Reanalyzed daily IASI/Metop-C ULB-LATMOS ammonia (NH3) L2 product (total column)

The reanalyzed daily IASI/Metop-C ammonia (NH3) total column dataset (ANNI-NH3-v4R) includes the results of a flexible and robust NH3 retrieval algorithm from measurements of the Infrared Atmospheric Sounding Interferometer (IASI). The method is based on the calculation of a hyperspectral range index (HRI) and subsequent conversion to NH3 columns via a neural network. With this method: (1) complete temperature and humidity vertical profiles can be accounted for, (2) third‐party NH3 vertical profile information can be used, (3) a full measurement uncertainty characterization is provided and (4) total column averaging kernels are provided. The ANNI-v4 IASI product builds on the heritage of version 1 ([Witburn et al., 2016](https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2016JD024828)), version 2 ([Van Damme et al., 2017](https://amt.copernicus.org/articles/10/4905/2017/)), improvements in the NN retrieval setup introduced in [Franco et al. (2018)](https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2018JD029633) and version 3 of the algorithm ([Van Damme et al., 2021](https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2018JD029633)) for the retrieval of volatiles organic compounds (VOCs). Recommendations for post-filtering the data and for averaging data in time or space are formulated in [Van Damme et al. (2017)](https://iopscience.iop.org/article/10.1088/1748-9326/abd5e0). Version 4 ([Clarisse et al., 2023](https://amt.copernicus.org/preprints/amt-2023-48/)) includes several corrections to improve consistency in time and across the different IASI instruments. It also introduces total column averaging kernels.

ANNI-NH3-v4R relies on ERA5 ECMWF meteorological input data ([Hersbach et al., 2020](https://rmets.onlinelibrary.wiley.com/doi/full/10.1002/qj.3803)), along with surface temperature and clouds retrieved from dedicated neural networks. This contrasts with the ANNI-NH3-v4 dataset, which is also available on the AERIS website, and which relies on Eumetsat IASI Level 2 meteorological input data (temperature and humidity profiles). For inter-annual comparisons and trends analyses, only the reanalyzed dataset should be used.

Current Version: V4.0.0R, available since 2023-07-03

Version history:
- V2.2.0R-I, available from 2018-06-15 to 2020-09-15
- V3.0.0R-ERA5, available from 2020-09-15 to 2022-01-14
- V3.1.0R-ERA5, available from 2022-09-14 to 2023-07-03

重新分析的每日IASI/Metop-C氨气（NH3）总柱数据集（ANNI-NH3-v4R）基于红外大气探测干涉仪（Infrared Atmospheric Sounding Interferometer, IASI）的观测数据，包含灵活鲁棒的氨气反演算法结果。该方法基于高光谱范围指数（hyperspectral range index, HRI）的计算，并通过神经网络将其转换为氨气总柱含量。采用该方法可实现：(1) 完整考虑温度与湿度的垂直廓线；(2) 引入第三方氨气垂直廓线信息；(3) 提供完整的测量不确定度表征；(4) 输出总柱平均核函数。

ANNI-v4 IASI产品继承了版本1（Witburn等，2016，https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2016JD024828）、版本2（Van Damme等，2017，https://amt.copernicus.org/articles/10/4905/2017/）的技术积淀，吸纳了Franco等（2018，https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2018JD029633）提出的神经网络反演设置改进，以及用于挥发性有机化合物（Volatile Organic Compounds, VOCs）反演的算法版本3（Van Damme等，2021，https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2018JD029633）。Van Damme等（2017，https://iopscience.iop.org/article/10.1088/1748-9326/abd5e0）提出了数据后滤波及时空平均的相关建议。版本4（Clarisse等，2023，https://amt.copernicus.org/preprints/amt-2023-48/）新增多项校正以提升不同时间维度及不同IASI仪器间的一致性，并首次引入总柱平均核函数。

ANNI-NH3-v4R依赖ERA5欧洲中期天气预报中心（European Centre for Medium-Range Weather Forecasts, ECMWF）气象输入数据（Hersbach等，2020，https://rmets.onlinelibrary.wiley.com/doi/full/10.1002/qj.3803），同时采用专用神经网络反演得到地表温度与云量信息。这与ANNI-NH3-v4数据集形成差异：后者同样可在AERIS网站获取，其采用欧洲气象卫星应用组织（Eumetsat）的IASI二级产品气象输入数据（温度与湿度廓线）。针对年际对比与趋势分析，仅可使用该重新分析数据集。

当前版本：V4.0.0R，自2023-07-03起正式发布。
版本历史：
- V2.2.0R-I，可用时段为2018-06-15至2020-09-15
- V3.0.0R-ERA5，可用时段为2020-09-15至2022-01-14
- V3.1.0R-ERA5，可用时段为2022-09-14至2023-07-03