Seasonal Sensitivity on COBEL-ISBA Local Forecast System for Fog and Low Clouds

Abstract

Skillful low visibility forecasts are essential for air-traffic managers to effectively regulate traffic and to optimize air-traffic control at international airports. For this purpose, the COBEL-ISBA local numerical forecast system has been implemented at Paris CDG international airport. This local approach is robust owing to the assimilation of detailed local observations. However, even with dedicated observations and initialization, uncertainties remain in both initial conditions and mesoscale forcings. The goal of the research presented here is to address the sensitivity of COBEL-ISBA forecast to initial conditions and mesoscale forcings during the winter season 2002-2003. The main sources of uncertainty of COBEL-ISBA input parameters have been estimated and the evaluation of parameter uncertainty on the forecasts has been studied. A budget strategy is applied during the winter season to quantify COBEL-ISBA sensitivity. This study is the first step toward building a local ensemble prediction system based on COBEL-ISBA. The conclusions of this work point out the potential for COBEL-ISBA ensemble forecasting and quantify sources of uncertainty that lead to dispersion. Accurate prediction of fog and low clouds is one of the main issues related to improving air-traffic management and safety. At Paris Charles de Gaulle (CdG) international airport, adverse ceiling and visibility conditions (visibility under 600 m and ceiling below 60 m) lead to the application of Low Visibility Procedures (LVP). Under these conditions, the airport take-off/landing efficiency is reduced by a factor of 2, causing aircraft delays. In this context, accurate short-term forecasts of LVP conditions are considered to be a priority by airport authorities. Unfortunately, current operational Numerical Weather Prediction (NWP) models are not able to provide detailed information due to their lack of both vertical and horizontal resolutions with respect to the typical length scale of fog. However, owing to higher vertical resolution and more detailed physics than 3D