Documentation of the Changes in the COSMO-Model
Version 5.09b

Contents:

1. Fix in TERRA for computation of subsurface runoff
2. Added a new diagnostic for the soil water budget
3. Some fixes for the diagnostic variables related to itype_hydmod=1
4. Technical Changes and Bug Fixes
5. Changes to the Namelists
6. Changes of Results

1. Fix in TERRA for computation of subsurface runoff

(by Ronny Petrik)

For the old hydrology scheme (itype_hydmod = 0) the calculation of the subsurface runoff has been improved, in order to better close the soil water budget.

For itype_hydmod = 1 a zero-flux condition is applied at the lower boundary of the hydrological active layers. No fix is required for that.

The changes have no impact on any prognostic variable.

The debug output has been modified to check the soil water budget at a specific grid point (first meteograph point).

Back to Contents

2. Added a new diagnostic for the soil water budget

(by Ronny Petrik)

Introduced a new field res_soilwatb for the soil water budget (residuum).

This field can be interpreted as an aggregation of all hydrological active soil layers. Therefore it is coded in GRIB as a "depthBelowLandLayer" with upper level 0 and lower level ibot_w_so (number of hydrological active soil layers; in TERRA called ke_soil_hy).

A logical switch in data_runcontrol (lres_soilwatb) is set to .TRUE., if RESID_WSO appears in an output list. Only then the necessary computations are done in sfc_terra.f90.

Back to Contents

3. Some fixes for the diagnostic variables related to itype_hydmod=1

(by Ulrich Schättler)

There are problems with these variables (wt_depth, wso_satlev, q_roff, wsoil_flx): (Modifications in sfc_terra.f90, sfc_interface.f90, src_output.f90)

• q_roff, wsoil_flx have a dimension 1:ke_soil+1, but are only computed for the hydrological active layers (ke_soil_hy=ibot_w_so). The loop, where these variables are set has been adapted accordingly.
• Also, for special soiltyps, the values for these diagnostic variables are not computed at all, leading to NaNs at some grid points. Therefore we set them to 0.0 in these cases.
• Because these variables are only computed over land grid points, the values over water remain with rinit_val=-99999999.0_wp, leading to problems for GRIB output. Therefore we included these values also in the copyToBlock, to give them meaningful initial values.
• Moreover, the output for WSOIL_FLX and Q_ROFF is only done for the hydrological active layers (ibot_w_so).
• The new diagnostic fields are only written for itype_hydmod==1

Back to Contents

4. Technical Changes and Bug Fixes

NONE

Back to Contents

5. Changes to the Namelists

NONE

Back to Contents

6. Changes of Results

NONE

Back to Contents