**4. Conclusions**

Within the ESA business applications demonstration project SnowSense (2015–2018), we successfully demonstrated a large scale snow hydrological monitoring service, by combining newly developed in situ stations based on signals of the Global Navigation Satellite System (GNSS), Earth Observation (EO) and hydrological modelling. With this combined approach, we present a reliable, and cost-efficient tool for the determination of snow cover properties like snow water equivalent (SWE), snow liquid water content (LWC), snow extent as well as run-off assessment, for real-time and forecast applications.

The GNSS in situ component was successfully applied and validated at the well-equipped study site NEIGE at Forêt Montmorency, Quebec, Canada. Furthermore, the entire SnowSense service providing modelled, in situ-, and EO-assimilated run-off was applied and validated at four run-off gauges within the Humber River and the Exploit River catchments on the island of Newfoundland, Canada.

The entire SnowSense service solution driven with an integrated numerical weather prediction (NWP) model for its application in this study in Newfoundland is capable of providing detailed knowledge on water stored as snow over large spatial scales. It is able to provide real-time and forecasted snow and run-off information and, if desired, also on reservoir status, which might be of grea<sup>t</sup> interest for hydropower plant operators. This information, which can be provided in various time steps, e.g., hourly up to daily, is especially needed in regions or catchments where in situ stations are only sparsely or non equipped catchments. The service is applicable at almost any location and was especially designed for remote locations, where access is limited and snow and run-off measurements were difficult up to now.

Within the project, the SnowSense service already reached a market dedicated design, based on the identification of potential customers (i.e., hydropower plants) and use cases (i.e., weather and climate observations, e.g., by national weather services).

**Author Contributions:** F.A., H.B. and W.M. conceived and developed the project idea; F.K., M.L., P.H. and F.A. were deeply involved in the development of the SnowSense algorithm, hardware development, and model integration; P.K. and H.B. were responsible for the PROMET model setup and the EO data assimilation. F.A. was the SnowSense project manager; F.A., P.H., F.K. and W.M. are involved in the patenting of the algorithm and the hardware; A.R., F.A. and F.K. wrote the paper.

**Funding:** The project was co-funded by the European Space Agency (ESA, 000113149/14/NL/AD) within the Business Application Demonstration Project SnowSense (2015–2018) (https://artesapps.esa.int/projects/ snowsense-dp).

**Acknowledgments:** We thank the editors for their work and the two anonymous reviewers for their constructive comments. Florian Appel, Franziska Koch, Patrick Henkel, Markus Lamm, Philipp Klug and Anja Rösel were co-funded by the European Space Agency (ESA, 000113149/14/NL/AD) within the Business Application Demonstration Project SnowSense (2015–2018) (https://artesapps.esa.int/projects/snowsense-dp), which is gratefully acknowledged. The authors would like to thank all involved staff members of the demo users Nalcor/NL Hydro and the Department of Municipal Affairs and Environment, Water Resources Management Division (WRMD) of the Government of Newfoundland and Labrador, for their active support to the project, and for providing their run-off data. We thank the staff from University Laval for the opportunity to set up our sensor at the NEIGE site in Forêt Montmorency and for the opportunity to participate in the inter-comparison study. For the developed algorithm for the GNSS-based SWE and LWC determination as well as the invention of the in situ hardware, two patents are pending. SnowSense -R is a registered trademark.

**Conflicts of Interest:** The authors declare no conflict of interest.
