PBLH devt3: Reconcile ROPP and TAMU-CC PBLH diagnostics

[Ian Culverwell]

The ROPP-9 beta reviewer noted a significant difference between the ROPP and TAMU-CC PBLH diagnostics in polar regions, especially Greenland and Antarctica. TAMU-CC generally produces PBLHs close to the minimum allowed value of 300 m, while ROPP generally returns much larger PBLHs. This may be due to different smoothing being applied in the two algorithms, or perhaps the interpolation of the profile data to higher (10 m) vertical resolution, as undertaken by the TAMU-CC algorithm, is playing a part. There is a suggestion that the ROPP algorithm, which uses a quadratic fit to the data around the maximum gradient to site the PBLH precisely, is placing it just below the 300 m threshold in situations where the gradients are largest at the bottom. This results in the PBLH being ruled invalid. The algorithm then investigates the next largest local maximum vertical gradient as a potential PBLH, and this is far up the profile in these situations. It is not clear that this is necessarily a bad thing, however, since a PBLH that is strongly tied to the arbitrary minimum possible value is clearly not a scientifically desirable feature, even if, fortuitously, the procedure gives accurate values in such circumstances.

The issue needs investigating.

[Ian Culverwell]

PBLH_antarctic_profiles.png

[Ian Culverwell]

Here, to illustrate the point, are 6 Antarctic T & N profiles, with the TAMU-CC and ROPP PBLHs overlain:

[Ian Culverwell]

Postpone to ROPP10.0.