Incorporating Subfreezing Road Probabilities into the WSSI

What tools are available to help forecasters anticipate if and when roads will become subfreezing?

Model of the Environment and Temperature of the Roads (METRo) model

METRo is a forecasting tool for road surface temperatures ( Crevier and Delage 2001 ). It is limited to Road Weather Information System (RWIS) point locations, and thus cannot be implemented CONUS-wide.

Probability of Subfreezing Road Temperatures (ProbSR)

ProbSR is a machine-learning algorithm that ties High-Resolution Rapid Refresh (HRRR) model surface variables to RWIS road temperatures, and outputs the probability that road surface temperatures will be <0°C ( Handler et al. 2020 ). Although it is trained using RWIS, the algorithm outputs subfreezing road probabilities for the entire CONUS.

ProbSR is currently a NWS-internal product available on the  MRMS Development Viewer .

The Weather Prediction Center ( WPC ) is currently developing a transportation-centric version of the Winter Storm Severity Index ( WSSI ), called WSSI Travel.

WSSI Travel is currently a WPC-internal prototype based on hourly HRRR data. It uses Geographic Information Systems (GIS) to communicate the spatial distribution of the potential impacts to surface transportation. It currently displays the overall impact in five levels (from "limited" to "extreme" impacts), as well as the impact from three sub-components:

• Snow Amount
  • Accumulated snowfall since HRRR initialization
  • Based on 2-day snowfall climatology developed for the operational WSSI (e.g., 4" of snow in Dallas has a higher impact than in Detroit)
• Snow Rate
  • Hourly snow rate
• Ice Accumulation
  • Ice accretion since HRRR initialization

Forecasters identified many use-case scenarios where the products would be useful for forecasting and messaging road hazards:

• When subfreezing roads are difficult to prescribe
  • e.g., late fall or early spring
• When subfreezing roads create hazardous road conditions for otherwise marginal or weak precipitation events
  • e.g., light snow, freezing drizzle, or flash freezes
• When the timing of the threat can increase transportation impacts
  • e.g., rush hours

Forecasters agreed that integration of ProbSR into WSSI Travel would be beneficial.

Integrate ProbSR outputs and non-meteorological factors into Precipitation Rate Components (Snow Rate and Liquid Rate).

WSSI Travel will now include impacts from all cool-season precipitation types! Users will also be able to visualize areas of mixed precipitation (e.g., mixed rain and snow).

ProbSR outputs will be used in WSSI Travel to help parameterize road surface conditions. These road conditions, along with other meteorological factors (e.g., wind speed, visibility, precipitation type and intensity), will provide a baseline for how disruptive the forecasted weather conditions are to surface transportation. For example, 0.5"/h snowfall is more disruptive on subfreezing roads, where the snow will begin to accumulate on untreated roads, than on above-freezing roads, where roads are more likely to remain wet.

Time-of-day and day-of-week (e.g., weekday vs. weekend) factors will be incorporated into WSSI Travel. These factors are based on ongoing research on the diurnal infleunces of winter weather to the following:

• Crash risk
• Crash severity
• Traffic volumes