[Dart-dev] DART/branches Revision: 12966
dart at ucar.edu
dart at ucar.edu
Thu Jan 31 16:57:39 MST 2019
nancy at ucar.edu
2019-01-31 16:57:39 -0700 (Thu, 31 Jan 2019)
80
add the compute_posterior item to the default namelist
and add it to the docs.
Modified: DART/branches/recam/assimilation_code/modules/assimilation/filter_mod.html
===================================================================
--- DART/branches/recam/assimilation_code/modules/assimilation/filter_mod.html 2019-01-31 23:48:23 UTC (rev 12965)
+++ DART/branches/recam/assimilation_code/modules/assimilation/filter_mod.html 2019-01-31 23:57:39 UTC (rev 12966)
@@ -98,6 +98,7 @@
output_mean = .true.,
output_sd = .true.,
write_all_stages_at_end = .false.,
+ compute_posterior = .true.
ens_size = 20,
num_groups = 1,
@@ -292,6 +293,14 @@
runtime, but can lower the cost of the job significantly in some cases.
</TD></TR>
+<TR><TD>compute_posterior</TD>
+ <TD>logical</TD>
+ <TD>If .false., skip computing posterior forward operators and do not write
+ posterior values in the obs_seq.final file. Saves time and memory.
+ Cannot enable posterior inflation and skip computing the posteriors.
+ For backwards compatibility the default for this is .true.
+ </TD></TR>
+
<TR><TD>ens_size</TD>
<TD>integer</TD>
<TD>Size of ensemble.
Modified: DART/branches/recam/assimilation_code/modules/assimilation/filter_mod.nml
===================================================================
--- DART/branches/recam/assimilation_code/modules/assimilation/filter_mod.nml 2019-01-31 23:48:23 UTC (rev 12965)
+++ DART/branches/recam/assimilation_code/modules/assimilation/filter_mod.nml 2019-01-31 23:57:39 UTC (rev 12966)
@@ -18,6 +18,7 @@
output_mean = .true.
output_sd = .true.
write_all_stages_at_end = .false.
+ compute_posterior = .true.
ens_size = 20,
num_groups = 1,
More information about the Dart-dev
mailing list