Astron. Astrophys. 362, 1072-1076 (2000)

4. Interpretation

The findings presented above could be interpreted in terms of both surface and atmospheric features. Lack of contrast at short wavelengths has been attributed to the spectral properties of iron oxides (Huguenin et al. 1977). This can explain the lowering of contrast of Martian features observed in the northern hemisphere (Vastitas and Acidalia). Viking images have shown the occurrence of albedo reversal at regional scales ( 300 Km) and it has been attributed to eolian deposits (Thomas & Veverka 1986). The classical contrast reversal phenomenon occurs among the Sinus Meridiani and Vastitas-Acidalia, with crossover appearing at 0.485 µm and Syrtis Major with Arabia, with crossover at 0.425 µm. The amount of reflectance increase at short wavelengths and position of crossover suggest that we are looking to some clouds or haze. Mars was observed during this opposition by the HST Wide field planetary camera. In Fig. 3, two color composite WFPC2 images taken on 10 and 30 March are shown. They are composed of individual red (673 nm), green (502 nm) and blue (410 nm) exposures. In the figure, a color composite at the same wavelengths obtained from our data is also shown. It has been stretched in order to have approximately the same color tint. It must be noted that the bandpass of the red and green WFPC2 camera filters are 5 and 3 nm, approximately the same of our spectrometer channels (5 nm). The blue filter has = 14 nm; for this reason we have taken an average of the 405, 410, 415 channels to generate the blue component in the Fig. 3 picture. On the WFPC2 images, a diffuse water ice haze is visible on the equatorial region. On the 30 March image it is more visible, probably due to a slightly different color stretch. Our image, taken on 20 March, shows a more prominent haze layer, which covers almost completely the Syrtis Major and Elysium regions while it is more diffused on Arabia. We have considered the limb cloud located approximately on Elysium as representative of the clouds and hazes visible on the blue images. The spectrum is an average of 44 pixels. In order to obtain the surface term, we sampled a spectrum at the same longitude of Elysium but just out of the cloud. We can assume that the surface close to the haze covered terrains has the same spectral response of the haze-underlying soils. The rationale of this approximation is due to the spectral homogeneity of bright regions, at least at the spatial scales involved in ground based observations. Fig. 4 shows the spectral dependence of brightening obtained by subtracting the surface spectrum to the cloud spectrum. The spectral brightening shown in Fig. 4 tends to be flat below 0.5 µm, with perhaps a small peak at 0.45 µm. We have modeled the observed spectral behavior in the 0.4-0.7 µm domain by using a discrete ordinates radiative transfer code (Disort, Stamnes et al. 1988). The martian atmosphere has been subdivided in two layers (0-10 Km, 10-20 Km) which take into account the vertical distributions of dust, water clouds and CO₂ Rayleigh scattering, by taking a surface pressure of 7 mbar. Following Clancy et al. 1996a, we assume 60% of the cloud opacity occurs in 0-10 Km layer and the remaining 40% in the 10 -20 Km layer. The condensation level of Mars water vapor is specified by the aphelion atmospheric temperature profile (Clancy et al. 1996b). The ozone absorption occurring above 20 Km altitude has been negletted because it affects only wavelengths shorter than 0.3 µm. Cloud and dust single scattering phase functions are taken from results of Clancy & Lee 1991. The dust single scattering albedo is adopted from Wolf et al. 1999, while for the clouds it is fixed to 1. The cloud and dust opacities are treated as wavelenght independent parameters and varied to achieve a consistent macth to the observed spectra. From previous studies this appear to be a reasonable approximation (Clancy et al. 1995; Smith et al. 1997; Wolff et al. 1999). Fig. 5 shows the results. The best fit to the data points has been obtained with a cloud opacity = 0.10 and dust opacity = 0. The residual is below 3% in the all wavelength range. Anyway, a model compatible with the data error bar is also obtained, by taking = 0.10 and = 0.15, through with a poorer fit. As shown in Fig. 5b, the residual is now larger, specially in the 0.5 0.7 µm domain. The case of a "pure" dusty atmosphere is shown in Fig. 5c. The figure shows how it is necessary to include some ice opacity to decrease the overall error fit. In summary, even though modelling pushes toward a dust free atmosphere, an higher dust opacity is not excluded. Wolff et al. 1999 report a diffuse dust opacity value = 0.3, measured at the end of March. Recently, the role of water ice clouds on martian climate has been revaluated. Clancy et al. 1996b showed the occurrence of low altitude (10 km) water vapor saturation around several Mars aphelions. During these periods the Mars atmosphere was 15-20 K colder than observed during the Viking mission. At these temperatures, water ice clouds form at low altitude, covering the 10^oS - 30^oN latitude region (James et al. 1994, Clancy et al. 1996b, Wolff et al. 1997). Temperatures 40 K lower than Viking mission were also reported by Pathfinder (Magalhães et al. 1999). An inversion at about 10 Km has been also observed which can lead to the formation of low-altitude clouds (Colaprete et al. 1999). In the 1997 opposition Mars was close to the aphelion and the same scenario probably occurred. We then suggest that the albedo reversal observed in our spectra and relative to several region of Mars, mainly located in the equatorial belt, is due to the scattering properties of low altitude water ice clouds. If the clouds condense around dust as nucleation centers, fine dust particles can be transported from one region to another, contributing to albedo variation of surface markings. Instances of albedo reversal reported by ground-based observers in the past could also be explained by the presence of low altitude water ice clouds (McCord 1969). On the other hand, occurrence of this phenomenon was reported also when Mars was much closer to the Sun (Thompson 1973). In this case, water ice clouds forming at higher altitudes could be invoked to explain the observations. Generally, the albedo features involved are located at equatorial-tropical latitudes. There is no clear evidence of a seasonal dependence of the reversal, mainly due to the sparse observations (Martin et al. 1992).

Fig. 3. Red, Green, Blue composite images of Mars during March 1997 (see text). They show limb clouds and an equatorial haze of H2O ice. The left and the right images were taken by HST on 10 and 30 March, respectively. The image in the center was taken on 20 March by the author and it is discussed in the text.

Fig. 4. Spectral dependence of brightening observed on Elysium. It has been obtained by subtracting a surface term to the cloud spectrum.

Fig. 5. Results of radiative transfer modeling of Elysium cloud. They have been obtained by using different dust and cloud opacities. See the text for details

© European Southern Observatory (ESO) 2000

Online publication: October 30, 2000
helpdesk.link@springer.de