Skill of an aircraft wake vortex transport and decay model using weather prediction and observation

In order to safely reduce aircraft separation during approach and landing, the wake vortex behavior along the glide path has to be known and predicted. This requires in principle the knowledge of wake vortex relevant meteorological parameters along the entire glide path where typically continuous meteorological measurements of all relevant variables are not feasible. For this purpose a one-year meteorological data base for the Frankfurt Terminal area has been generated using NOWVIV. The skill of the nowcasting system NOWVIV (nowcasting wake vortex impact variables) [1] to predict these environmental parameters is assessed. The core of NOWVIV is the mesoscale model MM5 where a Yamada & Mellor 2.5 level turbulence closure scheme is employed from which TKE is computed as a prognostic variable. Two nested domains with sizes of about 250x250 km^2 and about 90x90 km^2 centered on Frankfurt airport with grid distances of 6.3 km and 2.1 km, respectively, were used. The model employs 60 vertical levels such that in the altitude range of interest (z < 1100 m above ground) 26 levels yield a vertical resolution varying between 8 m and 50 m. Initial and boundary data were taken from the numerical data assimilation model LM (Local Model) of DWD (German Weather Service). This one-year data set comprises typical weather conditions and includes already typical features of a long-term surface wind climatology. It enables to test new operational concepts with realistic meteorological input and to estimate the potential for separation reduction. It may also be used within risk assessments for prototype wake-vortex advisory systems [1].

A subset of the one-year data base has been analysed in detail for a period of 40 days where a dedicated wake measurement campaign was carried out at Frankfurt airport in fall 2004. In total 231 wake vortex pairs from heavy aircraft in ground proximity were tracked and characterized by LIDAR. During this measurement campaign a SODAR/RASS and a LIDAR provided profile measurements of meteorological variables. These data are used to analyse the quality of the predicted profiles of wind, temperature and turbulence. Furthermore we carried out a skill analysis to investigate the potential of NOWVIV as a real-time predicion system where we focus on the ability to predict pre-defined cross wind thresholds. The skill of NOWVIV is compared to the skill of a simple cross wind persistence model based on SODAR measurements. Initially we assume that a given measured cross wind profile is valid over the whole forecast lead time and evaluate the skill of the forecast with increasing lead time every 10 minutes. Without further optimization of this forecast approach the results for the false alarm rate indicate on average a better skill of the short-term prediction system after one hour lead time compared to the persistence approach for a cross wind threshold of 2 m/s. Other cross wind thresholds yield similar results.

In a wake-vortex advisory system, the weather prediction and observation system is coupled to a wake vortex predictor. Therefore, we have to know how the predictive skill of the weather forcast system influences the predictive skill of the whole forecast system including the wake predictor. Consequently, we extend this analysis by coupling the Probabilistic Two-Phase wake vortex transport and decay model P2P [2] to the NOWVIV system and investigate the predictive skill of P2P for pre-selected confidence levels of vortex position and strength. The skill is assessed by comparing predicted wake vortex position and strength against Lidar measured wake observations. We focus here on the frequency of nonconservative predictions (NoCoPs), which refer to a situation where a wake vortex is predicted to be outside a predefined safety corridor while observations still indicate the presence of a wake vortex in the safety corridor. A NoCoP refers to a potential risk for a following aircraft which has to be avoided. In this study we choose a single runway approach scenario. The analysis is used to optimize the forcasting system by selecting proper safety allowances of the weather input in order to adjust the number of NoCoPs to an acceptable frequency. After optimizing the system for situations with wake measurements the overall potential effect on capacity is estimated for a 40 day period with respect to a ICAO baseline scenario.

References

[1] Gerz, T., Holzäpfel, F., Bryant, W., Köpp, F., Frech, M., Tafferner, A., and Winckelmans, G., "Research towards a wake-vortex advisory system for optimal aircraft spacing,'' Institut für Physik der Atmosphäre, DLR, Report No. 206, Oberpfaffenhofen, March 2005, to appear in Comptes Rendus Physique, Academie des Sciences Paris.

[2] Holzäpfel F., "Probabilistic Two-Phase Wake Vortex Decay and Transport Model,'' Journal of Aircraft, Vol. 40, No. 2, 2003, pp. 323-331.