Atmospheric water vapour is a key component of the Earth’s hydrological cycle, critical in shaping the global environment and supporting life on Earth as we know it. Manifold physical processes redistribute water from the oceans to the land and involve the formation of clouds, precipitation, and extreme weather events. Water vapour is also key in constraining the Earth’s energy balance. It is the most important natural greenhouse gas and constitutes a strong positive feedback to anthropogenic climate forcing from carbon dioxide. The water vapour feedback is critically important in understanding past and determining future climate change and its global and regional impacts.

There is consequently the need to consolidate our knowledge of natural variability and past changes in water vapour and to establish climate data records of both total column and vertically resolved water vapour for use in climate research. Such climate data records need to be homogeneous in space and time and have well-characterized uncertainties, which bears great challenges due to changing instrument characteristics and performances.

The Water Vapour Climate Change Initiative (WV_cci) tackles the challenges encountered in merging climate data records of water vapour, with the goal to provide climate modelers and researchers with long-term satellite records from current and past European (and other space agencies’) missions. Within its first phase (2018-2021), the WV_cci established user requirements through community involvement and workshops and produced climate data records (CDRs) of both total column (TCWV) and vertically resolved water vapour (VRWV). The CDRs have been quality controlled and extensively documented.

The TCWV CDRs include two products: the CCI TCWV-land (CDR-1), which is a gridded L3 data product over land based on ESA (MERIS, OLCI) and NASA instruments (MODIS), and the CM SAF/CCI TCWV-global (COMBI), which contains the CDR-1 over land, coasts, and sea-ice and the HOAPS microwave imager based TCWV data over ocean. Both datasets span the period from 2002-2017. The VRWV CDRs on the other hand include a zonal mean stratospheric product (CCI WV-strato), which contains a set of 11 different satellite limb sounders and extends between 1985-2019, and a prototype version of a CDR with specific focus on resolving the upper troposphere and lower stratosphere in 3D (CCI WV-UTLS) based on RAL IMS, ENVISAT MIPAS, and Aura-MLS data. A detailed overview of these CDRs will be provided in this presentation, along with first scientific results.

The WV_cci now has entered its second phase and this contribution will also provide an outlook on the steps forward which aim at improving the current WV_cci CDRs and gain a deeper knowledge of climate processes relating to water vapour.

Significant observational uncertainties limit our ability to assess changes in the vertical distribution of water vapour and therefore constrain the link with climate change. ESA’s Climate Change Initiative (CCI) was established to tackle the challenges encountered in merging climate data records (CDRs) of Essential Climate Variables and to provide climate modelers and researchers with stable long-term records from current and past European (and third-party) satellite missions.

As part of the CCI Water Vapour project, two vertically resolved water vapour data records have been developed and these were recently released. The first CCI H2O CDR (CCI-strato) provides monthly zonal mean profiles from the mid-1980s onwards covering the stratosphere and lower mesosphere, combining observations by limb and occultation sounders. The second CCI H2O CDR (CCI-utls) merges nadir and limb measurements to obtain latitude- and longitude-resolved monthly mean profiles between 2010-2014 covering the troposphere and the UTLS region.

In this contribution, we present our assessment of their potential for climate applications through comparison to other state-of-the-art CDRs used by the community. We start with an overview of the input profile data, the merging methods and the general characteristics of these CCI data records, and relate these to those of other water vapour CDRs developed by NOAA (SWOOSH) and NASA (GOZCARDS). This provides the necessary background to properly interpret the (dis)agreements between vertically resolved H2O CDRs.

All CDR time series were analysed using multiple linear regression (MLR) to estimate properties of the multi-annual mean, seasonal cycle, Quasi-Biennial Oscillation and long-term trends of water vapour across the stratosphere. The coherence of the zonal and vertical structure of each of these patterns is discussed. We conclude that the magnitude and structure of cyclic patterns inferred from the different CDRs are very similar. Residual, long-term water vapour changes are modelled as a piecewise continuous linear function inflecting in January 1998 and January 2004. The sign of the trend in all three periods is generally in good agreement between the CDRs. However, their magnitude can differ significantly in (part of) the stratosphere during the 1998-2003 and 2004-2019 periods. We related these differences to, e.g., differences in contributing sensors and data versions. In conclusion, the ESA CCI vertically resolved water vapour data records represent a valuable new, independent source of information for global long-term studies and climate applications.

The Spatial Heterodyne Observations of Water (SHOW) instrument is one of three instruments that are being studied as a potential Canadian contribution to NASA’s planned Atmospheric Observing System (AOS). SHOW observes limb scattered sunlight within a narrow spectral window (3 nm) in a vibrational absorption band of water near 1364.5 nm with an unapodized spectral resolution of 0.02 nm. The limb scattered radiance is inverted to obtain high vertical resolution (< 0.5 km) water vapour profiles in the UTLS region. The high SNR provided by the limb scattered signal and large instrument throughput facilitates rapid along track sampling (< 100 km) with < 10 % uncertainty. In this paper, we present the work that has been done to improve the forward model and retrieval approach in an effort to enhance sensitivities in the presence of clouds and high aerosol loads, which are known to pose challenges for this technique. It is shown that co-located observations of aerosol and cloud properties by the other two proposed Canadian instruments, ALI and TICFIRE, facilitate the retrieval of water vapour down to the cloud top and improve the overall sensitivity of the observations. We present a case study examining the ability to spatially and temporally resolve convective moistening of the lowermost stratosphere associated with overshooting tops. It is shown that the approach reveals important variability that is not captured by current state of the art satellite sensors.

In the upper troposphere and lowermost stratosphere (UTLS) region the air is coldest and driest, and the distribution of water vapour shows a large spatial and temporal variability. This makes it difficult to detect trends in water vapour records in this region. However, a data set which is ideally suited for long-term characterization of water vapour distribution in the UTLS is the IAGOS time series. IAGOS, starting in 1994. (In-Service Aircraft for a Global Observing System; www.iagos.org) is a European Research Infrastructure which uses commercial passenger aircraft for the continuous and global-scale observation of atmospheric composition, water vapour, and temperature. Together with the data from the IAGOS predecessor MOZAIC (Measurement of Ozone and Water Vapor by Airbus In-Service Aircraft), the IAGOS time series spans now more than 27 years and is comprised of more than 60000 flights. In this study we will use the IAGOS data to compare the distribution of Relative Humidity (RH) for different years and seasons and analyze the 27 years long time series for potential trends.

The upper troposphere and lower stratosphere (UTLS) are of special interest for dynamical and chemical processes as well as the associated radiative forcing. The water vapour in this region plays a key role in stratosphere-troposphere exchange but is subject to uncertain spatial distributions and temporal variations, and detection limits of instruments. Besides, a significant wet bias on the order of 100 % to 150 % in the stratospheric layers just above the tropopause has been found from ECMWF re-analysis humidity datasets compared with airborne in-situ measurements in the ML-CIRRUS campaign. And the dry bias of humidity in the UT also leads to the underestimation of contrail persistence relative to IAGOS/MOZAIC data. We were motivated by these challenges and built a Random Forest (RF) algorithm to improve relative and specific humidity data records.
This approach performed the model training of humidity predictions based on available parameters (modelled thermodynamic conditions and dynamical states, as well as measured water vapour from IAGOS/MOSAIC) with a high calculation speed. Considering the local atmospheric flow, we first assessed effects of weather conditions from a couple of hours before and potential influences of the vertical transport of UTLS substances on water vapor. Sensitivity analyses indicate that the combination of previous and current atmospheric states within the pressure difference of 100hPa presented the highest relationship with current humidity status. This points out to the need of accounting for temporal and spatial meteorological variables to evaluate atmospheric water vapor. We then identified the model and controlled the training accuracy to the correlation coefficient (CC) of 0.99. After applying our algorithm to ECMWF atmospheric datasets, overall algorithm results validated with the ML-CIRRUS campaign measurements showed that the estimated relative and specific humidity have a mean absolute error (MAE) of less than 20% and a root mean square error (RMSE) of no larger than 30%, which enhanced the average accuracy of formal model re-analysis humidity data. Based on the Random Forest outputs, we are capturing ice supersaturated regions favorable for contrails persistence in slow ascending or convective airmasses. This study addresses the significance of dynamical variables in water vapor forecasts and may also be valuable for improving model driven data and generating long-term water vapour essential climate variables (ECVs) for climate research.