Description of the COST-HOME monthly benchmark dataset and the submitted homogenized contributions

As part of the COST1 Action HOME2 a dataset has been generated that will serve as a benchmark for homogenisation algorithms. Members of the Action and third parties have been invited and are still welcome to homogenise this dataset3 . The results of this exercise was analysed to obtain recommendations for a standard homogenisation procedure and are described in an upcoming article (Venema et al., 2011). Chapter two discusses the generation of this benchmark dataset, the climate variables considered, which types of data are in the benchmark dataset, how they have been produced, the ways to introduce artificial inhomogeneities, and the additional specifications such as length, missing data and trends. This chapter is an updated version of a report (Venema et al., 2009), which was available to the participants. A draft for the properties of the benchmark was developed at a WG1 meeting and was approved by the Management Committee of the Cost Action. The homogenized data that was returned by the participant is described in Chapter 3. In total 25 contributions have been returned before the deadline at which the truth was revealed. Multiple late contributions, submitted after the deadline, are also described. The descriptions of the contributions were written by the respective participants.