Evaluation of the NCEP/NCAR and ECMWF 15-yr reanalyses over the data sparse Arctic Ocean

The presence of a floating ice field and blowing snow create difficult conditions for obtaining reliable measurements of atmospheric moisture and precipitation over the Arctic region. These limitations have led the application of the atmospheric moisture budget method (P-E) using reanalysis data. Many aspects of reanalysis data are of high quality over regions with sufficiently dense data. The accuracy of reanalyses is uncertain over areas with sparse data, particularly at high latitudes.

NCEP/NCAR Reanalysis (NNR) and ECMWF 15-year Reanalysis (ERA-15) are compared to two independent observed data sets, the Coordinated Eastern Arctic Research Experiment (CEAREX) and the Lead Experiment (LeadEx), for each component of the moisture budget equation and other meteorological variables. For the moisture field, both reanalyses are drier than the observations in terms of specific humidity and precipitable water at the CEAREX site. Both mean and transient eddy fluxes can be strongly influenced by the simulations of wind field. Generally, zonal fluxes are simulated better than meridional fluxes, particularly for ERA-15. NNR simulated stronger zonal flows during both experiment periods, but ERA-15 produced a better relationship with rawinsondes. Both reanalyses display large discrepancies in simulations of the v components, but exhibit strong agreements in the heights and temperature fields. Overall, ERA-15 performs much better than NNR during the two experiment periods.