Work

ACMANT: Homogenising temperature series with confidence and accuracy

Year: 2012
Type: conference-paper
Abstract: Changes in location, instrumentation or observing hours, and lot of other factors may cause artificial biases in observed climatic data.These undesired biases are often referred as inhomogeneities of time series, and homogenisation is applied for their elimination.For spatially dense networks of temperature observations the use of objective statistical methods is recommended.After a successful homogenisation the observable climatic characteristics (trend, variability, distribution function, etc.) are substantially closer to the reality than in the raw inhomogeneous series.ACMANT is a newly developed, fully automatic method for homogenising networks of monthly temperature series.Its excellent performance has been proved with a series of objective experiments, among others with the tests of the benchmark of the European project HOME.ACMANT is particularly good in the minimisation of root mean squared error, because it harmonises the work on different time scale in a unique way.
Language: English
Open Access status: green
Funder: FP7 Space
OpenAlex ID: W3024425400
FWCI:
Cites: 24
Cited by:
Related to: 8