In order to investigate hydraulic connections between the Brenta river and its groundwater aquifers, this work performs a stochastic analysis of measured groundwater temperature timeseries. The aim is to compare their trends to verify their mutual influence in stream seepage pathways. Additionally, the study wants to assess whether a large check dam, built across the river to increase aquifer recharge, is working properly. Field data were obtained by a water gauge placed directly into the stream, by two riverbank control points and by a monitoring well representing the regional groundwater thermal trend. Timeseries were evaluated by applying three different analyses - Fourier analysis, autocorrelation and sliding time-window Pearson correlation. Control points results were compared to the inflows to sort whether the aquifer is locally more affected by river dispersion or by regional groundwater flow. The three considered methods substantially agree in showing an increase of the seepage area upstream of the check dam and, at the same time, a marked reduction of groundwater recharge after few years of operation.