Simple statistical structure in time series for daily air flow characteristics

The structure function is used for quantifying of non‐stationarity in time series for certain daily air flow characteristics (DAFC). The National Center for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) reanalysis 30 year (1968–97) dataset given over 5° × 10° latitude longitude grid is used to calculate the DAFC series for the Baltic Sea region. The quantification on the basis of the Hurst exponent H enables us to discriminate between the strong synoptic scale non‐stationarity and a mild large‐scale one. Presented examples show that the change of H is accompanied by a change of the distribution for the series' increments. Long‐range increments for various DAFC appear to be approximately normally distributed with approximately constant variance independent of the increment length over a remarkably wide range. The property might be useful in describing a long‐range variability of the air flow characteristics.