Open AccessA piecewise linear model for detecting climatic trends and their structural changes with application to mesosphere/lower thermosphere winds over Collm, Germany
Author(s) -
Liu R. Q.,
Jacobi Ch.,
Hoffmann P.,
Stober G.,
Merzlyakov E. G.
Publication year - 2010
Publication title -
journal of geophysical research: atmospheres
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.67
H-Index - 298
eISSN - 2156-2202
pISSN - 0148-0227
DOI - 10.1029/2010jd014080
Subject(s) - thermosphere , autoregressive model , white noise , series (stratigraphy) , atmospheric sciences , zonal and meridional , climatology , mesosphere , meteorology , environmental science , mathematics , geology , econometrics , statistics , physics , ionosphere , geophysics , stratosphere , paleontology
A piecewise linear model is developed to detect climatic trends and their structural changes in time series with a priori unknown number and positions of breakpoints (BPs). The departure (i.e., the initial noise term) of trends from time series is allowed to be interpreted by the first‐ and second‐order autoregressive models. The goodness of fit of candidate models, if the residuals are accepted as normally distributed white noise, is evaluated using the Schwarz Bayesian Information Criterion (BIC). The uncertainties of all trend parameters are estimated using the Monte‐Carlo method. The model is applied to the mesosphere and lower thermosphere (MLT) winds obtained at Collm, Germany, during 1960–2007. A persistent increase after ∼1980 of the zonal prevailing wind is observed in all seasons and hence in the zonal annual mean based on the primary models. Trends of the meridional prevailing wind are different for different seasons. Several major trend BPs are identified in the annual mean zonal and meridional winds according to BIC. However, in view of the large wind variability before the late 1970s, alternative models are considered. This provides four additional minor breaks. In some cases, the initial noise must be further interpreted by autoregressive models, suggesting that other unidentified factors may also play a role.