Open AccessA physics‐based correction model for homogenizing sub‐daily temperature series
Author(s) -
Auchmann R.,
Brönnimann S.
Publication year - 2012
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 - 10.1029/2012jd018067
Subject(s) - homogenization (climate) , environmental science , meteorology , wind speed , energy balance , data assimilation , offset (computer science) , climate model , climate change , computer science , physics , geology , biodiversity , ecology , oceanography , biology , programming language , thermodynamics
A new physics‐based technique for correcting inhomogeneities present in sub‐daily temperature records is proposed. The approach accounts for changes in the sensor‐shield characteristics that affect the energy balance dependent on ambient weather conditions (radiation, wind). An empirical model is formulated that reflects the main atmospheric processes and can be used in the correction step of a homogenization procedure. The model accounts for short‐ and long‐wave radiation fluxes (including a snow cover component for albedo calculation) of a measurement system, such as a radiation shield. One part of the flux is further modulated by ventilation. The model requires only cloud cover and wind speed for each day, but detailed site‐specific information is necessary. The final model has three free parameters, one of which is a constant offset. The three parameters can be determined, e.g., using the mean offsets for three observation times. The model is developed using the example of the change from the Wild screen to the Stevenson screen in the temperature record of Basel, Switzerland, in 1966. It is evaluated based on parallel measurements of both systems during a sub‐period at this location, which were discovered during the writing of this paper. The model can be used in the correction step of homogenization to distribute a known mean step‐size to every single measurement, thus providing a reasonable alternative correction procedure for high‐resolution historical climate series. It also constitutes an error model, which may be applied, e.g., in data assimilation approaches.