A Numerical Study into the Dynamic Behaviour of a GPS dropsonde in a Prescribed Wind Field

The GPS dropsonde has now been in use for over a decade, providing measurements of atmospheric variables, like the wind velocity, air temperature and humidity, with unprecedented resolution and accuracy in hurricanes. One major issue in the interpretation of these measurements, however, is that since the dropsonde measurements are made in neither an Eulerian framework nor a perfect Lagrangian framework, they need to be interpreted carefully to provide the correct information about the characteristics of the hurricane boundary layer in a conventional Eulerian framework.

In this study we numerically model the fall of a GPS dropsonde through a simulated boundary layer wind field with prescribed mean and turbulent statistical characteristics. The falling dropsondes are then tracked and the wind field characteristics are analyzed using the wind finding equations of Hock and Franklin (1999). Mean wind profiles and the turbulent characteristics of the wind field derived from the dropsonde motion are compared with the prescribed values of the simulated boundary layer. The results of the comparison show that velocity of the dropsonde by itself can not represent the true wind velocity with satisfactory accuracy, especially when the wind velocity changes dramatically with respect to the spatial coordinates, however the wind finding equations provided by Hock and Franklin (1999) are sufficient to derive an accurate mean wind profile from the dropsonde measurements. In addition, the turbulence intensity can also be correctly found with the help of the wind finding equations. The derived turbulent shear stress, on the other hand, deviated from the prescribed value appreciably because the vertical wind velocity cannot be found accurately enough. Moreover, the impact of low-pass filter and different schemes for the calculation of the dropsonde accelerations, which are required in the corrections to the mean dropsonde velocity provided by the wind finding equations of Hock and Franklin (1999) were also examined.