There are 63 tests for Cruv7 forcing with Clm50 compsets, we should shift these to use Gswp3 as we don't need to test for all the possible DATM forcing types. We should have a few tests that include all the DATM forcing types, but don't need an extensive number of them.
We should cover matrixcn tests with both Cruv7 and other types of forcing. So we can keep those and we need the datm_ias_correct_cruv7 tests as well.
I think it should be reduced to these
ERI_D_Ld9.ne30_g17.I2000Clm50BgcCru*clm-vrtlay
ERI_D_Ld9.ne30_g17.I2000Clm50BgcCru*clm-vrtlay--clm-matrixcnOn
ERI_Ld9.f45_g37.I2000Clm50BgcCru*clm-nofire
ERI_Ld9.f45_g37.I2000Clm50BgcCru*clm-nofire--clm-matrixcnOn
ERP_D_Ld5.f10_f10_mg37.I2000Clm50BgcCru*clm-anoxia
ERP_D_Ld5.f10_f10_mg37.I2000Clm50BgcCru*clm-anoxia--clm-matrixcnOn
ERP_D_Ld5.f10_f10_mg37.I2000Clm50BgcCru*clm-ciso_flexCN_FUN
ERP_D_Ld5.f10_f10_mg37.I2000Clm50BgcCru*clm-ciso_flexCN_FUN--clm-matrixcnOn
ERP_D_P64x2_Ld3.f10_f10_mg37.I2000Clm50BgcCru*clm-flexCN_FUN
ERP_D_P64x2_Ld3.f10_f10_mg37.I2000Clm50BgcCru*clm-flexCN_FUN--clm-matrixcnOn_ignore_warnings
ERP_D_P64x2_Ld3.f10_f10_mg37.I2000Clm50BgcCru*clm-noFUN_flexCN
ERP_D_P64x2_Ld3.f10_f10_mg37.I2000Clm50BgcCru*clm-noFUN_flexCN--clm-matrixcnOn_ignore_warnings
ERP_D_P48x1.f10_f10_mg37.I2000Clm50BgcCru*clm-flexCNFUN
ERP_D_P48x1.f10_f10_mg37.I2000Clm50BgcCru*clm-flexCNFUN--clm-matrixcnOn
ERP_D_P48x1.f10_f10_mg37.I2000Clm50BgcCru*clm-noFUN_flexCN
ERP_D_P48x1.f10_f10_mg37.I2000Clm50BgcCru*clm-noFUN_flexCN--clm-matrixcnOn
SMS_D_Ld1_P48x1.f10_f10_mg37.I2000Clm50BgcCru.izumi_nag.clm-datm_bias_correct_cruv7
SMS_Ld1_PS.f09_g17.I2000Clm50BgcCru.derecho_gnu.clm-datm_bias_correct_cruv7
SMS_Ld1_PS.f09_g17.I2000Clm50BgcCru.derecho_gnu.clm-datm_bias_correct_cruv7
Then we can remove all but the I2000Clm50BgcCru compset for Cru forcing. This will likely shorten the testlist as there likely are duplicates for compsets with Gswp3
There are 63 tests for Cruv7 forcing with Clm50 compsets, we should shift these to use Gswp3 as we don't need to test for all the possible DATM forcing types. We should have a few tests that include all the DATM forcing types, but don't need an extensive number of them.
We should cover matrixcn tests with both Cruv7 and other types of forcing. So we can keep those and we need the datm_ias_correct_cruv7 tests as well.
I think it should be reduced to these
Then we can remove all but the I2000Clm50BgcCru compset for Cru forcing. This will likely shorten the testlist as there likely are duplicates for compsets with Gswp3