Using ASCAT scatterometer winds to evaluate relative biases in the QuikSCAT-derived wind vorticity
by S. Zecchetto and C. Accadia
Abstract
Relative wind vorticity (hereafter vorticity) is a crucial parameter to understand the spatial features of the wind field.
In the Mediterranean Sea, which is the area where this study is focused, these are particularly interesting because they are
often the effects of the interaction between the air flow and the orography. Wind vorticity has been derived both from QuikSCAT and ASCAT 12.5 km scatterometer data obtained from the NASA PODAAC data archive, and compared in the period
from March 2009 to November 2009. The monthly mean fields of the vorticity show discrepancies which need to be understood.
This work thus aims to understand the nature of these differences, to make the two vorticity data sets compatible and usable
as a common data set.
Results have been provided in terms of the relative bias in vorticity, which is the mean difference between the ASCAT
and QuikSCAT monthly mean vorticity averaged over the entire Mediterranean basin and the entire study period.
This difference (0.093 10^-5 1/s) is mainly due to a relative vorticity bias in the cyclonic component of vorticity,
rather than in the anticyclonic component, whose bias is four times smaller.
This bias does not depend significantly on
the variable accuracy of the wind speed and direction across the QuikSCAT swath.
The study led us to define and analyze
the so called vorticity noise, present especially in the QuikSCAT-derived vorticity, to understand if, and how, it can
contribute to the relative bias in vorticity. The contribution of this kind of noise on the mean vorticity has been found
relevant only for the QuikSCAT cyclonic vorticity. By applying a cyclonic de-noising to each swath of QuikSCAT,
the bias of -0.016 10^-5 1/s is obtained, drastically reduced with respect to the initial value.
This may be considered the typical bias over the Mediterranean Sea between the ASCAT and the QuikSCAT vortivity derived from the 12.5 km data, after applying the cyclonic de-noising to QuikSCAT vorticity fields.