The Polar WRF
Last modified on 2019-07-03
Statistics on Registered PWRF (Latest update 2018/01/22)
- 351 registered users of Polar WRF
- 262 International users
- 38 countries including USA
- 89 registered US users
Based on extensive experience with mesoscale modeling in the polar regions by the Polar Meteorology Group of the Byrd Polar and Climate Research Center at The Ohio State University, the Weather Research and Forecasting model (WRF) has been modified for use in the Polar Regions (referred to as the Polar WRF). A development approach is adopted similar to that used previously to implement the Polar version of the PSU/NCAR fifth generation mesoscale model (Polar MM5). The key modifications for Polar WRF are:
- Optimal surface energy blance and heat transfer for the Noah LSM over sea ice and permanent ice surfaces
- A fix to allow specified sea ice quantities and the land mask associated with sea ice to update during a simulation
Testing of Polar WRF over Arctic and Antarctic surfaces provides guidance on best choice of physics options. See the publications for guidance.
Model evaluations through Polar WRF simulations over Greenland and the Arctic Ocean (SHEBA site) have been performed, and the results are described in the articles provided below. Development studies have been performed for Arctic land (ARM sites in Alaska), and Antarctica.
Polar WRF is used by forecasters as part of the National Science Foundation sponsored Antarctic Mesoscale Prediction System (AMPS; link provided below) to meet the operational and logistic needs of the United States Antarctic Program (USAP). Under a collaborative project with the Polar Meteorology Group, AMPS simulations are performed at the National Center for Atmospheric Research twice per day (00Z and 12Z initializations), and cover progressively finer domains ranging from 45-km (covering most of the Southern Hemisphere) to 1.7-km (covering the region immediately surrounding McMurdo Station, the base of USAP operations). A 45-km resolution version of the Polar WRF is run here at the Byrd Polar and Climate Research Center twice per day for 5 days.
Status of Polar WRF:
Disclaimer: Polar WRF code is released and supported solely by the PMG and is currently based on standard WRF version:
February 6, 2017: Polar WRF 3.8.1 is available
Polar WRF 3.8.1 has been tested in Arctic ARISE simulations. Simulated cloud water was found to be different in the PWRF 3.8.1 compared to 3.7.1, with impacts on shortwave and longwave radiation and the near-surface temperature. This impact also occurs in WRF 3.8.1 and we are discussing possible causes with NCAR.
October 30, 2015: Polar WRF 3.7.1 is available
Polar WRF 3.7.1 has been tested in Arctic and Antarctic simulations. Results are similar to Polar WRF 3.6.1. We've included a fix that allows users of specified sea ice fraction to have this quantity update during a simulation when the SST update option is turned on. If you use standard WRF, sea ice fraction stays fixed in time during a simulation. There is also a fix for sea ice albedo option (1) that forecasts the sea ice albedo based upon temperature and snow depth. The snow fraction is now used in calculating surface fluxes of latent heat and sensible heat. There is also an option for reduced cloud liquid droplet concentration in the Morrison 2-moment microphysics. The reduce droplet concentration from 250 per cm cubed to 50 per cm cubed was already employed in the Regional Arctic System Model and can now be used with Polar WRF. The change tends to produce fewer, larger liquid cloud droplets. Precipitation is now more easily obtained and cloud liquid water amounts should be smaller. Tests show better transmission of solar radiation to the surface with the change.
December 10, 2014: Polar WRF 3.6.1 is available
Polar WRF 3.6.1 was tested with Arctic simulations. The modification files include a fix that allow sea ice fields and the associated land use to update during a simulation. As an additional comment, the use of the newer Noah-MP land surface model (sf_surface_physics = 4) is now encouraged by NCAR land surface modelers. This scheme, which is part of the standard WRF release, runs over soil (the Noah sea ice modules is still called over the oceans where there is sea ice). Noah-MP includes improved snow treatment and has been shown to have some reduced biases.
February 24, 2014: Polar WRF 3.5.1 is available
PWRF 3.5.1 now available. Polar WRF 3.5 was tested with Arctic simulations for 1998 and 2012. The results are in a manuscript submitted to Monthly Weather Review. See Hines et al. (2014) below. The updated version 3.5.1 has also been tested and shown to produce results very similar to 3.5. Polar WRF version 3.5.1 was sent to registered users on February 24, 2014. Future plans include updating Polar WRF for WRF version 3.6 and providing time-dependent data sets of sea ice concentration, sea ice thickness, snow depth on sea ice, and Arctic sea ice albedo through this website. These sea ice fields can be processed through the WRF preprocessing program metgrid into WRF sea surface fields. WRF 3.5/3.5.1 are compatible with these sea ice fields, however, gridded time and space-depedent fields of sea ice thickness, snow cover and albedo have not been easily obtainable by numerical weather prediction community.
October 18, 2012: Polar WRF 3.4.1 is available
WRF Model Version 3.4.1 was released by NCAR on August 16, 2012. The most recent polar modifications (primarily sea ice related) have been ported to WRF 3.4.1 and the resulting code run through a series of validation tests - this is Polar WRF 3.4.1 that is NOW available to the scientific community.
Extensive testing of Polar WRF including version 3.3.1 in Antarctica completed. See Bromwich et al. (2013) below.
November 2, 2011: Polar WRF 3.3.1 is available
Polar WRF 3.3.1 is available. Tested for limited periods over Greenland, Arctic Ocean, and Alaska.
April 25, 2011: Polar WRF 3.3 is available
WRF 3.3 was released to the community by NCAR on April 6, 2011. The Polar Meteorology Group will port all polar modifications to this release and provide a tar file of changes to interested users. Release date of Polar WRF 3.3 is yet to be decided. A suite of test simulations for Arctic sea ice (SHEBA case), the Greenland Ice Sheet, and Antarctic will be performed to verify the reliability of the modified codes.
August 23, 2010: Polar WRF 3.2.1 is available
WRF 3.2 was released to the community by NCAR on April 2, 2010. The Polar Meteorology Group has ported all polar modifications to this release and provides a tar file of changes to interested users as of August 2010. A suite of test simulations for Arctic sea ice (SHEBA case), the Greenland Ice Sheet, and Antarctic are performed to verify the reliability of the modified codes.
The modifications to WRF code for polar applications are available upon request from the Polar Meteorology Group to interested researchers.
October 5, 2009: Polar WRF 3.1.1 is available
WRF Version 3.1.1 was released by NCAR on July 31, 2009 with fractional sea ice now as a standard option in the downloadable code from NCAR. The Polar WRF modifications have now been updated for WRF 3.1.1 as of October 5, 2009. Important new features include the posibility of variable specified sea ice thickness and variable specified snow depth on sea ice. These features have been added in preparation for the Arctic System Reanalysis. Users should expect colder Arctic near-surface atmospheric temperatures and reduced ground heat flux. If you are still using WRF 22.214.171.124, with that previous upgrade in that standard release the Morrison mixed-phase microphysics (developed especially for Arctic stratus clouds and tested by Bromwich et al. ) became a standard option. All other polar physics for WRF 2.2 was implemented into WRF 126.96.36.199, and the modified code was tested fairly extensively. The fractional sea ice description developed by the Polar Meteorology Group (PMG; Le-Sheng Bai and Keith Hines) has been ported to the surface driver included in WRF 188.8.131.52, as well as the modifications to the NOAH LSM required for ice sheets, as described by Hines and Bromwich (2008).
April 9, 2009
WRF 3.1 has been released to the community by NCAR. The fractional sea ice developed by the Polar Meteorology Group is now a standard option within WRF. To use fractional sea ice set the "fractional_seaice" option to 1 in the WRF runtime namelist. Also, set the WPS initialization to ensure that the correct fractional sea ice data values (variable name SEAICE, it's best to include it in the metgrid variable table) are input to your model run. This is a contribution to the International Polar Year (March 2007-March 2009).
September 17, 2009
The full polar physics used with WRF 184.108.40.206 is being ported into a publicly-available modifications tar file for use with the most recent version of WRF, version 3.1.1. This will include the modified surface energy balance over permanent ice sheets, and modified heat transfer within permanent ice. New additions to the Noah LSM will be variable thickness of sea ice and variable snow cover on sea ice. Expected availability of Polar WRF 3.1.1 tar file is late September/early October 2009.
Future for Polar WRF: Polar WRF is a research modification of the standard WRF code and new capabilities will continue to be added. The polar capabilities in standard WRF will likely lag behind those available in the Polar WRF code from the Polar Meteorology Group (not necessarily supported by NCAR). The goals for the next phase of Polar WRF development target more detailed specified sea-ice description within the NOAH LSM, including variable ice thickness and surface albedo, along with a representation of Arctic melt ponds.
Testing of Polar WRF:by the Polar Meteorology Group for representative polar environments
- Greenland ice sheet: evaluation completed and described in Hines and Bromwich (2008).
- SHEBA (Surface Heat Budget of the Arctic Ocean) site in the Beaufort Sea in 1997-1998: evaluation has been completed and described in Bromwich et. al (2009).
- Arctic land areas (Department of Energy Atmospheric Radiation Monitoring (ARM) sites at Barrow and Atqasuk, Alaska): work completed and described in Hines et al. (2011).
- Evaluated for the greater Arctic by Wilson et al. (2011,2012).
- Antarctica: work completed by Francis Otieno, Elad Shilo and David Bromwich. In addition, work on this topic has been ongoing at NCAR (primarily) as part of the Antarctic Mesoscale Prediction System (AMPS) by Kevin Manning and Jordan Powers. See Bromwich et al. (2013).
Watch this location for further updates that will be issued when needed. We appreciate your interest and trust you will acknowledge our efforts on behalf of the scientific community in presentations and publications. Please keep us informed as to manuscripts on Polar WRF so that we can maintain an online archive of relevant publications. Research supported by US federal funding, primarily from the National Science Foundation.