Discussion about tangent linear and adjoint models, variational data assimilation, and other related issues.
Moderators: arango, robertson
- Posts: 6
- Joined: Sat Oct 18, 2014 4:11 pm
- Location: INCOIS, Hyderabad, INDIA
I am running W4DPSAS for the northern Indian Ocean with UKMO-EN4 obs. For a simulation of 5 days with 50 inner loops and 1 outer loop, I am getting the error:
Code: Select all
CONGRAD - Fatal error, not possitive definite algorithm:
cg_delta = -5.76565447E+09 (001, 022)
at the end of TLM for loop inner=22. Does this imply that the matrix corresponding to the conjugate gradient algorithm is not positive definite?! How can I solve this issue? This set-up has no balance operator.