The reduction of precipitation amount bias in the analysis of time series of wet deposition chemism

0029189 - GLÚ 2006 eng A - Abstract
Vach, Marek - Navrátil, Tomáš - Skřivan, Petr - Špičková, Jitka - Kubínová, Petra
The reduction of precipitation amount bias in the analysis of time series of wet deposition chemism.
[Eliminace chyby způsobené vlivem objemu srážek při statistickém zpracování časových řad chemického složení vzorků.]
Acid Rain 2005, 7th International Conference on Acid Deposition. Conference Abstracts. Prague: Ministry of the Environment CR ; Czech Hydrometeorological Institute; Charles University Environment Center; Czech Geological Survey, 2005 - (Hůnová, I.). s. 336-336. ISBN 80-86690-25-3.
[Acid Rain 2005. International Conference on Acid Deposition /7./. 12.06.2005-17.06.2005, Prague]
R&D Projects: GA ČR(CZ) GA205/04/0060
Institutional research plan: CEZ:AV0Z30130516
Keywords : wet deposition * time series * correlation analysis
Subject RIV: DD - Geochemistry

There is a problem of precipitation amount bias occurring in correlation and multivariate covariant matrix using analysis of surveyed elements and ions time series. If the time series of concentrations of elements and ions are evaluated the significant bias caused by sample dilution during more intensive precipitation event is documented by the negative correlation of concentrations of all studied substances with the precipitation amount. Resulting mutual correlation or covariant of the time series of concentrations of studied elements and ions could be then significantly overestimated. The correlation of time series of the deposition fluxes is biased with too pronounced, unreal compensation of the dilution effect. In order to approximate the correlated parameters to the actual content of studied substances in the atmosphere (unaffected with the precipitation amount Ap), we applied the simplified mathematical expression of the wash-out or rain-out process.
Permanent Link: http://hdl.handle.net/11104/0119055