From J. Brunhes to F. Durand-Dastès, different generations of geographers succeeded each other in the 20th century, showed that the geography of water was also a matter of time. Redefined as a social construct, water management is based as much on the natural rhythms of its availability as on the temporalities imposed by the economic and socio-cultural configurations that organize the accessibility of water, on intra-annual and to long term scales. The spatio-temporal inertia of hydraulics in the landscape and the weight of socio-political theories (Bichsel, 2016) have nevertheless led to “a contraction of time” in archaeological research by favoring a history of the intentional rather than the operational. This is illustrated by the historiography of the hydraulic networks of ancient Cambodia which made extensive use of the model of the “hydraulic city” to periodize the rise and fall of the "Angkorian civilization" (Groslier, 1979). Such a prism is not unrelated to the dominant use of photo-carto-interpretation – carried out at a large scale and relaunched by the LIDAR – which prevents to apprehend the function of hydraulic networks, such as the one in Banteay Chhoeu (still a matter of debate, see Bourdonneau, 2011). The ongoing study of this ancient network of canals implements a geoarchaeological approach focused on filling deposits. Their potential to record socio-environmental interactions and their fine temporalities has been demonstrated in other contexts (Purdue and Berger, 2015 ; Cez, 2019). A continuum of observations – stratigraphic, micromorphological and particle size analyzes – aims to characterize the deposition rhythm and sedimentation conditions in relation to the natural/controlled flow dynamics. The dating of these deposits (from the 7th century) contributes to render the place of this system in the hydraulic landscape making (chronology, function and temporalities).