Performing Hindcasts to evaluate COSMO-Model 5.05

04.07.2017

Introduction

With COSMO-Model Version 5.05 all physics are now in the blocked data format. The last two packages included from ICON, which are activated now, are the turbulence and the surface schemes. First experiments show heterogeneous results, which might be due to the fact, that the surface scheme has significant differences to the scheme used up to now in COSMO. Therefore, four hindcasts will be done to see whether a consolidated soil can be obtained. The hindcasts will be done for the three month March-May 2016. Bacy experiments will be done afterwards.

News from the Expert Meeting on 03.07.2017:

It was decided to perform another hindcast, similar to TEST_NEW, but with a different setting of the namelist variable tur_len = 300.0.
The Hindcast-configurations OPER_NOW and TEST_NEW will be continued in BACY-Experiments and will use their own soil data from the hindcasts.
The Hindcast-configuration TEST_OLD will be tested in a NUMEX Experiment similar to EXP 10416. Initialization will be from the soil from EXP 10416.

Setup of the Hindcasts

The following hindcasts will be performed:

OPER_NOW: Using the operational binary (Version 5.04d4) with the operational configuration from today.
TEST_OLD: Using a modified binary from the new version 5.05 and a configuration which should give results close to the operational version 5.04d4. To obtain that, some hard-coded switches have to be reset.
REF_NEW: Using the binary for version 5.05 with a configuration tested in hindcasts and NUMEX 10441. This configuration should be similar to the ICON configuration for the physics but remains conservative for plant evaporation.
TEST_NEW: Using the binary for version 5.05 with a configuration intended for operational use in the future. This configuration should be similar to the ICON configuration for the physics.
TEST_l300m: Similar to TEST_NEW, but with tur_len = 300.0.

Data for the Hindcasts

Initial and boundary data for these hindcasts are computed by INT2LM (Version 2.04a).

/lustre2/uscratch/uschaett/2016030100/COSMO_DE_input_old/
The tests OPER_NOW and TEST_OLD can use the same initial and boundary data available in this directory.
/lustre2/uscratch/uschaett/2016030100/COSMO_DE_input_ref/
The test REF_NEW has to use extra initial and boundary data because of using additional external parameters for the Tegen climatology.
/lustre2/uscratch/uschaett/2016030100/COSMO_DE_input_new/
The test TEST_NEW has to use extra initial and boundary data because of using different external parameters for PLCOV, LAI, ROOTDP, stomata resistance and emissivity.

The following parameters are different between the "old" and the "new" data. The REF_NEW test takes either the "old" or the "new" approach. See the Namelist Input for INT2LM for more details:

Parameter	OLD	NEW	Controlled by
`PLCOV`	Plant cover and leaf area index for the COSMO-Model and for a special day are produced by using the data sets for vegetation and for rest and modify these with an annual cycle.	Plant cover and leaf area index for the COSMO-Model and for a special day are produced by using only the data set for vegetation and an averaged ndvi ratio. This ndvi ratio is computed as a weighted mean between monthly mean values, which are taken from the external parameter data set for the COSMO-Model.	`itype_ndvi`
`LAI`			`itype_ndvi`
`ROOTDP`	Input from external parameter for the COSMO-Model is taken and modified with an annual cycle.	Input from external parameter for the COSMO-Model is taken without modifications.	`itype_root`
`AER_SO4`	No additional Fields for aerosol types are read from the external parameters. The COSMO-Model has to run with constant values for the aerosol distribution on rural areas, urban areas, desert areas or the sea.	Additional external parameters for the aerosol types of sulfate, mineral dust, organic, black carbon and sea salt are read as monthly mean values. Actual values for the current day are computed and given to the COSMO-Model.	`itype_aerosol`
`AER_DUST`
`AER_ORG`
`AER_BC`
`AER_SS`
`RSMIN`	No external parameter is taken. A constant value for all grid points is determined in the COSMO-Model.	Take a map from the external parameters for the minimum stomata resistance of plants.	`lstomata`
`EMIS_RAD`	No external parameter is taken. A constant value for all grid points is determined in the COSMO-Model.	Take a map from the external parameters for the thermal radiative surface emissivity.	`lemiss`

Because the hindcasts will be run for three month, we have to use the climate mode of INT2LM and COSMO, to also update the above parameters during the simulation.

Namelist Input for INT2LM

/CONTRL/	OPER_NOW, TEST_OLD	REF_NEW	TEST_NEW	Explanation
lbdclim	.TRUE.	.TRUE.	.TRUE.	Use the climate mode because we run for 3 months
lsso	.TRUE.	.TRUE.	.TRUE.	`SSO_STDH` and `SSO_SIGMA` are used by new schemes
itype_albedo	3	3	3	-
itype_aerosol	1	2	2	2 activates the Tegen climatology
itype_ndvi	0	0	1	1 activates a yearly cycle for `PLCOV` and `LAI` based on an averaged ndvi ratio.
itype_rootdp	0	0	4	4 takes the input from the external data set without modifications. This is done in the COSMO-Model now.
lemiss	.FALSE.	.FALSE.	.TRUE.	take a map from the external parameters for the thermal radiative surface emissivity.
lstomata	.FALSE.	.FALSE.	.TRUE.	take a map from the external parameters for the minimum stomata resistance of plants.

Namelist Input for COSMO

In COSMO several Namelist variables have to be adapted to either reproduce the operational settings from Version 5.04d4 or to choose an ICON-like configuration for the new physics. In addition there are two namelist variable, which are not available in the new binary any more (because they are related to the old physics).

/TUNING/	OPER_NOW	TEST_OLD	REF_NEW	TEST_NEW	TEST_l300m	Explanation
tkhmin	0.4	0.4	0.75	0.75	0.75	-
tkmmin	0.4	0.4	0.75	0.75	0.75	-
rat_sea	20.0	20.0	7.5	7.5	7.5	-
pat_len	500.0	500.0	750.0	750.0	750.0	-
tur_len	150.0	150.0	500.0	500.0	300.0	-
a_hshr	0.2	0.2	2.0	2.0	2.0	-
c_soil	1.0	1.0	1.75	1.75	1.75	-
wichfakt	0.0	N.A.	N.A.	N.A.	N.A.	-
/PHYCTL/	OPER_NOW	TEST_OLD	REF_NEW	TEST_NEW	TEST_l300m	Explanation
itype_vdif	-2	-1	1	1	1	-
ltkeshs	.FALSE.	.FALSE.	.TRUE.	.TRUE.	.TRUE.	-
itype_sher	1	1	0	0	0	-
imode_tran	1	1	0	0	0	-
imode_turb	1	1	1	1	1	-
icldm_tran	0	0	2	2	2	-
lconf_avg	.TRUE.	.FALSE.	.FALSE.	.FALSE.	.FALSE.	-
itype_albedo	3	3	3	3	3	-
itype_aerosol	1	1	2	2	2	-
itype_root	1	1	1	2	2	-
itype_heatcond	1	1	3	3	3	-
itype_evsl	2	2	4	4	4	-
cwimax_ml	-	1.0E-6	0.0005	0.0005	0.0005	-
idiag_snowfrac	-	1	20	20	20	-
lemiss	.FALSE.	.FALSE.	.FALSE.	.TRUE.	.TRUE.	-
lstomata	.FALSE.	.FALSE.	.FALSE.	.TRUE.	.TRUE.	-
lmulti_layer	.TRUE.	N.A.	N.A.	N.A.	N.A.	-
/IOCTL/	OPER_NOW	TEST_OLD	REF_NEW	TEST_NEW	TEST_l300m	Explanation
lbdclim	.TRUE.	.TRUE.	.TRUE.	.TRUE.	.TRUE.	-

Binaries to use

Also for the operational version 5.04d4, an extra binary has to be used, because of necessary updates for processing all fields with lbdclim=.TRUE. in GRIB 2. The binaries can be found in /e/uhome/uschaett/cosmo/BINARIES.

OPER_NOW: tstlm_f90_5.04d4_opernow
TEST_OLD: tstlm_f90_5.05_testold
REF_NEW, TEST_NEW: tstlm_f90_5.05_testnew