MDSS Challenges, Research, and Managing User Expectations Weather Issues

1
MDSS Challenges, Research, and Managing User
Expectations - Weather Issues -

• Bill Mahoney
• Kevin Petty
• National Center for Atmospheric Research
• (NCAR)
• MDSS Stakeholder Meeting
• Vienna, VA
• August 2006

2
Issue Statement

• The MDSS is a sophisticated technology that
• requires accurate weather information on very
  small
  
• scales. Almost all MDSS products (road and
• weather) are dependent on accurate weather data
• input.
• Given this dependency and the knowledge that
• weather prediction is still imprecise at these
• scales
• 1) How much utility does the system have?
• 2) What research is required to improve its
  skill?
  
• 3) How do we manage user expectations?

3
Primary MDSS Limitations

• Inconsistent weather prediction skill
• Lack of knowledge of actual road conditions
• Inconsistent pavement condition data quality
• Lack of knowledge of actual treatments
• Lack of universal methodology to incorporate
  actual treatment data
  

Actual Treatments
4
MDSS Weather Challenges
For optimal performance, weather forecasts need
to be very precise on city block spacial scales
and on a time scale of minutes!
While weather forecast precision is improving, i
t will be many years before the skill will match
the need. OK Now What?
Washington, D.C. mall region
5
The Weather Forecast EnterpriseWorldwide
Forecasting Issues
Why are weather forecasts often inaccurate?
There is a lack of global weather observations t
aken at high spatial and temporal resolution and
a lack of an ability to fully utilize current
observational data.
We dont know the state of the atmosphere very
well!
6
The Atmosphere is a Fluid
Water vapor false color imagery
The Earth System is well connected!
7
Primary Weather Prediction Limitations

• Lack of consistently accurate weather
  predictions
  
• Precipitation start and stop times
• Precipitation amounts (particularly for light
  events)
  
• Cloud cover (solar radiation)
• Water vapor (fog, frost, dew, etc.)
• Lack of accurate observational weather data
• Snow amounts (depth and liquid water equivalent)
• Freezing drizzle
• Insolation
• Radar data quality
• Lack of knowledge and utilization of land surface
  data
  
• Soil moisture
• Soil temperature
• Terrain features (e.g., road cuts, small lakes,
  clearings, forest boundaries)
  
• Snow cover albedo

8
MDSS Research Needs - Weather

• Weather Land Surface Modeling
• Boundary layer meteorology (friction, turbulence,
  water vapor flux, heat exchange, etc.) and land
  surface observations, data assimilation, and
  modeling
• Need to assess performance of surface weather
  forecasts using models that utilize new data
  assimilation methods and coupled land surface
  models.
• Need techniques to assess utility of solar
  radiation observations
  

9
SampleGeneral Errors in Air Temperature
Prediction
Raw models surface temperature errors 2.5 to
3.0 oC
MDSS Road Weather Forecast System Forecast
Lowest 10 meters of boundary layer
hard to predict! Radiational cooling layer
dominates errors.
10
Sample Large Variation in Predicted
Precipitation
11
Sample Direct Solar Radiation Prediction
Partly Cloudy Day
Model Output Differences
Clear Day
12
Deterministic vs. ProbabilisticForecasts
The chaotic character of the atmosphere coupled
with inevitable inadequacies in observations and
computer models, results in forecasts that always
contain uncertainties. The MDSS and its users
, who are risk managers, need to deal with these
uncertainties.
NAS Report July 2006
13
Communication of Uncertainty

• Given the weather and road condition prediction
  is not precise, we need to develop methods to
  convey uncertainty to decision makers.
  
• Need user working group to help define product
  concepts (text and graphics) for multiple
  variables.