Snow Cover Data Collection

Project Updates

• Recent Developments:

  • Snow sensors. The first prototype station moved outdoors at BRC in October 2026, and a second station is being built. Each station carries four ultrasonic snow depth sensors at different price points, tested side by side to find the lowest-cost sensor that is accurate enough to support a larger network. In bench testing, the station's long-range radio delivered 98.6% of its data over 25 days. A USU URCO grant (August 2026) is paying for sensor equipment.

    • This has two benefits: (1) real-time risk of hazardous road conditions; and (2) more information relevant to our forecasting of atmospheric inversions (cold pools) that settle atop the snow, and lead to high ozone events.
    • We built a labeling tool, and BRC students are labeling a random sample of the archive. A new machine learning model is in development to replace the current brightness-based method on basinwx.com/roads. Its first step, finding the road in each image, already matches hand-drawn road outlines about 89% of the time.
    • basinwx.com/roads drew about 850 visitors a month during its first winter. A BasinWx mobile app is in progress, along with a public survey asking drivers what road information they need and where.
    • In October 2026, USU submitted our proposal "Ground Truth: Teaching Utah's Traffic Cameras to Read the Road" to the state's Utah AI Moonshot program ($50,000). USU's President ranked it 4th of the 15 proposals the university submitted. UDOT is the anchor partner.

       

• Current and Upcoming Work:
  
  
  • Merging this effort of sensots with snow detection from UDOT camera. 
    
    
  • Deploying 6 to 8 stations across the Basin floor along the Duchesne, Roosevelt, and Vernal corridor this winter. Data will be published on basinwx.com.
    
  • Training the machine learning model on the labeled UDOT images.
    
    
  • Overhauling www.basinwx.com/roads as new products are developed.
    
    
  • Presenting the road camera work at the American Meteorological Society Annual Meeting in Denver, January 2027.
• Problems:
  
  
  • A planned agreement to host stations at a private company's tower sites did not work out because of its insurance requirements. We are now planning stations on BLM land and working on a land-use agreement. This may push some deployments later into the winter.

Snow Monitoring Stations

1. Winter ozone formation in the Uinta Basin requires a specific sequence of weather conditions. First, sufficient snowfall must persist after a winter storm. The snow cover, combined with high pressure and calm weather, creates a persistent cold pool (temperature inversion) that traps emissions from oil and gas operations in the Basin. These trapped emissions—primarily nitrogen oxides and volatileorganic compounds—are then exposed to sunlight. The bright snow surface reflects sunlight back into the atmosphere, driving chemicalreactions that convert the trapped pollution into ozone. Ozone is harmful because it irritates the respiratory system, aggravates asthma andother lung conditions, and can cause permanent lung damage with repeated exposure. The Uinta Basin is one of only two documentedlocations in North America where this phenomenon occurs, due to its unique combination of latitude, elevation, and terrain shape that allowsboth persistent snow cover and strong enough sunlight to produce unhealthy ozone concentrations.

2. Long distances between stations, once-daily reports with coarse precision, and mountains that block radar signals contribute to largedata gaps in the monitoring network on the Basin floor—particularly where oil and gas operations are concentrated (Figure 1).  The monitoring network has also deteriorated, with 11% of stations becoming inactive since 2018. Snow depth varies in ways current monitoringdoes not capture well, making observations less representative of initial conditions and limiting prediction accuracy. This project will deploy additional sensors and process available webcams for snow detection to address critical gaps.  These additional sensors will improve our quantitative ozone forecasting system, Clyfar (Lawson and Lyman, 2024), which is used to provide alerts to industry and others when high winter ozone is expected.

 We will deploy small, solar-powered, low-cost snow depth sensors at locations where coverage is missing. Each station includes an ultrasonicsnow depth sensor, temperature sensor, micro-controller with wireless capability, data storage, solar panel and battery system,weatherproof enclosure, and mounting hardware. Live data and daily summaries will be published on basinwx.com (our public datawebsite) for community access. We will use best practices for data quality control, including automated screening and manual checks to remove erroneous spikes, while still preserving genuine weather events. Filtered data and station status will be visible in real-time on basinwx.com.

Additional snow depth observations will improve Clyfar’s accuracy because snow cover is one of the key factors for high ozone formation in the Basin. More frequent measurements at more locations will help Dbetter understand when snow coverage may be“sufficient” to create the cold pool conditions that trap emissions. This is especially important because small differences in snow depth caninfluence whether a high ozone episode occurs or not.

 Our goal is to deploy sensors before the first significant winter snow of the 2026-27 winter season (anticipated October–November 2026) and conduct initial operations throughout winter 2026-27.  If the project is a success, operations will continue, and perhaps expand, in subsequent winters.