The method presented in this chapter computes an estimate of the location of a single gas sourcefrom a set of localised gas sensor measurements. The estimation process consists of three steps.First, a statistical model of the time-averaged gas distribution is estimated in the form of a two-dimensional grid map. In order to compute the gas distribution grid map the Kernel DM algorithm isapplied, which carries out spatial integration by convolving localised sensor readings and modelling theinformation content of the point measurements with a Gaussian kernel. The statistical gas distributiongrid map averages out the transitory effects of turbulence and converges to a representation of thetime-averaged spatial distribution of a target gas. The second step is to learn the parameters ofan analytical model of average gas distribution. Learning is achieved by nonlinear least squaresfitting of the analytical model to the statistical gas distribution map using Evolution Strategies (ES),which are a special type of Evolutionary Algorithms (EA). This step provides an analysis of thestatistical gas distribution map regarding the airflow conditions and an alternative estimate of thegas source location, i.e. the location predicted by the analytical model in addition to the location ofthe maximum in the statistical gas distribution map. In the third step, an improved estimate of thegas source position can then be derived by considering the maximum in the statistical gas distributionmap, the best fit as well as the corresponding fitness value. Different methods to select the mosttruthful estimate are introduced and a comparison regarding their accuracy is presented, based on atotal of 34 hours of gas distribution mapping experiments with a mobile robot. This chapter is anextended version of a paper by the authors (Lilienthal et al. [2005]).