Possible typos

[hjsong]

Hi,

I found some possible typos in Adjoint/ad_u3dbc_im.F and Tangent/tl_u3dbc_im.F

First, in Adjoint/ad_u3dbc_im.F, I suspect line 166, 168, 292, 294, 420 and 422.
For example in line 166, it is now

Code: Select all

tau=M3obc_in(inorth)

And I think it should be

Code: Select all

tau=M3obc_in(ng,inorth)

Same chages can be applied in other lines..

And almost same suspicious typos are in Tangent/tl_u3dbc_im.F.
They are in line 225, 227, 344, 346, 475 and 477.
I think 'ng' is missing in those lines as above.

Another suspicious typo in Tangent/tl_u3dbc_im.F is in line 99.
I guess

Code: Select all

DO k=1,N(ng)(ng)

should be changed like

Code: Select all

DO k=1,N(ng)

Thank you.

[arango]

Yes, good catch. Radiation conditions are very tricky in the adjoint because the radiation speeds need to be saved at every time-step for the basic state nonlinear solution. These boundary conditions were coded but not tested. We still need to add the appropriate code to process the in/out information forward and backwards.

Be aware that radiation conditions are not well possed in the nonlinear model. This is even worse in the tangent linear and adjoint models. This perturbations can growth even faster and the models become unstable.