IAGA paleointensity database: distribution and quality of the data set

The aim of this paper is to provide details about the design of the paleointensity database and focus on the distribution of the available data set. Version 2003 of the paleointensity database comprises 3128 data from 215 references. Only absolute paleointensity determinations from igneous rocks and baked contacts are considered regardless which method of paleointensity determination was used. The entries in the database correspond to cooling units. All polarities as well as all units are considered (normal, reverse or transitional). The Earth's magnetic field is the only surface trace of the working of the geodynamo and is therefore a very important tool for our understanding of the Earth's deep interior, especially its variation with geologic time. However, to study the long-term variation of the paleomagnetic field, non dipole effects have to be removed. In recent years, the production rate of paleointensity data has strongly increased and the most reliable methods were more frequently used. However, the present data set still has an important temporal bias with 35% of the rocks having ages less than 1 Ma, as well as a strong geographic bias towards the northern hemisphere especially North America and Europe. For most time intervals, secular variation cannot be averaged out, neither temporally nor spatially; therefore interpretation of the averaged field is very limited and is avoided here. For the future, a harmonization or a combination of all IAGA databases would be desirable. Furthermore, the input of raw data at the specimen level would be useful in order to allow reinterpretation of data with more developed and sophisticated methods based on our increasing understanding of rock magnetism. Finally new reliable data are very much needed. (C) 2004 Elsevier B.V. All rights reserved.