Workshop on Polar Simulations with the Weather Research and Forecasting (WRF) Model
A coupled atmosphere-–ocean–sea ice–land regional Arctic climate model (RACM) has recently been developed. The atmospheric model used in RACM is the Weather Research and Forecasting (WRF) model. The ocean and sea ice models are the same as those used in the NCAR Community Climate System Model (CCSM3), although used on a regional domain, and are the Los Alamos National Laboratory POP ocean model and CICE sea model. Land surface processes and hydrology are represented by the Variable Infiltration Capacity (VIC) model. These four climate system component models are coupled using the NCAR CCSM coupler CPL7. Initial results from this model will be presented that emphasize the model's ability to simulate the full annual cycle of atmosphere and land state. Results from a ten-year (1989-1999) RACM simulation will be presented and compared with uncoupled WRF-only simulations. The comparison will highlight differences between the atmosphere-land and fully coupled simulations. Future plans for RACM will also be presented, including the addition of ice sheet and dynamic vegetation models.
The atmospheric hydrologic cycle of the Polar version 3.1.1 of the Weather Research and Forecasting model (WRF) is examined for the year December 2006 – November 2007. The domain is similar to that of the Arctic System Reanalysis, an assimilation of model fields with Arctic observations being conducted partly by the Byrd Polar Research Center. Simulations are performed in 48 hour increments initialized daily at 0000 UTC, with the first 24 hours discarded for model spin-up of the hydrologic cycle and boundary layer processes. Precipitation analysis reveals a negative annual mean bias (-9.4%) in the polar region, with stations in the Canadian Archipelago particularly dry. Annual mean bias for the mid-latitudes is small and positive (4.6%), attributed to excessive precipitation during spring and summer when convection is high. An examination of precipitation within 4 major Arctic river basins shows large positive biases due to excessive convection in the summer as well. A sensitivity simulation using a nudging technique on the model’s atmospheric moisture decreases convection and improves the prediction of precipitation. Calculated cloud fraction shows the model has too little cloud cover, supported by an excess in shortwave radiation and a deficit in longwave radiation throughout the domain. The longwave bias is present regardless of the amount of cloud water or cloud ice in the model atmosphere, demonstrating a need to improve cloud effects on radiation in Polar WRF. This examination provides a benchmark of the hydrological cycle of Polar WRF and its use as ASR’s primary model.
As there are few in situ observations in and around Antarctica, it is important to assess how assimilating remotely-sensed observations, such as satellite-observed radiances, can fill this observational gap. Thus, a month-long study was conducted over the Antarctic to examine forecast and analysis sensitivity to the assimilation of microwave radiance measurements. Several experiments using both cyclic and non-cyclic initial conditions were configured to quantify the impact of radiance data assimilation (DA) and explore different approaches of radiance bias correction (BC). DA was performed using the Weather Research and Forecasting (WRF) model’s three-dimensional variational (3DVAR) algorithm, and the analyses initialized 72-hr Advanced Research WRF model forecasts.
The results demonstrate the critical importance of properly bias correcting raw radiance observations. When assimilating radiances using a “cold start” BC technique, forecasts and analyses were degraded compared to those from parallel experiments that only assimilated conventional (i.e., non-radiance) observations. However, when BC parameters were “spun-up” for several months before the assimilation period, radiance DA yielded forecast and analysis improvements compared to when only conventional observations were assimilated. The same general results regarding radiance DA were obtained for both the cyclic and non-cyclic experiments, but cycling led to poorer overall analyses and forecasts.
This presentation will discuss these results and describe a suggested method for radiance BC within limited-area domains.
The presence of surface melting on ice sheets and ice shelves marks an important climatic and geophysical threshold in the cryosphere. Wetting of snow reduces albedo and encourages additional melt, meltwater runoff contributes to mass loss from ice sheets, and penetration of meltwater to the glacier bed can lubricate faster flow and contribute to ice-sheet mass loss. Meltwater may also contribute to ice-shelf collapse through wedging open of crevasses.
While fringing ice-shelf collapse along the Antarctic Peninsula is probably the best known example of the cryospheric response to a warming atmosphere (and ocean), surface melting is also present in inland portions of West Antarctica. In addition to potentially contributing to ice sheet dynamics, surface melt occurrence is a valuable proxy for changing atmospheric temperature conditions.
Combining satellite remote sensing with atmospheric modeling (i.e., Polar WRF), we will be attempting to diagnose the meteorological conditions associated with surface melting on the Antarctic ice sheet and its fringing ice shelves. With these results, we plan to predict whether the regional warming associated with anticipated anthropogenic global warming and related atmospheric circulation changes will lead to a future increase of melting.
We are currently preparing case studies of selected West Antarctic melt events that document the utility and skill of our meteorological datasets (reanalyses, RCMs, selected CMIP5 GCMs) in the development of diagnostic tools for identifying surface melt as observed by satellite and simulated by models.
At CIRES/NOAA/ESRL, we have been using WRF to perform Arctic process studies. Two recent studies have explored the role of microphysical parameterizations for producing proper structure and physical interactions within Arctic mixed-phase stratocumulus clouds (AMPS) during MPACE, and to better understand the longevity of decoupled AMPS during the ISDAC field program. The first study showed the importance of interactions between snow and liquid water in order to maintain the observed amounts of supercooled liquid water in the AMPS. The maintenance of the liquid water in the cloud was only possible through either including a double-moment microphysics scheme, where both the number concentration and mass concentration are used for most microphysical constituents, or using a single-moment scheme with a snow size distribution intercept parameter (N0s) that was tuned to the Arctic conditions using the double-moment scheme. This intercept parameter was substantially smaller than typical mid-latitude values. Both simulations produced longer-lived supercooled cloud liquid water, which impacted the downwelling longwave and shortwave radiation as much as 40-80 W m-2, and produced surface radiation in much closer agreement with observations than when the clouds contained much less liquid water. The simulated liquid water contents and snow spectra are verified with observations as well. The second study illustrated the importance of an Arctic humidity inversion at cloud top for the persistence of decoupled AMPS. Budget results show that cloud liquid water is maintained in the upper entrainment zone near cloud top (within a temperature and humidity inversion) due to a down gradient transport of water vapor by turbulent fluxes into the cloud layer from above and direct condensation forced by radiative cooling. Liquid water is generated in the updraft portions of the mixed-layer eddies below cloud top by buoyant destabilization. These processes cause at least 20% of the cloud liquid water to extend into the inversion. In this decoupled system, the humidity inversion is the only source of water vapor for the cloud system, since water vapor from the surface layer is not efficiently transported into the mixed layer. Sedimentation of ice is the dominant sink of moisture from the mixed layer.
ASR land data assimilation
Land data assimilation is a vital component to Arctic System Reanalysis. The timing and amount of snow is a crucial component to both the energy and water cycles. The extent of the ASR domain also requires the accurate specification of vegetation condition. Observed albedo provides a constraint on the surface energy. The current ASR system incorporates NASA, NESDIS, and NOAA satellite observations for these vital surface properties.
An overview of the ASR system will be shown along with the land surface products used. Test results using the offline HRLDAS will be shown along with introductory results from coupled tests within the variational coupled atmosphere-ocean-land system.
Plans for a new land surface model for NCEP operations and for use with WRF
Over the past few years, a new land surface model has been developed to address some of the known deficiencies in the Noah LSM. This new model, called Noah-MP, is based on using multiple physics(MP) options for land surface processes and will be included in the next WRF release. In contrast to standard Noah, Noah-MP contains an explicit vegetation canopy, multi-layer snowpack, snow water holding capacity, options for vegetation shading, photosynthesis-based canopy resistance and a simple dynamical vegetation model. I will give a brief overview of Noah-MP and show some performance differences between Noah and Noah-MP.
Keith M. Hines
Testing and Formal Release of Polar WRF 3.3.1
The new version of Polar WRF, 3.3.1, is being released in conjunction with the Polar Simulations with the Weather Research and Forecasting (WRF) Model workshop. As before, the Polar WRF package consists of a series of modified files to replace selected files from the standard version of WRF that is obtained from NCAR. The new version of WRF has been tested over the Arctic Ocean in comparison to 1998 Surface Heat Budget of the Arctic Ocean (SHEBA) observations during January and August, and in comparison to 2007 Greenland observations during January and August. Polar WRF 3.3 has also been used in recent Arctic System Reanalysis tests that show regional performance exceeding the global ERA-Interim reanalysis. Polar WRF 3.3 performance over the Arctic Ocean is very similar to that of Polar WRF 3.2.1. Performance over Greenland meets or exceeds that of Polar WRF 3.2.1. A winter warm bias over the Greenland Ice Sheet is reduced. Capabilities of version 3.3.1 are similar to those of version 3.2.1 and are input through a series of compiler options in the Fortran files. These include treating specified variable sea ice thickness and snow cover on sea ice. An input Arctic sea ice thickness dataset from William Chapman of the University of Illinois is now extended to 2010. Furthermore, a time-varying, temperature based sea ice albedo calculation from Chapman is now included as an alternative to a specified seasonally-varying Arctic sea ice albedo. Polar WRF 3.3.1 can be obtained on request from Dr. David Bromwich.
Development and testing of Polar WRF, Part IV: Antarctica
Three recent versions of Polar Weather Research and Forecasting model developed at The Ohio State University are evaluated over Antarctica at 60 km to determine the impact of model improvements, large-scale circulation variability, and uncertainty in driving data. Differences in forecast skill between recent versions of Polar WRF are small. The overall model skill varies with season. Surface temperatures are colder than observed in the summer because of anomalously large heat fluxes into the snow and excess longwave loss. In the winter, mechanical mixing in the stable boundary layer due to positive wind speed bias results in a warm bias due to large downward fluxes of sensible heat. Wind speeds and other variables are forecast better away from the complex surface topography. The model also overpredicts the summer incident surface shortwave but underpredicts the longwave down at the surface in all seasons. Sensitivity experiments show that simulations with CAM and RRTMG physics produce modest reduction in the excess downwelling shortwave and corresponding biases in surface temperatures but including the gravity wave drag has little effect on the surface wind speed biases. Cloud radiation interaction is likely not well represented in the model leading to higher radiation prediction skill under clear but not cloudy skies. Antarctic simulations require careful treatment of snowpack temperatures because the excess downwelling shortwave radiation results in steep temperature gradients in the snow packs leading to anomalously large downward subsurface heat fluxes in summer. Polar WRF skill is robust even without nudging and is insensitive to variations in the large-scale circulation.
POSTER: David H. Bromwich
The ACCIMA project- coupled modeling of the high southern latitudes
Recent work in the climate modeling community has emphasized coupling of multiple modeling components so as to achieve flexible, quantitative, multi-disciplinary tools to address the various critical climate questions. Beyond the global efforts, coupled models are now being applied at higher-resolution regional scales. A new regional coupled modeling project is being applied for the West Antarctic Ice Sheet (WAIS), that represents about 10% of the volume of the entire Antarctic ice sheet. The WAIS is currently losing mass due to several of its outlet glaciers draining into the Amundsen Sea. The rate of this mass loss, over natural accretion through snowfall, has also been increasing in recent years, encouraging our efforts in the Atmosphere-Ocean Coupling Causing Ice Shelf Melt in Antarctica (ACCIMA) project. A team of researchers from Ohio State, Old Dominion University and New York University will develop and couple components of an earth system model for the Southern Ocean with a regional emphasis on the West Antarctic. The component system models to be coupled include the polar-optimized version of the Weather Research and Forecasting model (Polar WRF) for the atmosphere. The ocean component will be the Regional Ocean Modeling System (ROMS), and the sea ice component will be the Los Alamos sea ice model (CICE). Retrospective decadal simulations will be done to understand recent past variability. Downscaled future projections for Antarctica will be driven by the global National Center for Atmospheric Research (NCAR) Community Climate Model (CCSM or its equivalent). Resulting simulations will explore the possibility that increased basal melt of the seaward floating ice shelves of the region's terminal glaciers is increasing ice mass loss from West Antarctica.