About

Stage 1

The Ariel Spacemission Atmospheric Retrieval Challenge is designed to facilitate the comparison between atmospheric forward models as well as atmospheric retrieval codes in preparation for the Ariel red book.

The early stages of the challenge

At this stage, we will start off with model comparison exercises (see non-blind study below) and very simple blind retrieval comparisons. In stage 2 of the competition, we will provide models of varying complexity, fully blind, provided by a range of atmospheric forward model codes. These fully blind retrievals will be ranked according to retrieved parameter accuracies as well as forward model accuracies and the scores will be reported on a leader board.

The non-blind study

We will provide an initial four transmission/emission spectra of two hot-Jupiters and two super-Earths for model comparison. These will be relatively simple models to start with but at later stages of the retrieval challenge, these will be come harder.
In the non-blind study, we will provide the ground truth (i.e. the input values).

The blind study

Here, we will provide four extra planets, similar in type to the non-blind study but without the ground truth provided. In later stages of the competition we will expand the blind retrieval challenges to reflect a range of cloud and chemical models.