the sign of wind stress is opposite to the surface wind in CESM1-CAM5

I have run the coupled CESM1-CAM5 model, and checked the model output. I plotted the mean-state wind stress based on the TAUX and TAUY, but it displays an opposite pattern to the surface wind. In the tropical Pacific, it displays northwestern (southeastern) wind south (north) of the equator. Obviously, it will be right if it is multipied by (-1). I have searched in the internet but cann't find the result. Does anyone know the reason.
 
Back
Top