Version 2 of the IASI NH<sub>3</sub> neural network retrieval algorithm: near-real-time and reanalysed datasets

Abstract

Abstract. Recently, Whitburn et al. (2016) presented a neural-network-based algorithm for retrieving atmospheric ammonia (NH 3 ) columns from Infrared Atmospheric Sounding Interferometer (IASI) satellite observations. In the past year, several improvements have been introduced, and the resulting new baseline version, Artificial Neural Network for IASI (ANNI)-NH 3 -v2.1, is documented here. One of the main changes to the algorithm is that separate neural networks were trained for land and sea observations, resulting in a better training performance for both groups. By reducing and transforming the input parameter space, performance is now also better for observations associated with favourable sounding conditions (i.e. enhanced thermal contrasts). Other changes relate to the introduction of a bias correction over land and sea and the treatment of the satellite zenith angle. In addition to these algorithmic changes, new recommendations for post-filtering the data and for averaging data in time or space are formulated. We also introduce a second dataset (ANNI-NH 3 -v2.1R-I) which relies on ERA-Interim ECMWF meteorological input data, along with surface temperature retrieved from a dedicated network, rather than the operationally provided Eumetsat IASI Level 2 (L2) data used for the standard near-real-time version. The need for such a dataset emerged after a series of sharp discontinuities were identified in the NH 3 time series, which could be traced back to incremental changes in the IASI L2 algorithms for temperature and clouds. The reanalysed dataset is coherent in time and can therefore be used to study trends. Furthermore, both datasets agree reasonably well in the mean on recent data, after the date when the IASI meteorological L2 version 6 became operational (30 September 2014).

Stratospheric aerosol radiative forcing simulated by the chemistry climate model EMAC using aerosol CCI satellite data

“…It is based on a statistical regression technique and the use of a neural network trained on synthetic IASI data. A similar scheme has already been applied for the retrieval of NH3 (ammonia) (Whitburn et al, 2016 andVan Damme et al, 2017). As input variables it uses the IASI L2 pressure, humidity and temperature information, spectral information and a CALIPSO (Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation) derived dust altitude climatology.…”

Validation of the IASI FORLI/Eumetsat ozone products using satellite (GOME-2), ground-based (Brewer-Dobson, SAOZ) and ozonesonde measurements

“…The Eumetsat Level-2 data (pressure, water vapor, temperature and 5 clouds) are used as input in FORLI. It is worth mentioning that the Eumetsat dataset is not homogenous since it has been processed using different versions of the IASI Level-2 Product Processing Facility between 2008 (v4.2) and2016 (v6.2), as summarized in Van Damme et al (2017). The error budget of the retrieved O 3 profile shows that the dominant errors originate from the limited vertical sensitivity, from the measurement noise and from uncertainties in the fitted (water vapor column) or fixed (e.g.…”

“…In order to assess the latitudinal variability of IASI O 3 retrieval performance, the comparison is performed for six 30° latitude bands representative of the northern high latitudes (60-90°N), northern mid-latitudes (30-60° N), northern tropics (0-30°N), southern tropics (0-30°S), southern mid-latitudes (30-60°S) and southern high latitudes (60-90°S). 30 processing of the Eumetsat IASI Level-2 processor (see Table 1 in Van Damme et al (2017) for a summary of the changes in Eumsetsat Level-2 data), which is likely to induce some inconsistencies but no progressive drifts in the retrieved columns.…”

A physics-based approach to oversample multi-satellite, multi-species observations to a common grid

“…Moreover, many atmospheric species have strong spectroscopic signatures in the mid-infrared and can be retrieved from the Earth's thermal emission spectra col- 15 lected by satellite sensors such as MOPITT (Drummond et al, 2010), AIRS (Aumann et al, 2003), TES (Bowman et al, 2006), IASI (Clerbaux et al, 2009), and CrIS (Han et al, 2013). One species of particular significance to tropospheric chemistry and air quality is NH 3 (Baek et al, 2004;Paulot and Jacob, 2014), which has been successfully retrieved from TES (Shephard et al, 2011;Sun et al, 2015a), AIRS (Warner et al, 2016), IASI (Clarisse et al, 2010;Whitburn et al, 2016a;Van Damme et al, 2017), and CrIS (Shephard and Cady-Pereira, 2015;Dammers et al, 2017). 20 The retrieval results from satellite sensors are usually total or partial (e.g., tropospheric or planetary boundary layer, PBL) column density at individual satellite pixels, i.e., the Level 2 product.…”

“…and subsequent conversion to NH 3 columns via a neural network (Whitburn et al, 2016a;Van Damme et al, 2017). The IASI-NH 3 datasets are publicly available for both IASI-A and IASI-B, with the version 2 (Van Damme et al, 2017) presenting significant improvements over version 1 (Whitburn et al, 2016a), including the negative values that are crucial for observational error averaging near the detection limit.…”