Skip to main content
placeholder image

Quantification of WAFS weather forecast uncertainties and its effect on fuel burn using historical flight data

Conference Paper


Abstract


  • Weather conditions directly influence flight operations. Although daily weather forecast data has been employed when forming optimal flight trajectory and fuel requirements, no forecast is without error. In this regard, studies that sought to improve our understanding of weather uncertainties and their effect on trajectory and fuel burn prediction have shined a spotlight on future air traffic control, flight planning and fuel-saving strategies. Though such studies are not new, their findings are often difficult to apply to real-world airline operations. Most of these studies do not consider the fact that airlines, in practice, use lower resolution weather forecasts that are distributed from the World Area Forecast System (WAFS). In this paper, we introduce several data handling techniques to understand the cause of WAFS weather forecast error and quantify its effect on the predictions of cruise stage fuel burn. Using historical data, we merged the second-by-second recorded flight data with the corresponding WAFS weather forecast; we also calculated the deviation of different forecasted weather parameters to the realized weather conditions. Fuel flow deviations and overall cruise stage fuel burn deviations due to weather uncertainties were then modeled using the fuel consumption model. In summary, our analysis found that weather forecast errors increase with time elapsed from departure; we also found that weather forecast errors are route-specific. High variance in wind direction forecast error is found at low-latitude of the trans-Pacific routes and at high-latitude of the southern-hemispheric routes. Furthermore, an overestimation in forecast temperature is found in two southern-hemispheric routes. Based on the comparison between the performance under the forecasted weather and actual weather, the southern-hemispheric routes tend to overestimate fuel consumption with a median of up to 223 kg. An underestimation in temperature along with an underprediction in cruise stage fuel consumption with a median up to 202 kg is found for trans-Pacific and Asia-pacific routes.

UOW Authors


  •   Tang, Agnes (external author)
  •   Wu, Cheng-Lung (external author)
  •   Tan, David

Publication Date


  • 2018

Citation


  • Tang, A., Wu, C. & Tan, D. (2018). Quantification of WAFS weather forecast uncertainties and its effect on fuel burn using historical flight data. International Conference on Research in Air Transportation (pp. 1-8). ICRAT.

Start Page


  • 1

End Page


  • 8

Abstract


  • Weather conditions directly influence flight operations. Although daily weather forecast data has been employed when forming optimal flight trajectory and fuel requirements, no forecast is without error. In this regard, studies that sought to improve our understanding of weather uncertainties and their effect on trajectory and fuel burn prediction have shined a spotlight on future air traffic control, flight planning and fuel-saving strategies. Though such studies are not new, their findings are often difficult to apply to real-world airline operations. Most of these studies do not consider the fact that airlines, in practice, use lower resolution weather forecasts that are distributed from the World Area Forecast System (WAFS). In this paper, we introduce several data handling techniques to understand the cause of WAFS weather forecast error and quantify its effect on the predictions of cruise stage fuel burn. Using historical data, we merged the second-by-second recorded flight data with the corresponding WAFS weather forecast; we also calculated the deviation of different forecasted weather parameters to the realized weather conditions. Fuel flow deviations and overall cruise stage fuel burn deviations due to weather uncertainties were then modeled using the fuel consumption model. In summary, our analysis found that weather forecast errors increase with time elapsed from departure; we also found that weather forecast errors are route-specific. High variance in wind direction forecast error is found at low-latitude of the trans-Pacific routes and at high-latitude of the southern-hemispheric routes. Furthermore, an overestimation in forecast temperature is found in two southern-hemispheric routes. Based on the comparison between the performance under the forecasted weather and actual weather, the southern-hemispheric routes tend to overestimate fuel consumption with a median of up to 223 kg. An underestimation in temperature along with an underprediction in cruise stage fuel consumption with a median up to 202 kg is found for trans-Pacific and Asia-pacific routes.

UOW Authors


  •   Tang, Agnes (external author)
  •   Wu, Cheng-Lung (external author)
  •   Tan, David

Publication Date


  • 2018

Citation


  • Tang, A., Wu, C. & Tan, D. (2018). Quantification of WAFS weather forecast uncertainties and its effect on fuel burn using historical flight data. International Conference on Research in Air Transportation (pp. 1-8). ICRAT.

Start Page


  • 1

End Page


  • 8