Reading notes-14

Accurate modelling of the radiative transfer through the atmosphere plays an important role in the inversion.

The forward model adopted in the retrieval algorithm of AIUSis the RFM (reference forward model) from DUdhia, with the latest release being version v4.36(htt://eodg.atm.ox.ac.uk/RFM/index.html).

Von Clarmann et al have compared five forward models including the RFM. The inter-comparision experiment showed that an overall agreement between all models was reached, and RFM is proven to produce reliable simultations.

The inversion algorithm employed in this study is based on the OEM(optimal estimation method) proposed by Rodgers.

The OEM retrieval framework has been widely applied to the inversion of atmospheric state parameters using infrared and microwave remote sensing measurements, with nadir, occultation, or limb viewing.

OEM stabilizes the inversion process by taking into account the statistical information about the atmospheric variability, which has been investigated by many studies.

The forward model generates a numberical simulation of measurements based on the given atmospheric state. In other words, the accuracy of the simulated measurements depends on the reliability of the atmospheric parameters used in the forward model.

In the retrieval scheme, a dataset of integrated atmospheric profiles based on the MLS( Microwave Limb Sounder) level 2 products, ACE-FTS Climatology-Version 3.5, and the AFGL(Air Force Geophysics Laboratory) atmospheric models was generated.

The MLS level 2 (v4.2) products between 2014 and 2016 were considered. The species profiles are classified and stored month by month for eath set of products.

Then, the monthly mean profiles are acquired and classified into different coordinate grids, which is discretized with a 5° latitude and 30° longitude spacing.

That is, both of MLS and ACE-FTS datasets are classified by a monthly and geolocation grid. A diagram for constructing the integrated atmospheric profiles is shown in Figure 7.

The next step was to combine the two sets of profiles. Since ACE-FTS and AIUS have similar instrument characteristics, the ACE-FTS product was chosen in the case that the profile of a particular species at the same geolocation and time can be found in both ACE-FS and MLS datasets.

Finally, the remaining species were read from the AFGL dataset. The profiles imported from the AFGL dataset include N3, HBr, HI, PH3, H2S, F11, F12, F13, F21, F22, F114, F115, and HNO4. These species profiles from FASCOD(Fast Atmospheric Signature Code) Models 1-6 are resampled and integrated into the monthly geographic grids. In the end, the integrated atmospheric profiles dataset was produces.

Before performing the retrievals for the AIUS measurements, the retrieval experiments were carried out based on the simulated spectra and ACE-FTS instrument to test and verify our algorithm.

We had no detailed parameters of ACE-FTS instrument, such as instrument line shape, the antenna response, etc. Thus, the parameters of AIUS were adopted to perform the retrieval experiments since the two instrument characteristics were similar. We adopted three ACE-FTS level 1 data sets to perform the O3 retrieval experiments.

Information of the ACE-FTS level-1 data used in this study is listed in Table 2. All the a priori profiles of O3 and the other species were taken from the dataset of inteegrated atmospheric profiles described in Section 3.3

The retrieval profiles and the relative differences are shown in Figure8. The solid lines represent the O3 profile retrieved by the AIUS retrieval algorithm and the dotted lines are ACE-FTS O3 products， which were resampled to the retrieval grid.

引自《Monitoring Trace Gases over the Antarctic Using Atmospheric Infrared Ultraspectral Sounder Onboard GaoFen-5: Algorithm Description and First Retrieval Results of O3, H2O, and HCl》