUCEY dependent spectral variation on the Asteroid Eros and new evidence for the presence of an olivine-rich silicate assemblage. Icarus 155, 181–188. McSween, H. Y. 1977. Carbonaceous chondrites of the Ornans type: A metamorphic sequence. Geochem Cosmochem. Acta 41, 477– 491. McSween, H. Y., and M. E. Bennett III 1991. The mineralogy of ordinary chondrites and implications for asteroid spectrophotometry. Icarus 90, 107–116. Morris, R. V., R. A. Score, C. Dardano, and G. Heiken 1983. Handbook of Lunar Soils, Parts 1 and 2. NASA Planetary Materials Branch Publication 67, JSC Publ. 19069. Morris, R. V., D. C. Golden, and J. F. Bell III 1997. Low-temperature re?ectivity spectra of red hematite and the color of Mars. J. Geophys. Res. 102, 9125– 9134. Moroz, L., U. Schade, and R. W¨ sch 2000. Re?ectance spectra of olivinea orthopyroxene-bearing assemblages at decreased temperatures: Implications for remote sensing of asteroids. Icarus 147, 79–93. Pieters, C. M., L. A. Taylor, S. K. Noble, L. P. Keller, B. Hapke, R. V. Morris, C. Allen, D. S. McKay, and S. Wentworth 2000. Space weathering on asteroids: Resolving a mystery with lunar samples. Meteorit. Planet. Sci. 35, 1101–1107. Roush, T. L. 1984. Effects of Temperature on Remotely Sensed Ma?c Mineral Absorption Features. Masters thesis, University of Hawaii. Roush, T. L., and R. B. Singer 1986. Gaussian analysis of temperature effects on the re?ectance spectra of ma?c minerals in the 1-um region. J. Geophys. Res. 91(B10), 10301–10308. Roush, T. L., and R. B. Singer 1987. Possible temperature variation effects on the interpretation of spatially resolved re?ectance observations of asteroid surfaces. Icarus 69, 571–574. Schade, U., and R. Wasch 1999. NIR re?ectance spectroscopy of ma?c minerals in the temperature region between 80 and 473 K. Adv. Space Res. 23, 1253– 1256. Singer, R. B. 1981. Near-infrared spectral re?ectance of mineral mixtures: Systematic combinations of pyroxenes, olivines, and iron oxides. J. Geophys. Res. 86, 7967–7982. Singer, R. B., and T. L. Roush 1985. Effects of temperature on remotely sensed mineral absorption features. J. Geophys. Res. 90(B14), 12434–12444.

scanned monochromatic light source with high in-band but low total irradiance might be an effective approach to measurement of very low temperature spectral properties. Finally, use of this technique for compositional analysis is currently limited by the small number of materials for which properties have been measured. In particular pyroxenes and olivines of different compositions, plagioclase, and glasses would be important to characterize.
ACKNOWLEDGMENTS
The authors thank Jack Mustard and an anonymous reviewer for their thoughtful comments and Anders Meibom and Tom Burbine for the meteorite samples. This research was supported in part by NASA grants NAG5-8891 and NAG53766.

REFERENCES
Binzel, R. P., and S. Xu 1993. Chips off of asteroid 4 Vesta: Evidence for the parent body of basaltic achondrite meteorites. Science 260, 186–191. Hapke, B. 2001. Space weathering from Mercury to the asteroid belt. J. Geophys. Res. 106, 10,039–10,074. Henderson, B. G., and B. M. Jakosky 1994. Near-surface thermal gradients and their effects on mid-infrared emission spectra of planetary surfaces. J. Geophys. Res. 99, 19063–19073. Hinrichs, J. L., P. G. Lucey, M. S. Robinson, A. Meibom, and A. N. Krot 1999. Implications of temperature-dependent Near-IR spectral properties of common minerals and meteorites for remote sensing of asteroids. Geophys. Res. Lett. 26, 1661–1664. Lucey, P. G., K. Keil, and R. Whiteley 1998. The in?uence of temperature on the spectra of the A-asteroids and implications for their silicate chemistry. J. Geophys. Res. 103, 5865–5871. Lucey, P. G., J. Hinrichs, M. Kelly, D. Wellnitz, N. Izenberg, S. Murchie, M. Robinson, B. E. Clark, and J. F. Bell III 2002. Detection of temperature-


赞助商链接

更多相关文章:
更多相关标签:

All rights reserved Powered by 甜梦文库 9512.net

copyright ©right 2010-2021。
甜梦文库内容来自网络,如有侵犯请联系客服。zhit325@126.com|网站地图