Dark spots are short-lived vortices in Neptune's atmosphere that appear every few years but only last a few months before disappearing. The first dark spot discovered was the 'Great Dark Spot', seen by Voyager 2 in 1989, which had a size similar to that of Jupiter's Great Dark Spot, but which disappeared after the Voyager 2 flyby. Dark spots have since  been seen occasionally by the Hubble Space Telescope, but only in a few broad-wavelength channels and so their vertical level and darkening mechanism has remained poorly constrained. To better determine their nature we need observations with a spectrometer, able to measure their continuous reflectance spectrum at medium resolution, but this has never before been achieved. In addition, because of Neptune's vast distance from the Earth the only telescope able to spatially resolve such dark spots has been, to date, the Hubble Space Telescope, and dark spots have never before been observed with ground-based telescopes.

In 2018 a new dark spot was discovered by HST using its WFC3 instrument, called NDS-20128 ('Northern Dark Spot - 2018').  As soon as I heard of NDS-2018's discovery,  I applied for time (along with several colleagues) to try and observe it with the Multi Unit Spectroscopic Explorer (MUSE) instrument at the European Southern Observatory's Very Large Telescope in Chile. MUSE is an Integral Field Unit spectrometer, where the received light over the instrument's field of view is split off to several CCDs to allow spectroscopic imaging. The resulting data product are spectral 'cubes' of data, where each pixel of the reconstructed 300x300 pixel images is a complete spectrum covering 475-933 nm at over 3000 wavelengths, with a spectral resolving power of ~3000.  In its Narrow Field Mode, the MUSE field of view is 7.5 x 7.5 arcsec, which compares well with Neptune's disc diameter of ~2.4 arcsec. In addition, MUSE is equipped with adaptive optics, which corrects in real time for atmospheric turbulence and which can achieve a spatial resolution of ~0.1 arcsec, using a Laser from 578-605 nm to create a synthetic guide star. Given the observed size of the NDS-2018 dark spot, we hoped this spatial resolution would be enough to resolve the dark spot and thus recover its spectrum, giving us not only the first ever detection of a Neptunian dark spectrum from the ground, but also uniquely recovering its reflectance spectrum.

Our application for time was accepted and observations of Neptune were made on October 17th and 18th, and November 13th, 2019. Five observations within this set were predicted to include the NDS-2018 feature, but initial analysis of the data was disappointing. Although the MUSE AO system achieves remarkable spatial resolution, the performance of the system is best at longer wavelengths, but deteriorates at shorter wavelengths, where dark spots are visible. Hence, although we thought we could JUST pick out the spot in some of our MUSE cubes, the results were inconclusive. 

To make the most of these data we needed a way to improve the spatial resolution. Many techniques exist for achieving such 'spatial deconvolution', but none were quite appropriate for enhancing the resolution of a faint dark spot against a brighter background. Hence, we had to develop our own, bespoke deconvolution algorithm, which we call 'Modified-CLEAN', which is applied to each wavelength in the recorded MUSE cubes using the observed Point Spread Function at that wavelength of the calibration standard star.

Modified-CLEAN, once tuned, has proven to be remarkable effective at de-blurring the images in our Neptune MUSE cubes and has enabled us to make the first ever detection of a dark spot from the ground, but more importantly recover its reflection spectrum. Analysing the data with our NEMESIS radiative transfer and retrieval code, we have shown that NDS-2018 is caused by a darkening of the particles in a layer of haze and H₂S cloud at ~5 bar, which we call 'Aerosol-1'. The darkening is wavelength dependent and only applies at wavelengths less than ~650 nm. What causes the darkening remains a mystery, however! 

A wholly unexpected discovery from this analysis was the presence of a bright cloud just to the southwest of NDS-2018. Bright 'companion' clouds are often seen with dark spots, but these are usually upper troposphere methane ice clouds at pressures of 200 - 600 mb. The new spot we discovered, however, was based much deeper, again down in the Aerosol-1 cloud/haze layer at ~5 bar, which we call DBS-2019 (Deep Bright Spot 2019). The spectral signature of DBS-2019 is consistent with a spectrally-dependant brightening the particles in this layer at wavelengths longer than ~650 nm. Again, what causes this is a mystery, but it is intriguing that both the NDS-2018 and DBS-2019 features are caused by a spectrally-dependent darkening or brightening of the same layer, which suggests some connection. What we do know, however, is that DBS-2019 was short-lived. Only the spectral resolution of instruments such as MUSE is able to distinguish such deep clouds from upper tropospheric methane ice clouds, but imaging observations would have detected the spot had they been viewing Neptune at the same time. However, imaging observations taken  just a couple of weeks before and after our observations on 18th October revealed no trace of DBS-2019.

VLT/MUSE has proven itself to be a very powerful tool for understanding clouds and hazes in Neptune's atmosphere. Now that we have analysed the NDS-2018 and DBS-2019 features, the next step is to look more generally at the variations in brightness across the disc in the different slices to understand how Neptune's aerosol structure varies with latitude.