Can we quantify the variability of soil moisture across scales usingElectromagnetic Induction ?Jérémy Robinet (1), Christian von Hebel (2), Jan van der Kruk (2), Gerard Govers (1), Jan Vanderborght (1,2)(1) Department of Earth and Environmental Sciences, Katholieke Universiteit Leuven, Leuven, Belgium , (2)Forschungszentrum Jülich, Agrosphere (IBG-3), Jülich, GermanySoil moisture is a key variable in many natural processes. Therefore, technological and methodological advance-ments are of primary importance to provide accurate measurements of spatial and temporal variability of soilmoisture. In that context, ElectroMagnetic Induction (EMI) instruments are often cited as a hydrogeophysicalmethod with a large potential, through the measurement of the soil apparent electrical conductivity (ECa).To our knowledge, no studies have evaluated the potential of EMI to characterize variability of soil mois-ture on both agricultural and forested land covers in a (sub-) tropical environment. These differences in land usecould be critical as differences in temperature, transpiration and root water uptake can have significant effect,notably on the electrical conductivity of the pore water.In this study, we used an EMI instrument to carry out a first assessment of the impact of deforestation andagriculture on soil moisture in a subtropical region in the south of Brazil. We selected slopes of different topogra-phies (gentle vs. steep) and contrasting land uses (natural forest vs. agriculture) within two nearby catchments. Atselected locations on the slopes, we measured simultaneously ECa using EMI and a depth-weighted average ofthe soil moisture using TDR probes installed within soil pits.We found that the temporal variability of the soil moisture could not be measured accurately with EMI,probably because of important temporal variations of the pore water electrical conductivity and the relativelysmall temporal variations in soil moisture content. However, we found that its spatial variability could beeffectively quantified using a non-linear relationship, for both intra- and inter-slopes variations. Within slopes,the ECa could explained between 67 and 90% of the variability of the soil moisture, while a single non-linearmodel for all the slopes could explain 55% of the soil moisture variability. We eventually showed that combin-ing a specific relationship for the most degraded slope (steep slope under agriculture) and a single relationshipfor all the other slopes, both non-linear relations, yielded the best results with an overall explained variance of 90%.We applied the latter model to measurements of the ECa along transects at the different slopes, which al-lowed us to highlight the strong control of topography on the soil moisture content. We also observed a significantimpact of the land use with higher moisture content on the agricultural slopes, probably due to a reducedevapotranspiration.