Minor tweaks to configurations

* drop 16 km and add 1 km to ensemble.  16 km appears too coarse to be
  very useful, especially for halfar, where it makes a dome of only 7 grid cells
* adjust halfar target_velocity accordingly (accounting for diffusive cfl)
* move start time of year 1 to the test group level so it applies to both tests

Author: matthewhoffman

compass/landice/tests/mesh_convergence/halfar/halfar.cfg

@@ -39,7 +39,7 @@ nonperiodic = True
 # and an advective CFL condition.
 # Note that Halfar is subject to a more restrictive diffusive CFL so this must be
 # artificially inflated
-target_velocity = 3000.0
+target_velocity = 30000.0
 
 # the duration (years) of the run
 duration = 200

compass/landice/tests/mesh_convergence/halfar/namelist.forward

@@ -1,4 +1,3 @@
 config_velocity_solver = 'sia'
 config_thickness_advection = 'fo'
 config_tracer_advection = 'fo'
-config_start_time = '0001-01-01_00:00:00'

compass/landice/tests/mesh_convergence/mesh_convergence.cfg

@@ -2,7 +2,7 @@
 [mesh_convergence]
 
 # a list of resolutions (km) to test
-resolutions = 16, 8, 4, 2
+resolutions = 8, 4, 2, 1
 
 # number of mesh cells in x and y for 1 km resolution.  Other resolutions have
 # the same physical size.  The default is approximately square, because of the

compass/landice/tests/mesh_convergence/namelist.forward

@@ -2,3 +2,4 @@ config_velocity_solver = 'none'
 config_unrealistic_velocity = 1.0e36
 config_thickness_advection = 'fo'
 config_tracer_advection = 'fo'
+config_start_time = '0001-01-01_00:00:00'