A study on kinematic trajectory calculations with different schemes using wind field of varying temporal frequency

Abstract

Computation of kinematic trajectories has got immense importance in assessing transport pathways of tracers and evaluating airflow patterns within meteorological systems. In a Lagrangian description of the transport of air pollutants, the kinematic trajectory of an air parcel is estimated using observed wind data or the output of a Numerical Weather Prediction (NWP) model. In this study, a non-hydrostatic mesoscale model is used to generate the meteorological data for a period, during which an active Western Disturbance (WD) crossed the Indian region. The 'true atmospheric trajectories (forward)' are constructed using model-simulated wind with 12 minutes time interval. The true trajectories are then compared to the trajectories that were constructed using simulated wind data with degraded temporal frequency of 36 min. and 1, 2, 3 and 6 hour. Five schemes including the well-known Petterssen scheme are used to calculate 72-hour forward trajectories. Comparison of the results using different schemes show that in terms of accuracy and computational efficiency, the midpoint method of integration (scheme-4) is comparatively better and the simplified Innocentini's scheme, which accounts for the spatial dependence of the wind field all along the particle path, gives results similar to the Petterssen scheme. Further, it is shown that when wind data are of low temporal frequency, use of smaller time steps will not improve the results.