Abstract Water storage dynamics modulate fluxes within catchments, control the rainfall-runoff response and regulate the velocity of water particles through mixing associated processes. Tracer-aided models are useful tools for tracking the interactions between catchment storage and fluxes, as they can capture both the celerity of the runoff response and the velocity of water particles revealed by tracer dynamics. The phase-space reconstruction of modelled systems can help in this regard; it traces the evolution of a dynamic system from a known initial state as phase trajectories in response to inputs. In this study, we compared the modelled storage-flux dynamics obtained from the application of a spatially distributed tracer-aided hydrological model (STARR) in five contrasting long-term research catchments with varying degrees of snow influence. The models were calibrated using a consistent multivariate methodology based on discharge, isotope composition and snowpack water equivalent. Analysis of extracted modelled storage dynamics gave insights into the system functioning. Large volumes of total stored water needed to be invoked at most sites to reconcile celerity and travel times to match observe discharge and isotope responses. This is because changes in dynamic storage from water balance considerations are small when compared to volume of storage necessary for observed tracer dampening. In the phase-space diagrams, the rates of storage change gave insights into the relative storage volume and seasonal catchment functioning. The storage increase was dominated by hydroclimatic inputs; thus, it presented a stochastic response. Furthermore, depending on the dominance of snow or rainfall inputs, catchments had different seasonal responses in storage dynamics. Decreases in storage were more predictable and reflected the efficiency of catchment drainage, yet at lower storages the influence of ET was also evident. Activation of flow paths due to overland and near-surface flows resulted in non-linearity of catchment functioning largely at high storage states. The storage-discharge relationships generally showed a non-linear distribution, with more scattered states during wettest condition. In turn, all the catchments exhibited an inverse storage effect, with modelled water ages decreasing with increasing storage as lateral flow paths were activated. Insights from this inter-comparison of storage-flux-age dynamics show the benefits of tracer-aided hydrological models in exploring their interactions at well-instrumented sites to better understand hydrological functioning of contrasting catchments.

中文翻译：

---

使用示踪剂辅助径流模型通过集水区相互比较揭示的对比存储-通量-年龄相互作用

摘要 蓄水动力学调节流域内的通量，控制降雨-径流响应，并通过混合相关过程调节水粒子的速度。示踪辅助模型是跟踪流域蓄水量和通量之间相互作用的有用工具，因为它们可以捕获径流响应的速度和示踪动力学揭示的水粒子速度。在这方面，建模系统的相空间重建可以提供帮助；它跟踪动态系统从已知初始状态的演化，作为响应输入的相轨迹。在这项研究中，我们比较了在五个具有不同雪影响程度的对比长期研究集水区中应用空间分布示踪辅助水文模型 (STARR) 获得的模拟存储通量动态。使用基于排放、同位素组成和积雪水当量的一致多变量方法校准模型。对提取的建模存储动态的分析提供了对系统功能的深入了解。大多数站点需要调用大量的总储存水，以协调速度和旅行时间，以匹配观察到的排放和同位素响应。这是因为与观察到的示踪剂衰减所需的存储量相比，考虑到水平衡的动态存储的变化很小。在相空间图中，蓄水变化率可以洞察相对蓄水量和季节性集水区功能。蓄水量增加主要是水文气候输入；因此，它呈现随机响应。此外，根据雪或降雨输入的主导地位，集水区在蓄水动态方面有不同的季节性响应。蓄水量的减少更容易预测，反映了集水区排水的效率，但在蓄水量较低时，ET 的影响也很明显。由于陆上和近地表流动而激活的流动路径导致流域的非线性主要在高蓄水状态下发挥作用。储放电关系一般呈非线性分布，在最湿条件下具有更多的分散状态。反过来，所有集水区都表现出逆向蓄水效应，随着侧流路径被激活，模拟的水龄随着蓄水量的增加而降低。