Modelling soil–water interactions provides important outputs for agriculture management and environmental monitoring. Most of the existing models rely on soil hydraulic properties (SHP) as input data. Geostatistical approaches based on stochastic simulations provide the spatial distribution of SHP and the uncertainty attached to their estimates. Frequently, SHP are measured in different data supports thus one needs to guarantee the integration of these different supports to provide a coherent and reliable model. One possible solution is to use block sequential simulation (Liu and Journel, 2009). Our work presents an application of this algorithm to map and quantify the spatial uncertainty of total porosity. Based on the simulation outputs and on the comparison with direct sequential simulation that does not account for the multiple supports of the data, we concluded that block sequential simulation should be used to produce reliable input data to dynamic soil–water models.