On Wed, May 16, 2018 at 2:13 AM, Nanda Kishore Reddy <nandu.eng@xxxxxxxxx> wrote:
I am trying to calculate the brunt vaisala frequencyby using the pyferret. For this, I computed the density (ρ) with the help of temperature and salinity fields, but I am getting difficulties to calculate the dρ/dz. Is there any ferret function is available to calculate the density gradient (dρ/dz) in the ferret.
But, insitu density gradient doesn't give you a "correct" BVf. I don't know how much precision you need, but a quick check would be to compare BVfs computed from insitu density and from potential density. If the difference is significant to you, you need a more precise method. Ideally you use the locally-referenced potential density. But to do that, calculating some density and taking its derivative is not sufficient. (There should be some formula to calculate BVf with compressibility taken into account . . . )
Another potential issue is the vertical resolution of your temperature and salinity data. CTD profiles, for example, almost always include all sorts of vertical scales. So, BVfs inherently depend on the vertical-smoothing scale you choose. If the T-S data comes from a general circulation model, this problem would be much less severe below the mixed layer unless the model resolves internal waves.
Cheers,
Ryo