Future Initiatives

Applications for the "In-house" Weather Analyst

Weather affects a wide range of industries from energy to insurance to commodities investing. As such, many organizations employ "in-house" meteorologists and atmospheric scientists. Private sector scientists are inevitably at the forefront of day-to-day corporate activities and often lack the time and resources to build background tools to make them more efficient and effective in their duties. Some such tools are currently available but not customizable or operationally supported. Existing tools are also scattered throughout the world of cyberspace and difficult to locate. SPHEAR proposes a centralized network of tools such as:

• Real-time performance diagnostics for the primary global, mesoscale, and tropical forecasting models. Provide the private-sector meteorologist with a suite of tools to analyze model skill in real-time. Goal: help the "in-house" meteorologist produce forecasts more efficiently and effectively.
• Tools to better understand the variables that lead to hemispheric jet-stream patterns within and beyond the two-week forecast window. Quantify which atmospheric and/or oceanic signals have the most power to explain significant changes in global atmospheric patterns within and beyond the two-week forecast window. Goal: give the "in-house" meteorologist a suite of products to better predict intra-seasonal weather patterns.
• Multi-model ensemble techniques for weather prediction and probabilistic forecast application. Weather forecast accuracy can be significantly improved using multi-model ensembles. Probabilistic forecast output makes industrial application more efficient. Goal: give the "in-house" meteorologist methods for constructing and communicating probabilistic temperature forecasts via multi-model ensembles.

Event, Hazard, and Industry Modeling

Private sector science teams often utilize vendor modeling services and/or models from the government sector. SPHEAR's academic partners pioneered the underlying research behind many of these models. Vendor models are often fully proprietary closed systems and the user has little ability to observe or adjust the internal workings. Government models are often inflexible with no ability for user customization. SPHEAR proposes the development of "open source" models and modules to specifically address industry needs such as:

• Catastrophic Events Leading to Financial Loss. Develop synthetic catalogs of physical hazards; engineering and damage functions; and financial loss components in an "open source" environment so as to be fully customizable by the user. Goal: provide industry a transparent, peer-reviewed, alternative to vendor catastrophe models.
• Coupled Ocean/Atmosphere Model for Weekly to Seasonal Climate Prediction. Develop research models that emulate and extend the current NCEP Coupled Forecast System. Goal: provide additional tools and data toward seasonal and intraseasonal weather prediction.
• Model Business Impact from Weather Events. Direct relationships exist between weather phenomena (such as temperature) and industrial applications (such as energy demand). Goal: provide models that convert primary weather inputs (such as rainy days in summertime) to important business variables (such as amusement park ticket sales).

Quantifying Extreme Events

How likely is a deadly heat wave this summer? How often do catastrophic floods affect densely populated areas? What's the risk of extreme temperatures in any given winter? What are the odds that multiple severe hurricanes will move through crucial energy-production regions in a given year? These are all regular questions from business leaders to their staff. Yet methodologies for constructing a robust climatology for such extreme events are relatively unknown in the private sector. Further, the question as to how oceanic and atmospheric cycles or climate change itself might impact such events is unexplained. SPHEAR proposes to quantitatively address issues such as:

• Temperature and Precipitation Extremes. Quantify the probability that certain thresholds of extreme heat, extreme cold, or extreme precipitation would be exceeded in any given year for specified regions around the world. Further determine the oceanic and/or atmospheric variables that condition the base-line probability for such events. Goal: provide tools for base-line risk assessment and real-time conditional probability of extreme events.
• Event "Clustering" in Natural Hazards and Extreme Events. Determine whether winter storms, hurricanes, severe thunderstorms, and/or earthquakes demonstrate grouping or clustering characteristics. Goal: develop methodologies for simulating such events in a risk modeling framework.
• Seasonal and Intraseasonal Hurricane Risk. Develop peer-reviewed and scientifically accepted methodology for determining seasonal and intraseasonal tropical cyclone risk. Goal: provide real-time tropical cyclone risk assessment.

Questions?

Contact the executive director of SPHEAR at sphear@sio.ucsd.edu