@article{PAR00010200, title = {{U}se of gaseous tracers ({CFC}s and {SF}6) and transit-time distribution spectrum to validate a shallow groundwater transport model}, author = {{M}ol{\'e}nat, {J}{\'e}r{\^o}me and {G}ascuel-{O}doux, {C}. and {A}quilina, {L}. and {R}uiz, {L}.}, editor = {}, language = {{ENG}}, abstract = {{C}atchment-scale solute transport models constitute potentially powerful tools of water resources management for predicting the evolution of water quality in response to land use modifications. {V}alidation of the models remains a crucial step in order to get reliable prediction. {H}owever, pertinent data for the validation step are lacking. {W}hen catchment solute transport models are applied to shallow groundwater catchments, catchment-scale model validation relies on the validation of the groundwater solute transport model. {H}ere we studied the interests of two approaches for groundwater solute transport model validation: the use of (i) gaseous atmospheric tracers and (ii) of spectral properties of transit time distributions derived from tracer concentration time series of rainfall and stream base-flow. {T}he {K}ervidy-{N}aizin catchment, a 5 km(2) catchment in {W}estern {F}rance was chosen as case application. {T}hree simulations with a shallow groundwater transport model were performed using three sets of effective porosity, respectively. {W}e then investigated the ability of four gaseous tracers ({CFC}-11, {CFC}-12, {CFC}-113 and {SF}6) and transit time distribution spectra to identify the most realistic simulations. {T}he three simulations led to mean transit times that varied over a wide range of values, from 1.2 to 12.1 yr. {S}imulated groundwater mixing ratio differences for the three simulations did not exceed 5% for {CFC}-11, regardless of the groundwater location. {T}he values of the {SF}6, {CFC}-12 and {CFC}-113 mixing ratio differences between the two simulations that contained the two shortest mean transit times were within expected measurement error of the tracer concentrations, regardless of the groundwater location. {I}n contrast, the mixing ratio differences between the two simulations with the shortest mean transit times and the simulation containing the largest one exceeded 28% at all groundwater locations for {SF}6 and 9% in groundwater discharging into the stream for {CFC}-12 and {CFC}-113. {T}he spectra of the three simulated transit time distributions do not match the transit time distribution spectrum derived from chloride concentration time series. {S}o these results show that transit time distribution spectrum, and in much less extend {SF}6, {CFC}-12 and {CFC}-113 mixing ratios are sensitive enough to effective porosity values to be used as validation variables in shallow groundwater transport modelling. {T}ransit time distribution spectrum constitutes an integrative variable able to discriminate between shallow groundwater transport simulations leading to very close mean transit times. {I}n the present work, the discrepancy between the simulated spectra and the spectrum derived from the chloride concentration may come from other catchment processes than groundwater transport process, or from a wrong modelling of the groundwater transport process. {G}lobally, deriving reliable transit time distribution spectrum requires high-frequency and extensive time series of a conservative tracer measured in stream water and rainfall, which is not straightforward to collect. {T}he low sensitivity of simulated {CFC}-11 tracer mixing ratios to simulated groundwater transit times, and thus to model parameters, constituted a shortcoming to use it as a validation variable of groundwater solute transport models.}, keywords = {{V}alidation ; {S}olute transport model ; {C}hloride concentration ; {S}pectral properties ; {CFC} ; {FRANCE}}, booktitle = {}, journal = {{J}ournal of {H}ydrology}, volume = {480}, numero = {}, pages = {1--9}, ISSN = {0022-1694}, year = {2013}, DOI = {10.1016/j.jhydrol.2012.11.043}, URL = {https://www.documentation.ird.fr/hor/{PAR}00010200}, }