Met Office Hadley Centre observations datasets |
> Home > QUARC > Data > Validation experiments > |
These files contain the breakpoints that were derived in each error model and applied to the daytime and nightime control data sets (see Titchner et al., 2009 for more information). A positive (negative) breakpoint is said to occur when there is an increase (decrease) in the mean of the later part time series compared to the mean of the earlier part of the time series. The breakpoint values given in the files below are actually the inverse of the error model breakpoints, and are therefore equivalent to the adjustments required for homogenisation. They should be subtracted from the control data anomalies at and before the breakpoint time to create each error model input data set. Breakpoint timings are given as YYYYMM where YYYY is the year and MM is the first month of the season (e.g. 195712 denotes the December, January and February winter season in 1957/58).
Current Understanding error model | Many Small Breaks error model | Removal of Signal error model | Few Large Breaks error model |
Commercial and media enquiriesYou can access the Met Office Customer Centre, any time of the day or night by phone, fax or e-mail. Trained staff will help you find the information or products that are right for you. |
Maintained by: Holly Titchner |
© Crown Copyright
|