Speaker
Description
ALTIUS (Atmospheric Limb Tracker for the Investigation of the Upcoming Stratosphere) is the upcoming stratospheric ozone monitoring mission of ESA’s Earth Watch program. ALTIUS consists of three 2D high resolution imagers: UV (250-355 nm), VIS (440-675 nm) and NIR (600-1040 nm) channels. Each channel is independent of the others and takes snapshots of the atmosphere in limb geometry at requested wavelengths. Stratospheric ozone profiles are the mission’s primary objectives. ALTIUS will measure in three different observation modes to maximize the spatial coverage: limb scattering on the dayside of the orbit, solar occultation at the terminator and stellar occultations on the nightside of the orbit.
The forward model used for the limb scattering L2 processing chain is a neural network trained on a large set of radiance profiles computed by a Monte-Carlo Radiative Transfer Model (RTM) called Smart-G. This approach is an attempt to make the retrieval algorithm benefit from the accuracy of a Monte-Carlo model, while circumventing the large computation load. While the neural network used as a proxy for Smart-G is orders of magnitude faster than a Smart-G simulation and provides good accuracy, it also comes with limitations in the number of state vector elements that can be used.
In this work, a parallel retrieval algorithm was built to explore the high-resolution capabilities of ALTIUS. In this retrieval algorithm, the forward model is Smart-G itself without the use of a proxy. Good Ozone retrieval performance is achieved on simulations and the vertical resolution of ALTIUS is assessed. Challenges caused by the use of a non-deterministic model (i.e. Monte Carlo) are discussed.
