A review of probability aggregation methods in Earth sciences Alessandro Comunian, Denis Allard and Philippe Renard Abstract The need of combining in a probabilistic framework different sources of information is a frequent task in earth sciences. This can occur for example when modeling a reservoir using direct geological observations, geophysics, remote sens- ing, training images etc. For example, the probability of occurrence of a certain lithofacies at a certain location can easily be computed conditionally on the event observed at each source of information. The problem of aggregating these different conditional probability distributions into a single conditional distribution arises as an approximation to the inaccessible genuine conditional probability given all infor- mation. This paper makes a formal review of most aggregation methods proposed in the literature with a particular focus on their mathematical properties. Calibration of the aggregated probability distribution is of particular importance. It is known that linear aggregation operators are not calibrated. Here, we show that if a cali- brated log-linear pooling exists, then it is the log-linear pooling with parameters estimated from maximum likelihood. Simulations in a spatial context illustrate the performance of these operators. Alessandro Comunian NCGRT, University of New South Wales, Sydney 2052 NSW, Australia e-mail: a.comunian@unsw.edu.au Denis Allard INRA, UR546 BioSP, Site Agroparc 84914 Avignon, France e-mail: allard@avignon.inra.fr Philippe Renard CHYN, University of Neuchˆ atel, Emile-Argand 11, 2000 Neuchˆ atel, Switzerland e-mail: philippe.renard@unine.ch Ninth International Geostatistics Congress, Oslo, Norway, June 11. – 15., 2012