|
||||||||||||
Computational error was not a significant source of trajectory error. However, spatial and temporal resolution of the meteorological data may be important. We can examine the effect of meteorological data resolution by re-running the forward calculation from 600 m but using the North American Regional Reanalysis (NARR - 32 km every 3 h) instead of the WRF27UW (27 km every 1 h). Although the spatial resolution is similar, the WRF data have a more frequent temporal interval. In addition, the NARR data provided initial and boundary conditions for the WRF calculations, so one would not expect large differences between these two data sets.
A trajectory represents the integration of the position vector in space and time through the time- and space-varying wind field. In the atmosphere, the wind direction and speed is a continuous field, but the numerical representation used in these calculations is discreet in space and time. How well the gridded meteorological data represents the real atmospheric conditions can vary from day-to-day and depends upon the location of regions with sharp gradients or cols with respect to the grid points. This is probably the most important source of trajectory error. 1 s |