摘要: In this study, we have examined the systematic humidity errors of both global meteorological analyses and a global numerical weather prediction (NWF) model (in various versions) initialized from them. Upon determining a definition... 展开 In this study, we have examined the systematic humidity errors of both global meteorological analyses and a global numerical weather prediction (NWF) model (in various versions) initialized from them. Upon determining a definition for systematic error, we obtained the individual forecast bias and the ensemble bias over a sequence of forecast runs for each of several versions of the modeL We found that model formulations designed to reduce systematic humidity forecast error with respect to one set of initial conditions did not produce the corresponding error reductions when initialized from a different set of meteorological analyses. Thus, both model formulation and initial conditions affect systematic humidity forecast error Statistics gathered from several operational mesoscale NWP models suggest that each model has its own unique error characteristics. (AN). 收起