Abstract. The interconnection of local downslope flows of different intensities with the turbulent characteristics and thermal structure of the atmospheric boundary layer (ABL) is investigated through observations. Measurements are carried out in a relatively flat area 2 km away from the steep slopes of the Sierra de Guadarrama (central Iberian Peninsula). A total of 40 thermally driven downslope events are selected from an observational database spanning the summer 2017 period by using an objective and systematic algorithm that accounts for a weak synoptic forcing and local downslope wind direction. We subsequently classify the downslope events into weak, moderate and intense categories, according to their maximum 6 m wind speed. This classification enables us to contrast their main differences regarding the driving mechanisms, associated ABL turbulence and thermal structure, and the major dynamical characteristics. We find that the strongest downslope flows (U > 3.5 m s−1) develop when soil moisture is low ( < 0.07 m3 m−3) and the synoptic wind not so weak (3.5 m s−1 < V850 < 6 m s−1) and roughly parallel to the direction of the downslope flow. The latter adds an important dynamical input, which induces an early flow advection from the nearby steep slope, when the local thermal profile is not stable yet. Consequently, turbulence driven by the bulk shear increases up to friction velocity (u*) ≃ 1 m s−1, preventing the development of the surface-based thermal inversion and giving rise to the so-called weakly stable boundary layer. On the contrary, when the dynamical input is absent, buoyancy acceleration drives the formation of a katabatic flow, which is weak (U < 1.5 m s−1) and generally manifested in the form of a shallow jet below 3 m. The relative flatness of the area favours the formation of very stable boundary layers marked by very weak turbulence (u* < 0.1 m s−1). In between, moderate downslope flows show intermediate characteristics, depending on the strength of the dynamical input and the occasional interaction with down-basin winds. On the other hand, by inspecting individual weak and intense events, we further explore the impact of downslope flows on CO2 variability. By relating the dynamics of the distinct turbulent regimes to the CO2 budget, we are able to estimate the contribution of the different terms. For the intense event, indeed, we infer a horizontal transport of 67 ppm in 3 h driven by the strong downslope advection.