UPP Users Page

Unified Post Processor Software Package

The Unified Post Processor Version 3.0 was released on May 5, 2015.

UPP Version 3.0

Issues with run_unipost_minutes

Problem: Typos in run_unipost_minutes script cause failure when using script due to incorrect valid time field.

Solution: Use the updated script here: run_unipost_minutes

Incorrect map data output in grib2 files for Lambert

Problem: Latitude, longitude, dx, dy, were being output in the grib2 files incorrectly resulting in projection errors.

Solution: Download the new grib2_module.f to UPPV3.0/src/unipost, rename it to grib2_module.f. Clean (./clean) and recompile (./compile) under UPPV3.0/. This fix pertians to grib2 output format only and specifically for Lambert projections.

Fix for Floating Point Exception in CALMICT (Ferrier Phyiscs)

Problem: Float point exceptions can occur when using Ferrier based physics due to diving by zeros.

Solution: Download CALMICT.f.fix to UPPV3.0/src/unipost, rename it to CALMICT.f. Clean (./clean) and recompile (./compile) under UPPV3.0/

Updated xml files for grib2

Updated postcntrl.xml and post_avblflds.xml files to include more examples and fix a few syntax errors. Note that grib2 functionallity is still limited and recommend for model output that does not require destaggering or regridding, e.g. wrf-arw, since a grib2 tool parallel to copygb is not yet supported for grib2.
postcntrl.xml
post_avblflds.xml

Runtime error with grib2 output using Gfortran

Problem: Code compiled with gfortran may experience a failure while reading in the xml files for grib2 production. Runtime error will appear in unipost_d01.00.out log file as "STOP" after in "readxml,bf readxml,size" calls. gfortran returns a "1" from the size function when it should be "0" as in intel, pgi.

Solution: A workaround is to modify the two lines in UPPV3.0/src/unipost/READ_xml.f file that use the size function so that it looks for return values <=1, instead of just ==0:

(size(post_avblflds%param)<=1)

(size(paramset)<=1)

Recompile the code after making the modification and re-run.

* This is not considered a permanent fix but rather a workaround at this time while futher development is happening.

Modified effective radius for cloud ice for Ferrier and Ferrier-Aligo microphysics

Problem: Cold bias detected when computing synthetic satellite products for Ferrier and Ferrier-Aligo microphyscs. the output field.

Solution: Testing of new effective radius improved synthetic satellite results. This modification is recommended in particular for HWRF users and those using synthetic satellite products. In the file UPPV3.0/src/unipost/CALRAD_WCLOUD_newcrtm.f, change the effective radius from 25 to 75 on lines 778 and 1334. The lines should read:

atmosphere(1)%cloud(2)%effective_radius(k) = 75.

Clean and recompile the code.

Composite reflectivity for ARW being set to zeros

Problem: The array being filled with ARW composite reflectivity is being reset to zero resulting in zeros in the output field.

Solution: Download INITPOST.F.arw-comp-refl-fix to UPPV3.0/src/unipost, rename it to INITPOST.F and recompile (./compile) under UPPV3.0/

Grib2 output not working for all models or cases

Problem: Grib2 functionality has not yet been fully tested and extended to all models and cases. Additional enhancements for further flexibility are underway at NCEP and will be provided as they become available. At this time, it is recommended to only attempt grib2 output with ARW when not utilizing copygb.

Solution: Check back for additional updates.

UPPV3.0 depends on WRFV3.7 if using Ferrier microphysics

Problem: If using UPPV3.0 with model output from a model version of WRF less than v3.7, there will be a mismatch in the microphysics number option for Ferrier physics. This is because UPP supports the new Ferrier-Aligo microphysics and this capability is not yet in WRF.

Solution: If using model output from WRFV3.7 or above, no change required. If using model output from WRFV3.6.1 or below, edits to the metadata in your wrfouts* must be made to change the microphysics option from mp_physics=5 to mp_physics=95. For example, one could use NCO tool ncatted:

ncatted -O -a MP_PHYSICS,global,o,c,95 wrfout_d02_2010-06-25_12:00:00

Projection Issues in Southern Hemisphere

Problem: Errors for some grid projections in Southern Hemisphere. Work is underway to consolidate various fixes.

Solution: In the meantime, fix files for known problems (Lambert and Mercator projections) are provided here. w3fi74.f.SP-fix.

Radar Reflectivity with WRF or NMMB output and Thompson microphysics requires radar reflectivity output from model.

Problem: Radar reflectivity diagnostics can now be output directly by the model (WRF/NMMB) for certain microphysics. Currently UPP is hardcoded to either use the model output directly or compute reflectivity fields internally. When using Thompson microphysics in either WRF or NMMB, UPP is currently hardcoded to use model-derived reflectivity. Thus, when requesting fields such as radar on model levels, composite reflectivity, reflectivity 1km AGL, and echo tops when running Thompson microphysics, these fields must be writted to the model output files; UPP will not compute it internally.

Solution: Work is underway to provide more user control to select where radar reflectivity diagnostics are coming from for each microphysics, i.e. direct from model output (WRF/ NMMB) or calculated within UPP code.

Previous release notes

UPP v2.2

UPP v2.1

UPP v2.0

UPP v1.1

UPP v1.0

UPP beta v0.5