Constraints on Neptune's haze structure and formation from VLT observations in the H-band

A 1-dimensional microphysics model has been used to constrain the structure and formation of haze in Neptune's atmosphere. These simulations were coupled to a radiative-transfer and retrieval code (NEMESIS) to model spectral observations of Neptune in the H-band performed by the SINFONI Integral Field Unit Spectrometer on the Very Large Telescope (VLT) in 2013. It was found that observations in the H-band and with emission angles ≤60° are largely unaffected by the imaginary refractive index of haze particles, allowing a notable reduction of the free parameters required to fit the observations. Our analysis shows a total haze production rate of (2.61 ± 0.18) × 10−14 kg m−2 s−1, about 10 times larger than that found in Uranus's atmosphere, and a particle electric charge of q = 8.6 ± 1.1 electrons per μm radius at latitudes between 5 and 15° S. This haze production rate in Neptune results in haze optical depths about 10 times greater than those in Uranus. The effective radius reff was found to be 0.22 ± 0.01 and 0.26 ± 0.02 μm at the 0.1 and 1-bar levels, respectively, with haze number densities of 8.48+1.78−1.31 and 9.31+2.52−1.91 particles per cm3. The fit at weak methane-absorbing wavelengths reveals also the presence of a tropospheric cloud with a total optical depth >10 at 1.46 μm. The tropospheric cloud base altitude was found near the 2.5-bar level, although this estimation may be only representative of the top of a thicker and deeper cloud. Our analysis leads to haze opacities about 3.5 times larger than that derived from Voyager-2 observations (Moses et al., 1995). This larger opacity indicates a haze production rate 2 times larger at least. To study this difference haze opacity or production rate, we performed a timescale analysis with our microphysical model to estimate the time required for haze particles to grow and settle out. Although this analysis shows haze timescales (~15 years) shorter than the time lapsed between Voyager-2 observations and 2013, the solar illumination at the top of the atmosphere has not varied significantly during this period (at the studied latitudes) to explain the increase in haze production. This difference in haze production rate derived for these two periods may arise from: a) the fact that in our analysis we employed spectral observations in the infrared (H-band), while Moses et al. (1995) used photometric images taken at 5 different filters in the visible. While high-phase-angle Voyager observations are more sensitive to small haze particles and at altitudes above the 0.1-bar level, the haze constraints derived from VLT spectra in H-band are limited to pressures greater than 0.1 bar. As a result of the different phase angles of the two set of observations, differences in the estimation of M0 may arise from the use of Mie phase functions as well. b) our 1-dimensional model does not account for latitudinal redistributions of the haze by dynamics. A possible meridional transport of haze with wind velocities greater than ~0.03 m s−1 would result in dynamics timescales shorter than 15 years and thus might explain the observed variations in the haze production rate during this period. Compared with our estimations, photochemical models point to even larger production rates on Neptune (by a factor of 2.4). Assuming that the photochemical simulations are correct, we found that this discrepancy can be explained if haze particles evaporate before reaching the tropospheric-cloud levels. This scenario would decrease the cumulative haze opacity above the 1-bar level, and thus a larger haze production rate would be required to fit our observations. However, to validate this haze vertical structure future microphysical simulations that include the evaporation rates of haze particles are required