Producing a Synthesis Forecast and Uncertainty Estimation From Multiple, Disparate Forecast Products

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Wednesday, 7 January 2015: 9:30 AM
123 (Phoenix Convention Center - West and North Buildings)
Matthew S. Wandishin, NOAA/ESRL/Global Systems Division/CIRES/Univ. of Colorado, Boulder, CO; and G. Layne, M. A. Petty, and B. J. Etherton

The Integrated Support for Impacted air-Traffic Environments (INSITE) is a web-based application that provides visualization and interrogation capabilities of potential weather-related impacts to en-route air traffic. A key component of INSITE is the synthesis product: post-processing of disparate (human-generated categorical, model-based probabilities, high-resolution deterministic) forecasts to produce a single picture displaying a blended common variable that measures weather-caused constraint to the flow of commercial aviation.

The synthesis production includes the following steps:

1. Weather forecasts are translated into constraint fields, calibrated, and combined with air-traffic information

2. Confidence measures are derived from historical performance in the neighborhood of the forecast point

3. Individual constraint fields are blended to produce the forecast synthesis

4. A consistency measure is produced based on the similarity between the synthesis and the component forecasts.

This presentation will detail the steps listed above, such as the construction of forecast and observation distributions from the neighborhood constraint values, the use of forecast confidence information to perform a weighted blending of the individual forecast distributions, and the implicit accounting for weather-feature orientation in the calculation of forecast confidence and consistency.