Local prediction of avalanche danger on mountain roads

The avalanche prevention and control system, Smarty Snow, enables the assessment of avalanche risk in order to prevent damage to infrastructure.

In the Pyrenees, as in many mountainous regions around the world, avalanches play a significant role in winter communication management. In this article, we explain how prediction teams work to evaluate the avalanche danger on affected roads. Despite the uncertainties involved in avalanche prediction, a rigorous working system allows for the collection of meaningful field data to make the best possible assessments.

At Arantec, we have been developing the Smarty Snow system for years, which combines avalanche prevention and control, snow management, and winter road maintenance. Through the use of nivometeorological stations, the Smarty Snow system monitors the risk of snow avalanches in the mountains to prevent damage to infrastructure. Parameters such as snow depth, surface temperature, ambient temperature, relative humidity, solar radiation, wind speed and direction, and precipitation are monitored using Smarty Planet nivometeorological stations.

Working System

By studying the terrain and analyzing the historical records of avalanches, an atlas of avalanche-prone areas affecting specific roads is created. Each avalanche zone, consisting of a start zone, track zone, and runout zone, is mapped and characterized by various parameters, including code and name, orientation, slope angle, terrain type, elevation of the start zone, vertical drop to the road, recorded avalanches, maximum observed reach, and more. This cartographic database serves as the foundation for ongoing work and the storage of avalanche activity season after season.

Figura 1: Cartografía de las zonas de aludes que afectan la carretera A-2606 a Baños de Panticosa
Figure 1: Cartography of avalanche-prone areas affecting the A-2606 road to Baños de Panticosa.

To monitor the meteorological conditions and snowpack throughout the season, a network of observation is designed, consisting of:

  • Local observers, who live and work near the study area, who share daily meteorological data and avalanche activity.
  • Nivometeorological stations, located near the avalanche release areas, which provide continuous information on air temperature and humidity, wind, precipitation, radiation, snow temperatures, and snow depth.
  • Webcams, located on the opposite slope of the mountain, which provide images of snow coverage and avalanche activity on the study slopes.

Figure 2: Automatic station located in La Raca for avalanche prediction on the N-330a road in Somport.

This remote monitoring, along with the weather forecast monitoring, allows for the detection of the need for specific field visits. These visits can have three main objectives:

  • Analyzing the snowpack before a snowfall event.
  • Monitoring a snowfall or snowpack warming event.
  • Tracking the evolution of the snowpack after a period of instability.

Field work consists of conducting stratigraphic profiles and stability tests, along with observing recent signs such as avalanches, wind effects, and temperature effects. Based on this data, a Nowcast is created, providing a snapshot of the current state of the snowpack in the relevant orientations and altitudes.

Figure 3: Field work in the avalanche starting zone of Gabarda for avalanche prediction on the A-2606 road in Baños de Panticosa.

To obtain an avalanche hazard prediction or Forecast, we start with the current state of the snowpack (nowcast) and estimate how it will change based on the expected weather. This requires consulting the weather forecast through numerical models at different scales. Weather and snowpack conditions are closely related and fluctuate according to known patterns.

The final product is a Local Avalanche Danger Bulletin (Boletín del Peligro de Aludes Local or BPAL), which estimates the probability and size of avalanches that could occur in the avalanche-prone areas defined in the atlas.

Figure 4: Example of a Local Avalanche Danger Bulletin for the A-2606 road to Baños de Panticosa. It uses a 5-level avalanche danger scale. This scale is a customized adaptation of the European Avalanche Danger Scale based on the probability of affecting roads.

Once the local avalanche danger is established, the next step is to manage the risk. If the risk exceeds the acceptable level for a particular type of road at a given time, actions are taken to either reduce exposure (limiting or suspending traffic) or mitigate the avalanche danger through artificial triggering.

There are many systems for artificial triggering, all aimed at causing avalanches to reduce the danger or clearing the starting zones to prevent further accumulation of snow. If no artificial triggering system is available, road closures may be extended for several days, waiting for the snowpack to stabilize naturally.

Figure 5: Artificial triggering system with Gazex.

Figure 6: Artificial triggering system with DaisyBell.

On the other hand, the presence of defense structures reduces the likelihood of avalanches reaching the road. The counterpart of these systems is their extremely high cost. This includes examples such as avalanche nets, barriers, and visors.

Figure 7: Avalanche barriers

Figure 8: Avalanche deflector in the Senarta I avalanche area on the A139 road to Llanos del Hospital.

Figura 9: Containment barrier in the avalanche area of Viseras on the A-2606 road to Baños de Panticosa.

Clarifying Concepts

There are some important terms in the world of avalanches that are commonly used incorrectly. We provide a brief explanatory explanation below.

  • Avalanche Hazard vs. Avalanche Risk: The hazard depends on the mountain conditions and is the probability of a specific intensity (number and size) avalanche occurrence. The risk is a function of the hazard, exposure, and vulnerability. Managers install anti-avalanche systems and apply artificial triggering methods to modify the hazard. They open and close roads to regulate exposure and provide avalanche rescue training to operators to reduce their vulnerability.


  • Regional Avalanche Hazard vs. Local Avalanche Hazard: Often, the regional avalanche hazard bulletin for an area (e.g., AEMET – Pyrenees – Ribagorza) and the local avalanche hazard bulletin for a specific road (e.g., A139 to Llanos del Hospital) establish different hazard levels for the same day, even though they are in the same geographical area. These approaches may seem contradictory, but they are not because a regional prediction takes into account all orientations in different altitude zones of the mountain range, while a local prediction only considers specific avalanche zones with defined orientations and altitudes. This leads to different avalanche problems and, consequently, different probabilities and sizes of expected avalanches.

How we work at Arantec Engineering

Monitoring the snow volume of large areas without the need for costly trips, having historical data on snow behavior and evolution, knowing the snow depths in different points in real-time, knowing the snow temperature, and having remote information on the status of avalanche-prone areas and reducing the risk for people through our nivometeorological stations allow us to control the risk of snow avalanches in the mountains and reduce the costs of manual measurements by reaching many more measurement points.

A clear example of this is the installation of Arantec’s automatic nivometeorological station in the Pyrenees of Huesca, located on the right slope of the A-2606 road that provides access to the Panticosa spa and ski resort from the town. Historically, this road has been temporarily closed due to intense snowfall episodes or minor snowslides from the slope.

The nivometeorological station allows the measurement of the following variables: wind speed and direction, ambient temperature and relative humidity, rainfall, solar radiation, and snow depth and quantity.

The station is equipped with a remote web-cam that provides a view of the mountain and the road to have a system for monitoring and surveillance of possible avalanches. The dome-type camera, in addition to having remote access and a 360° field of view, captures periodic images every 30 minutes with a resolution of 2 Megapixels (1920 x 1080). The nivometeorological station and web-cam are powered by photovoltaic panels and communicate data via GSM-4G.

The Government of Aragon, through the Department of Territorial Development, Mobility and Housing, which is responsible for the conservation and maintenance of roads in the Autonomous Community of Aragon, as well as informing the population about their status and ensuring mobility and road traffic, has access to the Smarty Planet platform. Through this platform, they can remotely monitor all the parameters provided by the two equipment mentioned above, establish corresponding thresholds for each parameter, and set up alert mechanisms to optimize road management.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top