Contents:
Background
Objectives
Methodology and Validation
Data Requirements
References
Further information

The diagnostic subproject proposed here is a continuation of Diagnostic Subproject No. 8 (Polar Processes and Sea Ice) in AMIP-I. This subproject has focussed on the AMIP model-simulated fields that are most relevant to the forcing of sea ice. Examples are the near-surface temperatures (Tao et al., 1996), precipitation (Walsh et al., 1997), surface winds (Cattle, ongoing) and the surface energy fluxes over sea ice (Maslanik, ongoing). In addition, the AMIP simulations of moisture fluxes and energy budgets over the Antarctic and Greenland ice sheets are being examined by Bromwich and by Kattsov and Meleshko. A substantial part of the collective effort in Subproject No. 8 has been invested in the enhancement of the datasets required to assess the polar output of the AMIP models. In the papers referenced above, for example, gridded datasets available through PCMDI have been augmented with Arctic-specific data on temperature and precipitation from various sources. The results of the collective effort of our AMIP Phase I diagnostic subproject are being consolidated into a report for submission to the WCRP.

 Several of the individual efforts in our subproject are ongoing and will continue into the period of AMIP-II. These studies will clearly benefit from the extended simulation periods, from the inclusion of the additional variables in the Standard Output and from the use of more realistic sea ice boundary conditions (concentrations) vis-a-vis AMIP-I. More importantly, however, the next phase of a polar diagnostic subproject will mesh with two major efforts that are coming to fruition subsequent to our AMIP-I subproject. The first is the completion of the atmospheric reanalysis projects at NCEP and ECMWF. While the reanalysis products will be the cornerstone of many AMIP-II diagnoses and assessments, their derived fields (e.g., surface fluxes, radiative fields) provide an unprecedented opportunity to extend the diagnoses of polar simulations in climate models. Second, 1997 marked the start of several coordinated Arctic field programs: SHEBA (Surface Heat Budget of the Arctic Ocean), DOE's ARM (Atmospheric Radiation Measurement) program on the North Slope of Alaska and Adjacent Arctic Ocean, and NASA's FIRE-III campaign over the Arctic Ocean [see Randall et al. (1998) for an overview]. All of these programs are motivated by the need for data to test and improve the simulation of polar processes in global climate models. Thus the opportunities for meaningful collaboration between the polar and global climate modeling communities will accelerate dramatically in the next few years, and our proposed AMIP-II subproject is intended to provide a significant bridge between the two communities.

 The earlier subproject also included an examination of snow cover in the AMIP-I simulations (Frei and Robinson, 1995). Because snow cover extends well beyond the polar regions, and indeed varies primarily in middle latitudes during the northern winter, we have concluded that the snow cover diagnosis in AMIP-II is more logically a separate diagnostic subproject. Consequently, D. Robinson (Rutgers) and R. Brown (Atmospheric Environment Service) will submit an AMIP-II subproject proposal that will specifically address hemispheric snow cover.
 

While the scope of the proposed polar subproject is intended to be sufficiently broad to allow for the entrainment of additional investigators having polar interests, several foci have emerged: (1) polar clouds and radiative fluxes, (2) the polar water vapor fluxes (lateral as well as surface fluxes), and (3) downslope airflow in the vicinity of ice sheet margins. In addition, we hope to stimulate AMIP-II experiments pertaining to (a) the parameterization of horizontal thickness variations and leads in pack ice, and (b) the inclusion of ice-phase microphysical processes in polar clouds. These foci are addressed sequentially in the following paragraphs.

 (1) Polar clouds and their radiative interactions are the scientific drivers of the ARM, SHEBA and FIRE programs. For the evaluation and development of parameterizations of cloud/radiative interactions, the field phases (1997-98) of these programs have been designed to produce an Arctic dataset containing vertical profiles of cloud and aerosol characteristics (e.g., ice and liquid water); radiative fluxes at the surface, top of atmosphere and within the atmosphere; surface albedo and the interfacial fluxes of sensible and latent heat. These variables will be compiled into averages for GCM grid-sized areas. We will use this dataset as a benchmark for evaluating the cloud and radiative fields in the AMIP-II simulations. The evaluations will draw upon the following variables in the AMIP-II standard output (cf. Tables 1 and 2 in AMIP-II Guidelines):

 Table 1: cloud fraction

  cloud amount (surface and satellite views)
  cloud liquid water
  cloud ice
  extinction coefficient
  cloud emittance

 Table 2: surface incident and reflected SW radiation

  surface downwelling and upwelling LW radiation
  TOA reflected SW radiation
  TOA outgoing LW radiation
  net radiation at model top
  surface sensible heat flux
  surface latent heat flux
  surface evaporation (+ sublimation) rate
  vertically integrated cloud water
  vertically integrated cloud ice

In using the AMIP-II standard output, we will work primarily with the climatological (17-year) monthly means. It will be necessary to assume that the SHEBA/ARM/FIRE datasets are representative of the corresponding observational climatological means. Our experience with the AMIP-I output suggests that the across-model scatter of the means of key Arctic variables (e.g., surface radiative fluxes, surface evaporation) is considerably larger than the estimated variability of the decadal-scale means of the observational values.

(2) Atmospheric water vapor transports are central components of the hydrological cycles of the polar regions. Our AMIP-I work included an intercomparison of the surface evaporative fluxes in the various models, together with an attempt to diagnose the model-to-model differences in these fluxes (Walsh et al., 1998). However, a more comprehensive diagnosis of the hydrologic cycle was limited by the availability of output required for diagnostic computations of poleward moisture fluxes (i.e., high-frequency v and q with adequate vertical resolution). The inclusion of the mean product for v and q for 17 levels in the AMIP-II output (Table 1) will permit more complete closure of the models' polar moisture budgets and will permit direct comparisons with corresponding quantities in the reanalyses. We will perform systematic comparisons of the atmospheric moisture fluxes and flux convergences in the AMIP-II models (and in the reanalysis output) for several regions: the Arctic polar cap (70? -90? N), Greenland and Antarctica. Bromwich will assume the lead role in the studies of the Greenland and Antarctic regions, while Walsh will work with M. Serreze (U. Colorado) in the moisture budget study for the central Arctic. A key issue of the regional budget studies will be the partitioning of the moisture "source term" between advective influx and surface evaporation (sublimation).

(3)  Downslope or katabatic winds are prominent features of the coastal climates of Antarctica, Greenland and, to a lesser extent, the smaller icecaps of the Northern Hemisphere. Because these winds lead to coastal polynya formation by advecting sea ice offshore, they play important roles in air-sea energy exchange, sea ice formation (i.e., new ice growth) and the associated input of salt to the upper ocean (Liu et al., 1997). The ability of models to capture the katabatic wind regime will be a key issue for coupled model simulations of atmosphere-ocean interaction in the areas surrounding the ice sheets. While katabatic winds have been reproduced in short simulations with limited-area models of the atmosphere (Bromwich et al., 1994), their simulation in global models has received little attention. We propose to assess the AMIP-II model simulations of the katabatic wind regimes over Antarctica and Greenland by evaluating the downslope wind component relative to the synoptic-scale pressure gradients in the same regions. The difference between the downslope wind component and the value implied by the pressure gradient will provide an index of the katabatic component of the wind. Since the ability of the reanalyses to capture katabatic winds has yet to be determined, the verification will draw upon the routine wind reports from coastal stations of Antarctica and Greenland, as well as from the automated weather station (AWS) network of Antarctica and from buoys on the Antarctic ice shelves.

The major part of the katabatic wind assessment will utilize monthly mean fields of the AMIP-II Standard Output: surface (10 m) winds, mean sea level pressure and surface pressure (Table 2 of AMIP-II Guidelines). The katabatic assessment will also require the surface topography grid of each model (Table 5). In addition, we will select several models for an examination of the high-frequency variations of the downslope wind component in order to assess the models' abilities to simulate Katabatic Surge Events (Liu et al., 1997), which represent an interplay between offshore synoptic conditions and the downslope drainage phenomenon. This sub-task will require the six-hourly values of surface (10 m) wind and pressure (Table 6) from a subset of 3-5 models. A few annual cycles of the high-frequency output from these models should be sufficient. We will inform all the AMIP-II modeling groups of these needs, thereby offering any interested groups the opportunity to provide this output either through the AMP-II Supplementary Output or directly to our diagnostic subproject (D. Bromwich).

In addition to the analyses of AMIP-II standard and supplementary output as described above, we will work within the AMIP-II framework to encourage numerical experimentation on two topics: (1) the role of ice-phase microphysics in the seasonal cycle of Arctic cloudiness (Beesley and Moritz, 1999) and the model sensitivities to sea ice thickness and the thickness distribution (Rind et al., 1997). A first step will be preliminary experiments with at least one model. The NCAR CSM is a natural candidate because it is available to the broad community of university scientists and we have already contacted the CSM Polar Working Group (J. Weatherly, co-chair) in order to initiate action on this issue. Other models available to us for such purposes are the MGO and UKMO global models. The second step, contingent on the results from the first step, will be an experimental subproject proposal to AMIP-II. The experimental subproject will complement the diagnostic subproject discussed here, but it will not involve all participants in the diagnostic subproject.

References

Beesley, J. A., and R. E. Moritz, 1999: Towards an explanation of the annual cycle of cloudiness over the Arctic Ocean. J. Climate, 12, in press.

Bromwich, D. H., Y. Du, and T. R. Parish, 1994: Numerical simulation of winter katabatic winds from West Antarctica crossing Siple Coast and the Ross Ice Shelf. Mon. Wea. Rev., 122, 1417-1435.

Frei, A., and D. A. Robinson, 1995: Northern Hemisphere snow cover extent: Comparison of AMIP results to observations. Proceedings of the First International AMIP Scientific Conference (Monterey, CA), World Climate Research Programme, WCRP-92, WMO/TD-No. 732, 499-504.

Liu, Z., M. L. Van Woert, and D. H. Bromwich, 1997: Winter atmospheric forcing of the Ross Sea polynya. Antarctic Continental Shelf Oceanography, Antarctic Research Series, American Geophysical Union, Washington, DC, in press.

Randall, D., J. Curry, D. Battisti, G. Flato, R. Grumbine, S. Hakkinen, D. Martinson, R. Preller, J. Walsh, and J. Weatherly, 1998: Status of and outlook for large-scale modeling of atmosphere-ice-ocean interactions in the Arctic. Bull. Amer. Meteor. Soc., 79, 197-219.

Rind, D., R. Healy, C. Parkinson, and D. Martinson, 1997: The role of sea ice in 2 × CO2 climate model sensitivity. Part II: Hemispheric dependence of sea ice thickness and extent. Geophys. Res. Lett., 24, 1491-1494.

Tao, X., J. E. Walsh, and W. L. Chapman, 1996: An assessment of global climate model simulations of Arctic air temperatures. J. Climate, 9, 1060-1076.

Walsh, J. E., V. Kattsov, D. Portis, V. Meleshko and participating AMIP modeling groups, 1998: Arctic precipitation and evapotranspiration: Model results and observational estimates. J. Climate, 11, 72-87.
 

LLNL Disclaimers

UCRL-MI-127350