Path-Integrated tropospheric water vapor from a mountain-to-mountain microwave link: a summer/autumn NDSA campaign compared with ERA5 and instrumental data

Water vapor (WV) plays a fundamental role in tropospheric processes, as most atmospheric moisture is confined to this layer. However, homogeneous and globally distributed observations of the lower troposphere—up to about 5–6 km altitude—remain limited. Filling this observational gap would significantly improve short-term climate analyses and the performance of numerical weather prediction (NWP) models.

Within theoretical activities supported by ESA, a novel retrieval concept called Normalized Differential Spectral Attenuation (NDSA) was developed to estimate integrated water vapor (IWV) from microwave attenuation measurements in the 17–21 GHz frequency range along tropospheric propagation paths. The method is based on the estimation of a spectral sensitivity coefficient (S), defined as the differential attenuation between two closely spaced carrier frequencies with a relative separation smaller than 2%. We demonstrated a linear relationship between S and IWV, enabling a simple and robust retrieval scheme. These investigations also highlighted the suitability of NDSA for spaceborne applications, including co- and counter-rotating Low Earth Orbit (LEO) satellite geometries. The Italian Space Agency funded the SATCROSS project to assess the technological feasibility of a dedicated satellite mission and to develop a ground-based prototype capable of performing NDSA measurements on terrestrial microwave links at 19 GHz.

A critical step toward an operational space-based system is the quantitative assessment of the accuracy and reliability of IWV estimates derived from the prototype through validation against independent observing techniques. A first validation campaign was in 2024, comparing IWV retrieved by the NDSA prototype with measurements from a MAX-DOAS instrument observing the same atmospheric volume along a 91 km link between the meteorological station “Giorgio Fea” (San Pietro Capofiume, 10 m a.s.l.) and the climate observatory at Mount Cimone (2165 m a.s.l.). Additional reference data were provided by radiosondes, hygrometers, and GNSS. While the results were encouraging, significant signal scintillation affected the NDSA measurements due to a large fraction of the link remaining within the terrain boundary layer.

The present work focuses on a second campaign carried out in 2025 along a 160 km high-altitude microwave link connecting Mount Cimone to Mount Amiata (1738 m a.s.l.). For the first time, the NDSA prototype was tested on a link with nearly constant height and limited ground influence, closely approximating the geometry of a LEO-to-LEO satellite link with a tangent height of about 2000 m. This setup enabled verification of the theoretical relationship between the spectral sensitivity parameter S and IWV, with particular attention to the linear model coefficients reported by the authors in previous papers. ERA5 reanalysis data (25-km linear res.), integrated along the full link, were also compared with in situ hygrometer measurements and GNSS-derived IWV. Overall, IWV estimates from the different techniques show good agreement in capturing daily and seasonal variability, while ERA5 systematically underestimates IWV due to its coarser resolution. At shorter timescales, discrepancies increase during periods of enhanced tropospheric turbulence, induced by air mass movements. Criteria for real-time identification of high-scintillation conditions were defined, demonstrating the capability of NDSA to detect precipitation while preserving WV information.