Neptune’s Cloud Structure and Activity: Ground-based Monitoring with Adaptive Optics


Since August 1995, near-infrared images of Neptune have regularly been obtained with the University-of-Hawaii adaptive optics system mounted on the f/35 Cassegrain focus of the Canada-France-Hawaii telescope. These images reveal Neptune’s cloud structure with an angular resolution reaching 0.12” in the H band. Using adaptive optics (AO), long-term monitoring of Neptune’s cloud activity is now possible from the ground with an angular resolution close to the telescope diffraction limit. Since our first observation of Neptune in August 1995 (Roddier et al. 1997), we have regularly used the University-of-Hawaii (UH) AO system and HgCdTe infrared camera to observe Neptune at the f/35 infrared Cassegrain focus of the CFHT. Let us recall that the UH AO system differs from other AO systems developed for defense applications by its wave-front sensing and compensation technique (Roddier 1988, Roddier et al. 1991). Wave-front sensing is done with an array of photon-counting avalanche photodiodes. Neptune itself is used as a guide source. First operated in 1994, the 13actuator UH AO system has now been upgraded to 36 actuators. Most of the images presented here were obtained with the new system. It should be noted that a 19-actuator AO system of the same kind has been built by the Canada-France-Hawaii Telescope (CFHT) corporation and has been offered to its users since 1996. The images presented here were obtained from data taken in October 1996, July 1997, and November 1997. Observations were made in the near infrared using standard wide-band J, H, and K’ filters, and two narrow-band filters: one centered at 1.72 μm (inside the methane absorption bands) and one centered at 1.56 μm (outside the methane absorption bands). As we shall see, differences in the opacity of Neptune’s atmosphere inside and outside of the methane absorption bands can be used to isolate high altitude atmospheric layers from lower ones. Figure 1 shows twelve images of Neptune (numbered from 1 to 12) as a representative sample of the data we obtained with our adaptive optics system. They have been only background-subtracted and flat-fielded. No deconvolution was applied. The display is linear. The images are ordered by date increasing from left to right and by wavelength increasing from top to bottom. To save space, a July 1997 image has been put in the same column as a Nov. 1997 image. Exposure times were all limited to 300 s to avoid image blurring caused by the rotation of Neptune. The Oct. 1996 and July 1997 images were obtained with the 13-channel AO system and are of poorer optical quality, especially the July images which were obtained under poor seeing conditions. The Nov. 1997 ones were obtained with the new 36-channel AO system and are practically diffraction limited (for information on the performance of AO systems, see Roddier 1998). The angular resolution of these images was estimated on field stars. For the best H band images, the point spread function (PSF) has a full width at half maximum (FWHM) of 0.12 arcsec. This corresponds to 2500 km on Neptune’s visible cloud tops at disk center. The sampling rate is 0.035 arcsec/pixel. The Neptune’s radius and orientation was taken from the JPL Horizon online program accessible by Telnet ( 6775). Neptune’s radius was taken to be 2.26 arcsec in Oct. 1996, 2.34 arcsec in July 1997, and 2.24 arcsec in Nov. 1997. The inclination and the position angle of Neptune’s north pole were respectively taken to be 25.8 ̊ and 1.9 ̊ in Oct. 1996, 26.5 ̊ and 358.9 ̊ in July 1997, 26.2 ̊ and 359.9 ̊ in Nov. 1997. When trying to locate surface features, a large uncertainty comes from the determination of the disk center. To minimize this uncertainty, we have computed how the clouds would appear when the observer faces Neptune’s south pole. Assuming that the cloud bands are parallel to the equator, the disk center was chosen so that they appear best as concentric rings around the south pole. The transformed images were empirically rotated until common features coincide. An example is given at the bottom of Fig. 2, where images (5), (7), and (11) of Fig. 1 have been transformed, rotated, and added. We believe this method has allowed us to determine the image center with an accuracy better than one pixel or 0.035 arcsec. On all the images presented here, a thin line was used to mark the following latitudes: 60 ̊ N, 30 ̊ N, 0 ̊, 30 ̊ S, and 60 ̊ S. The uncertainty on the latitudes is ± 3 ̊ for the data taken with the old AO system and ± 2 ̊ for the data taken with the new system. The accuracy is currently limited by the uncertainty on the pixel size. In the future a more accurate calibration method will be used allowing latitudes to be estimated with an uncertainty of ± 1 ̊. On some images of Fig. 1 two meridians have been drawn pointing to particular cloud features. They have been rotated with a rotation period of 17 h matching that of the cloud bands at the latitude of 43 ̊ S (see below). All the images in Fig. 1 show bright isolated clouds. They have a maximum contrast in the methane absorption bands around 1.72 μm (4th row), where they appear in front of a dark highly absorbing atmosphere. It implies they are high-altitude, possibly stratospheric clouds (Sromovsky et al 1995, p. 28) A precise determination of their altitude would require both accurate photometry and a good estimate of the methane mixing ratio which is beyond the scope of this short note. According to extinction curves from Baines et al. (1995), they may form at a pressure of the order of 0.1 bar or at an even higher altitude. On all our images, we have estimated the latitude of the cloud bands. For the Oct. 1996 images, the cloud latitudes were found to be 48 ̊ S, 35 ̊ S, and 26 ̊ N, that is within the error bars the same as that of our Aug. 1995 observations (Roddier et al. 1997). HST observations made in March 1996 (Hammel and Lockwood 1996 and 1997) and in Aug. 1996 (Sromovsky et al. 1996) also show activity at these latitudes. By contrast, images taken in July 1997 reveal cloud activity at 45 ̊ S, 27 ̊ S, and 40 ̊ N, and images taken in Nov. 1997 at 68 ̊ S, 43 ̊ S, 27 ̊ S, and 40 ̊ N. Therefore, a change in the latitude location of Neptune’s cloud activity may have possibly occurred between Oct. 1996 and July 1997. In the southern hemisphere cloud activity now appears over a wider range of latitudes. Also the cloud band at 26 ̊ N has apparently disappeared and a band has appeared at 40 ̊ N. Cloud activity at 40 ̊ N was also seen by Voyager in 1989 and by HST in 1991 (Sromovsky et al. 1995). According to Hammel and Lockwood (1996 and 1997) northern cloud activity peaked around 1994 and has decayed since (see also Sromovsky et al. 1996). We have made rough estimates of the respective contribution of the north and south hemisphere to the overall cloud brightness integrated over each of our data sets. The ratio of these contributions (north/south) was found to be 2.9 in Aug. 1995 (1 night), 2.5 in Oct. 1996 (2 nights), 0.6 in July 1997 (1 night), and 0.7 in Nov. 1997 (3 nights). Hence, our observations appear to confirm the decrease in the northern cloud activity observed by others. Neptune’s cloud activity may have switched back from north to south toward a configuration more similar to that observed by Voyager 2 in 1989. If confirmed, this result might support the conjecture that Neptune’s cloud activity has a long term component related to the solar cycle (Lockwood and Thompson 1986, Roddier et al. 1997) From observations of the high altitude clouds made a few hours or a few days apart we have attempted to estimate the rotation period of the observed clouds. On Oct. 26, 1996, the bright spot seen on image (3) at 26 ̊ N was observed in K’ over a 1 h interval. Its rotation period was found to be 18.2 ± 0.5 h. Although there is a large uncertainty due to the short time interval, this value is consistent with the rotation periods observed by Voyager at the same latitude (Sromovsky et al. 1993). On Nov. 10, 1997, image (5) shows two cloud spots at a latitude of 43 ̊ S. The same spots are also seen two days later on image (11). For clarity, both images show a pair of widely spaced meridians (35 ̊ apart) drawn between the two spots. Assuming no drift or other evolution of the clouds occurred during the three revolution interval, the rotation period was found to be 17.0 ± 0.1 h, in perfect agreement with Voyager observations at the same latitude (Sromovsky et al. 1993). The clouds we have tracked are probably the same type of clouds that have been tracked on the Voyager images. However, the fact that they appear within the methane absorption bands implies that they are high in the atmosphere, certainly higher than the dark spots. Evidence of this can also be found in the shadows cast by these clouds as seen on some Voyager images. Since both type of features have about the same rotation velocities, Neptune’s atmospheric motion must be relatively independent of the altitude, at least up to 0.1 bar above which the wind speed may start to decrease (French et al. 1996). An interesting cloud feature can be seen on image (11) taken on Nov. 12, 1997. The feature appears near the west limb at a high south latitude (68 ̊ S). However, no such cloud is seen earlier. Given its latitude one would indeed expect a rotation period of 13 h. That would bring it very close to the south limb on earlier images, where it may be difficult to see. However, the cloud feature may well have formed in less than 24 h. As pointed out by one referee, this cloud feature is probably what has been described as a south polar feature (SPF) by Sromovsky et al. (1993). SPFs are rapidly evolving and rapidly moving features which have been observed on Voyager images at latitudes between 68 ̊ S and 74 ̊ S (Limaye and Sromovsky 1991). We now discuss differences between images taken at different wavelengths. Images taken inside the methane bands, such as image (11) taken at 1.72 μm, and those taken outside, such as image (10) taken at 1.56 μm, show the most striking differences. Image (11) shows bright, high contrast cloud features on top of a dark background. Image (10) shows similar features in the southern hemisphere but with a lower contrast. The same effect can be seen in the HST images taken inside and outside of the 0.89 μm methane bands. This effect has been interpreted as due to a difference in the altitude at which light scattering is seen to occur (Hammel and Lockwood 1997). Because outside the methane bands Neptune’s atmosphere is more transparent one sees deeper, and new atmospheric layers appear. Both image (10) and image (11) show the same isolated highaltitude clouds, but in addition image (10) also shows lower atmospheric layers. These are better seen by subtracting image (11) from image (10). A slight correction has been applied to one of the images to account for the planet’s rotation, as described in Roddier et al. (1997). This was done to image (10). When subtracting the two images a relative weight was applied and empirically determined to make the high altitude clouds best disappear, leaving only the low altitude ones. The result of the subtraction is displayed with two different intensity levels on image (f) and (f’) of Fig. 2. Image (f) shows that the high altitude clouds have indeed fairly uniformly disappeared, which indicates they are mostly optically thin. Image (f’) displays lower intensity levels, and shows that most of Neptune’s south hemisphere (up to at least 30 ̊ S) is covered with haze. Photometric profiles taken along meridians show a maximum haze near 70 ̊ S from which the amount of scattered light decreases almost linearly from south to north without any sharp discontinuity. The same southern haze can be seen on image (a) (Fig. 2) obtained by subtracting image (2) (1.72 μm) from image (1) (1.56 μm), and on image (d) obtained by subtracting image (8) (1.72 μm) from image (7) (H band). In addition, a localized spot of haze can also be seen in the northern hemisphere where bright high altitude clouds appear, i.e., 26 ̊ N on image (a) and 40 ̊ N on image (b). This could be due to an incomplete removal of the high latitude northern clouds which are brighter than the southern ones on which the weights have been determined. Neptune’s atmosphere is also relatively transparent in the J band so that both image (9) taken in the J band and image (10) taken at 1.56 μm show deep layers and look similar. One can again remove the contribution of the high altitude clouds by subtracting image (11) (1. 72 μm) from image (9). The resulting image (e) indeed shows the low altitude haze with a maximum brightness in the south hemisphere. However, comparison with image (f’) shows more clouds near the west limb. In the same way, image (c) shows the result of subtracting image (8) from image (6). Note that the haze now extends all the way from the southern hemisphere to the north spot, as also seen on HST images (Hammel et al. 1995). Hence the cloud layers seen on images (c) and (e) differ somewhat from those seen on images (a), (d), and (f’). The difference may be a difference in altitude or in chemical composition. In the K’ band, Neptune’s atmosphere is much less transparent (See Fig. 3 of Baines et al. 1995). One can see features which are much less deep than at 1.56 μm, but still somewhat deeper than at 1.72 μm. One can indeed see differences between image (11) taken at 1.72 μm and image (12) taken in the K’ band. Compared to the former, clouds in the latter have less contrast, and northern clouds are brighter compared to southern ones. This is clearly seen in image (g) obtained by subtracting image (11) from image (12). We have also subtracted the cloud features seen on image (2) from those observed on an image similar to (3) but taken at about the same time (6:02). The result is shown on image (b). Both (b) and (g) show similar structures, which may be explained by differences in altitude or differences in chemical composition. In this case, differences due to an inadequate weighting of the images can be excluded, since the relative brightness of the northern and southern hemisphere is opposite in 1996 to that in 1997. In conclusion, the images reveal a complex cloud structure. Low altitude hazes seem to cover most of the southern hemisphere with possible isolated areas over northern active regions. Thin high-altitude clouds form over the haze in the south hemisphere and at latitudes where activity occurs in the north. A more quantitative photometric analysis of such images coupled to a good model of Neptune’s atmosphere should clearly enable us to better distinguish the various layers, determine their altitude and perhaps their chemical composition. This research is supported by NASA grant NAG5-3731. The UH AO systems were built under NSF grants AST 93-19004 and AST 96-18852. We thank C. Dumas and L. Close for helping us acquire the data, and H. Hammel and an unknown referee for their numerous comments. REFERENCES BAINES, K. H., H. B. HAMMEL, K. A. RAGES, P. N. ROMANI, and R. E. SAMUELSON 1995. Clouds and hazes in the atmosphere of Neptune. In Neptune and Triton (D. P. Cruikshank, Ed.) pp. 489546. Univ. of Arizona Press, Tucson. FRENCH, R. G., C. A. MCGHEE and B. SICARDY 1996. Neptune’s stratospheric winds from three central flash occultations. Bull. Am. Astr. Soc.28, 1078. HAMMEL, H. B., LOCKWOOD, G. W., MILLS, J. R., and BARNET, C. D. 1995. Hubble Space Telescope Imaging of Neptune’s cloud structure in 1994 Science 268, 1740-1742. HAMMEL, H. B. and G. W. LOCKWOOD 1996. HST imaging of Neptune in 1994-1996 at multiple wavelengths. Bull. Am. Astron. Soc.28, 1077. HAMMEL, H. B. and G. W. LOCKWOOD 1997. Atmospheric Structure of Neptune in 1994, 1995, and 1996: HST imaging at multiple wavelengths. Icarus 129, 466-481. LIMAYE, S. S. and SROMOVSKY, L. A. 1991. Winds of Neptune: Voyager observations of cloud motions. J. Geophys. Res. 96, 18941-18960. LOCKWOOD, G. W., and THOMPSON, D. T. 1986. Long-term brightness variations of Neptune and the solar cycle modulation of its albedo. Science 234, 1543-1545. RODDIER, F. 1988. Curvature sensing and compensation: a new concept in adaptive optics. Appl. Optics 27, 1223-1225. RODDIER, F., M. J. NORTHCOTT, and J. E., GRAVES 1991. A simple low order adaptive optics system for near-infrared applications. Pub. Astron. Soc. Pac. 103, 131-149. RODDIER, F., C. RODDIER, A. BRAHIC, C. DUMAS, J. E. GRAVES, M. J. NORTHCOTT, and T. OWEN 1997. First ground-based adaptive optics observations of Neptune and Proteus. Planet. Space Sci. 45, 1031-1036. RODDIER, F. 1998. Maximum gain and efficiency of adaptive optics systems. Pub. Astron. Soc. Pac. (to appear in the July 98 issue). SROMOVSKY, L. A., LIMAYE, S. S., AND FRY, P. M. 1993. Dynamics of Neptune’s major cloud features. Icarus 105, 110-141. SROMOVSKY, L. A., LIMAYE, S. S., AND FRY, P. M. 1995. Clouds and circulation on Neptune: Implications of 1991 HST observations. Icarus 118, 25-38. SROMOVSKY, L. A., S. S. LIMAYE, K. H. BAINES, G. S. ORTON and A. P. INGERSOLL 1996. Coordinated HST and IRTF observations of Neptune’s cloud structure and circulation in 1996. Bull. Am. Astron. Soc 28, 1077 Figure captions: Fig. 1. Adaptive optics images of Neptune. Each column corresponds to a different date and each row to a different wavelength. To save space column (2) shows data taken at two different dates. The dates (UT) of the observations are indicated on top and bottom. The time (UT) is indicated under each image. The filters are indicated on the left. These are: a standard J-band filter, a 120-nm bandwidth filter centered at 1.56 μm, a standard H-band filter, a 120-nm bandwidth filter centered on the methane absorption bands at 1.72 μm, and a so-called K’ filter with a 1.94-2.29 μm band pass. The time of mid-exposure is indicated below each image. On each image thin lines indicate Neptune’s equator together with latitudes of 30 ̊ N, 30 ̊ S, and 60 ̊ S. On two images a pair of meridian is shown with a separation of 35 ̊, and a rotation period of 17 h. Fig. 2. Except for the last row, these images display brightness differences. Several of them display both positive and negative values and have a gray background corresponding to a value of zero. They show Neptune’s brightness distribution after subtraction of the brightness distribution in the corresponding 1.72 μm image as indicated on the left. Dates (UT) are indicated on the top. On these images, thin lines indicate the latitudes at which high altitude cloud bands are observed (see text). The last row shows a combination of images (5), (7), and (11) after transformation to show Nep10/26/96 11/10/97 11/11/97 11/12/97 10/26/96 7/17/97 11/10/97 11/11/97 J band

Cite this paper

@inproceedings{Roddier1998NeptunesCS, title={Neptune’s Cloud Structure and Activity: Ground-based Monitoring with Adaptive Optics}, author={François Roddier and Claude Roddier and John E. Graves and Malcolm Northcott}, year={1998} }