MCPL: Future Plans and Ongoing Extensions

<Under Construction !!>
Feedback appreciated -- CJC

All of the following have been implemented and/or are currently undergoing beta-test.

Add extra EPA CMAQ output variables, as required for the CMAQ air quality model: these include surface and 3-D dot point Jacobians (redundantly, in both the time-independent (GRID_[CRO|BDY]_[2|3]D and the time-stepped MET_[CRO|BDY]_[2|3]D files), and Jacobian-coupled air and water-vapor densities, Jacobian-coupled entropy and Jacobian-coupled contravariant wind components in the MET_[CRO|BDY]_[2|3]D and MET_DOT_3D files, respectively.
For nonhydrostatic versions of MM5, only.

Coding and alpha-test in progress.

Add CCM2 RTTEN temperature tendency due to radiation (requires IFRAD = 3).
Coding complete; beta-test in progress.

Add Ohio State LSM (ISOIL = 2) and Pleim-Xiu LSM/PBL (ISOIL = 3, IBLTYP(<nest> = 7)) related variables in MET_CRO_2D.
For MM5V3 only, and turned on only if OSU or P-X is active.

Undergoing beta-test.

Add observation height variables TA2, QV2, UX10, VX10, and WSPD10, in MET_CRO_2D.
For MM5V3 only, and turned on only for MRF PBL: It turns out that these are very useful for comparing with observational meteorology, as well as for coupling with models normally driven by observations. It is also best done within MM5 PBL schemes, where it can be done consistently with the rest of the physics. Implementing this, however, involves PBL-scheme modifications for the user who wishes to take advantage of it.

Undergoing beta-test. Contact the authors for information about modifications to HIRPBL and GSPBL that generate these variables within those PBL schemes.

Remove variable RLD from MET_CRO_2D.
It turns out that the situation for longwave radiation in MM5 is more complex than originally thought, including the fact that the meaning of GLW depends upon the physics options chosen. RLD is not what we thought it was, and it looks quite complex to decide just what it should be, and then to implement that.

COMPLETED (i.e., RLD deleted as an output variable).

SLPRES option in file MET_CRO_2D.
Benjamin-Miller (1990) algorithm for calculating sea level pressure option, activated by defining preprocessor symbol MCPL_BMP, which then uses new routine MM2BMP for the calculation.

Coding COMPLETED. Undergoing test at this time.

NOTE as of January 2001: Results from it do not look as good as we'd like, at least running over the Eastern US at 15KM resolution. BM will be an option but not one that we recommend nor use as our default.

Wind component variables UX, VX at cross points in files MET_CRO_2D (layer-1 extraction for that file) and MET_CRO_3D (fully 3-D), using new routine MMXUSDV.

Coding COMPLETED. Undergoing test at this time.

More Detailed KF output
Since precipitation and updraft/downdraft behavior turn out to be exceedingly sensitive to the numerics used, the KF event files now export the KFPARA-internal versions of layer thickness DZQ and pressure thickness DP0 for the model layers at cloud inception.


Task-Parallel MCPL()
For MM5V2.11 and later (which implement OpenMP parallelism) make separate OpenMP tasks for each output file in MCPL_OUT(). This will allow greater efficiency on moderately-parallel platforms.

Coding COMPLETED and VALIDATED. On SGI's, requires obscure run-time flags that set the per-thread stack-size and memory limits or else it will seg-fault.

New Variables SLPRS, DZ1000_850, DZ1000_500, DZ850_700 in file MET_CRO_2D.
Compute sea level pressure and thicknesses between the 1000mB and 850mB, 1000mB and 500mB, and 850mB and 700mB pressure surfaces, and put them out as new variables in file MET_CRO_2D.


New file PRES_CRO_3D
contains 3D decoupled cross-point variables, interpolated to the standard pressure surfaces. Note that this file uses a pressure vertical coordinate instead of a sigma coordinate.

Output variables for this file are controlled by PC3_VARS


New file PRES_DOT_3D
contains 3D decoupled dot-point variables, interpolated to the standard pressure surfaces. Note that this file uses a pressure vertical coordinate instead of a sigma coordinate; decoupling of the winds requires that the output window is a proper window of the MM5 domain (e.g., there is no value for PSTAR available at MM5 dot-point (1,1) ).

Output variables for this file are controlled by PD3_VARS.


More general output-window structure
supporting multiple output windows per M5 domain. This will make MM5 more useful to the environmental modeling community in general (for coupled models work) at the cost of some additional complexity for run script configuration.
COMPLETED in version "C" and later, of MCPL, controlled by environment variable EM_OUTGRIDS which is a list of how many output grids there are per MM5 domain, so that setenv EM_OUTGRIDS 1,2,2,2,3,3,3,3,3,3 says that MM5 domain 1 has 1 output grid, MM5 domain 2 has 3 output grids, and MM5 domain 3 has 6 output grids; EM_I0 (etc.) will then be expected to be a list with 1+3+7=10 entries.


Support for KF cloud event files
via new ENTRY MCPL_KF to be called from MM5 module KFPARA.
COMPLETED in version "KF" and later, of MCPL. Requires MM5V2.11 or later if MM5 is to be run in parallel (in order for "critical sections" protecting the (serial) I/O to the KF-files.


Back to MCPL() Contents