Opened 13 years ago
Last modified 12 years ago
#234 new defect
Discrepancy between minROPP and LevMarq minimisers — at Initial version
| Reported by: | Ian Culverwell | Owned by: | Ian Culverwell |
|---|---|---|---|
| Priority: | normal | Milestone: | Whenever |
| Component: | ropp_1dvar | Version: | 5.0 |
| Keywords: | minropp levmarq minimisation | Cc: |
Description
ROPP5 testing has shown large differences in retrievals using these two minimisers (eg see #232).
Huw Lewis documented similar differences (if not quite as bad) in GSR 6.
We should find out why minROPP and LevMarq differ.
Here are some preliminary thoughts:
Plots of the convergence of both schemes (in terms of J and dx, the normalised change in the state from one iteration to the next) are listed below, for the 8 profiles in IT-1DVAR-03, for reference with Huw's earlier work in GSR-03 and GSR-06
Refrac, J
Change history (9)
by , 13 years ago
| Attachment: | Test4a_refrac_J.gif added |
|---|
by , 13 years ago
| Attachment: | Test4a_refrac_dx.gif added |
|---|
by , 13 years ago
| Attachment: | Test4a_bangle_J.gif added |
|---|
by , 13 years ago
| Attachment: | Test4a_refrac_dx.2.gif added |
|---|
by , 13 years ago
| Attachment: | Test4_multi_1.gif added |
|---|
by , 13 years ago
| Attachment: | Test4_multi_2.gif added |
|---|
by , 13 years ago
| Attachment: | Test4_multi_3.gif added |
|---|
by , 13 years ago
| Attachment: | Test4_multi_4.gif added |
|---|
by , 13 years ago
| Attachment: | Test4a_bangle_dx.gif added |
|---|
Note:
See TracTickets
for help on using tickets.
