PROSNOW is a prediction model for meteorological and snow conditions in ski resorts. It enables real-time optimisation of grooming and snowmaking in ski resorts, thereby supporting their activities under increasingly challenging operating conditions due to climate change.
Ski tourism faces climate-related challenges at various interconnected time scales: how to efficiently adapt to ongoing climate change, now and in the future, using effective methods to cope with variable and declining snow cover, without committing to detrimental effects to the local and global environment in the long term? Several studies have documented the long term impacts and risks of various climate change scenarios on snow reliability in ski resorts in various mountain areas, and the need for the development of early warning systems to enhance the adaptation capacity of socio-economic sectors under current and future climate conditions.
In this context, the initial motivation of the PROSNOW project was to contribute to developing early-warning systems to better cope with the impact of the variability of meteorological and snow conditions on ski resorts operating conditions, and better adapt to increasingly challenging conditions in the future. One key requirement for early-warning systems is the provision of impact-based forecasts, at relevant time scales, designed to adapt operations, in real-time, to expected upcoming conditions, in order to reduce their potentially negative impacts, or benefit from upcoming opportunities. Given the main characteristics of the ski tourism business, in particular snow production in early winter, potentially annihilated by warm spells, anticipating meteorological and snow conditions at time scales from days to weeks, to months, is the relevant time scale to optimize snow management in mountain ski resorts.
The core concept of PROSNOW, developed through the EU H2020 PROSNOW from 2017 to 2020, was to focus on the snow cover on ski slopes, in the most integrative possible way, at time scales relevant to real-time decision making for snow management. Indeed, ski resorts managers are used to rely on weather forecasts, and intellectually combine this information with observations of the snow cover within the ski resort, taking into account their knowledge about snow processes. Snow cover modelling, i.e. the direct simulation of the state of the snow cover based on past and/or future meteorological conditions, has not been used for real-time ski resorts management hitherto, and this is the main gap that the PROSNOW project has contributed bridging. Indeed, one of the main benefits of numerical models is their capacity to integrate various sources of information and provide quantitative predictions beyond what the human mind can intuitively infer.
The project PROSNOW has brought together 13 academic and industrial partners operating in the field of the mountain snow cover in more or less direct relationship with ski tourism, but with different perspectives on how to approach this medium. Numerical snow cover models simulate the evolution of the snow cover, based on the time evolution of meteorological conditions (temperature, precipitation, wind speed, incoming radiation etc.), and the PROSNOW project has enabled further development of such models to account for grooming and snowmaking in a physically sound and operationally meaningful manner (Hanzer et al., 2020). This was developed thanks to the interactions between the diversity of partners involved in PROSNOW. Such models need to be fed by meteorological information, and the PROSNOW project has consolidated methods for feeding such snow cover models using not only past observations but also forecasts across weather (days to weeks) to seasonal (weeks to months) time scales. Because of the various sources of uncertainty involved, so-called “ensemble forecasting” was implemented, which provides a range of meteorological scenarios at various lead times, based on the output of numerical weather models, based on which probability ranges for the upcoming conditions can be inferred.
The PROSNOW system developed during the project combines an ensemble of meteorological and seasonal forecasts with an ensemble of model configurations, enabling to quantify the impact of various snow management tactics (e.g. various snow productions approaches, including no production), and therefore provide objective information to guide decision making in real time. Based on a detailed representation of the geographical organization of the ski resort, the system makes it possible to identify the subsectors of the ski resorts with most challenging snow conditions, and provide forecast of water consumption for snowmaking associated with each option. Thanks to snow cover modelling, the system makes it possible to provide outputs directly in the form of snow cover height or snow mass, which can be displayed on a ski resort map and compared to in-situ measurements. In fact, in-situ measurements of snow depth (using sensors embarked on grooming machines) or snow production can be used, in the PROSNOW system, to adjust the simulations and enhance the realism of the results, thereby fostering adoption and uptake by the users. Functional choices relevant to the development of the main configurations handled by the system and the design of the user-facing tool were developed thanks to a co-design approach involving pilot ski resorts. These specifications were confirmed by a wider-ranging survey across ski resorts in the European Alps, which further indicated the strong desire by the ski tourism industry to better manage its resources and a widely-shared opinion that better and more customized forecasts could indeed save resources and optimize their use.
The PROSNOW project has developed a dedicated Advanced Programming Interface (API) for exchanging snow-related information between ski industry service providers and stakeholders, thereby enabling an acceleration of data sharing in this sector. PROSNOW has also delivered improvements of existing open-source software elements, with immediate benefits to their community of users (e.g., snow cover models, or tools for the post-processing of seasonal forecasts). Scientific results from the PROSNOW project have been published in open access journals, with several publications still under preparation. The PROSNOW service is now brought to market across various countries in Europe. See online demo : https://prosnow.org/get-to-know-prosnow.
