An isotopic and geochemical study of snowmelt runoff in a small arctic watershed

‘Old’ water contributions to snowmelt runoff in a stream can be defined as water which was stored in the catchment prior to the start of the runoff event in question. We used mass balance techniques for natural oxygen-18 and several chemical parameters (electrical conductivity, Ca and Mg) to estimate the magnitude and timing of ‘old’ water contributions to snowmelt runoff in the Apex River watershed during the 1983 field season. The Apex River catchment is located in the southern part of Baffin Island, Northwest Territories, Canada. The watershed has an area of 60 km², it is in the zone of continuous permafrost, and the geology is dominantly Precambrian gneiss with sporadic, thin, glacial overburden. The isotopic data indicate that for the snowmelt season of 1983 (third highest peak discharge of 11 years of record), approximately 50 per cent of the peak stream discharge consisted of ‘old’ water. Our data also suggest that about 60 per cent of the entire 1983 hydrograph was ‘old’ water. The chemical parameters give old water contributions which are at least 10 per cent less than the isotopically-derived estimate, but they are consistent with the isotopic estimate during peak flow.     

Measurement and prediction of water movement in a field soil: The matrix-macropore dichotomy

The saturated and unsaturated flow properties of a field soil under two tillage treatments were obtained with ponded rings and disc permeameters of dissimilar radii. No difference was observed between tillage treatments but the flow properties displayed a distinct macropore-matrix dichotomy, with K changing by an order of magnitude as ψ₀ went from just - 30 mm to zero. Accurate prediction of time to incipient ponding was achieved using both numerical and analytical models calibrated with field hydraulic properties that were characteristic of the soil matrix. However, extension of the numerical model to the prediction of the wetting front development under non-ponded conditions was less accurate due to localized preferred wetting It is hypothesized that at this site, localized concentration of rainfall and hence, preferred wetting, May, occur by interception and stemflow associated with lines of standing stubble present in the original seeding slots.     

Combined use of isotopic and hydrometric data to conceptualize ecohydrological processes in a high‐elevation tropical ecosystem

Few high‐elevation tropical catchments worldwide are gauged, and even fewer are studied using combined hydrometric and isotopic data. Consequently, we lack information needed to understand processes governing rainfall–runoff dynamics and to predict their influence on downstream ecosystem functioning. To address this need, we present a combination of hydrometric and water stable isotopic observations in the wet Andean páramo ecosystem of the Zhurucay Ecohydrological Observatory (7.53 km²). The catchment is located in the Andes of south Ecuador between 3400 and 3900 m a.s.l. Water samples for stable isotopic analysis were collected during 2 years (May 2011–May 2013), while rainfall and runoff measurements were continuously recorded since late 2010. The isotopic data reveal that andosol soils predominantly situated on hillslopes drain laterally to histosols (Andean páramo wetlands) mainly located at the valley bottom. Histosols, in turn, feed water to creeks and small rivers throughout the year, establishing hydrologic connectivity between wetlands and the drainage network. Runoff is primarily composed of pre‐event water stored in the histosols, which is replenished by rainfall that infiltrates through the andosols. Contributions from the mineral horizon and the top of the fractured bedrock are small and only seem to influence discharge in small catchments during low flow generation (non‐exceedance flows < Q₃₅). Variations in source contributions are controlled by antecedent soil moisture, rainfall intensity, and duration of rainy periods. Saturated hydraulic conductivity of the soils, higher than the year‐round low precipitation intensity, indicates that Hortonian overland flow rarely occurs during high‐intensity precipitation events. Deep groundwater contributions to discharge seem to be minimal. These results suggest that, in this high‐elevation tropical ecosystem, (1) subsurface flow is a dominant hydrological process and (2) (histosols) wetlands are the major source of stream runoff. Our study highlights that detailed isotopic characterization during short time periods provides valuable information about ecohydrological processes in regions where very few basins are gauged. Copyright © 2016 John Wiley & Sons, Ltd.     

Mixing of event and pre‐event water in a shallow Entisol in sloping farmland based on isotopic and hydrometric measurements,SW China

Water percolation and flow processes in subsurface geologic media play an important role in determining the water source for plants and the transport of contaminants or nutrients, which is essential for water resource management and the development of measures for pollution mitigation. During June 2013, the dynamics of the rainwater, soil water, subsurface flows and groundwater in a shallow Entisol on sloping farmland were monitored using a hydrometric and isotopic approach. The results showed that effective mixing of rainwater and soil water occurred in hours. The rebound phenomenon of δD profiles in soils showed that most isotope‐depleted rainwater largely bypassed the soil matrix when the water saturation in the soil was high. Preferential‐flow, which was the dominant water movement pattern in the vadose zone, occurred through the whole soil profile, and infrequent piston‐flow was mainly found at 20–40 cm in depth. The interflow in the soil layer, composed of 75.2% rainwater, was only generated when the soil profile had been saturated. Underflow in the fractured mudrock was the dominant flow type in this hillslope, and outflow was dominated by base flow (groundwater flow) with a mean contribution of 76.7%. The generation mechanism of underflow was groundwater ridging, which was superimposed upon preferential‐flow composed mainly of rainwater. The quick mixing process of rainwater and soil water and the rapid movement of the mixture through preferential channels in the study soil, which shows a typical bimodal pore size distribution, can explain the prompt release of pre‐event water in subsurface flow. Water sources of subsurface flows at peak discharge could be affected by the antecedent soil water content, rain characteristics and antecedent groundwater levels. Copyright © 2016 John Wiley & Sons, Ltd.     

An infiltration model based on flow variability in macropores: development, sensitivity analysis and applications

Simulating infiltration in soils containing macropores still provides unsatisfactory results, as existing models seem not to capture all relevant processes. Recent studies of macropore flow initiation in natural soils containing earthworm channels revealed a distinct flow rate variability in the macropores depending on the initiation process. When macropore flow was initiated at the soil surface, most of the macropores received very little water while a few macropores received a large proportion of the total inflow. In contrast, when macropore flow was initiated from a saturated or nearly saturated soil layer, macropore flow rate variation was much lower. The objective of this study was to develop, evaluate, and test a model, which combines macropore flow variability with several established approaches to model dual permeability soils. We then evaluate the INfiltration–INitiation–INteraction Model (IN³M) to explore the influence of macropore flow variability on infiltration behavior by performing a sensitivity analysis and applying IN³M to sprinkling and dye tracer experiments at three field sites with different macropore and soil matrix properties. The sensitivity analysis showed that the flow variability in macropores reduces interaction between the macropores and the surrounding soil matrix and thus increases bypass flow, especially for surface initiation of macropore flow and at higher rainfall intensities. The model application shows reasonable agreement between IN³M simulations and field data in terms of water balance, water content change, and dye patterns. The influence of macropore flow variability on the hydrological response of the soil was considerable and especially pronounced for soils where initiation occurs at the soil surface. In future, the model could be applied to explore other types of preferential flow and hence to get a generally better understanding of macropore flow.     

Differences in gravity‐dominated unsaturated flow during autumn rains and snowmelt

Five tracer experiments have been performed in a coarse‐textured soil near the new main airport at Gardermoen, Norway. In two lysimeter walls, 30 and 40 measuring points form the basis for spatial moment calculations. Although experiments were performed under different meteorological conditions (autumn and snowmelt) and at two different sites, the ratios of centres of vertical mass over cumulative infiltration were of the same order of magnitude, indicating a gravity‐dominated flow. Two‐dimensional transport simulations with SUTRA (Voss, 1984), with a priori estimated input parameters and random fields of soil hydraulic properties revealed a relatively good agreement with the experimental results. Three possible sources of heterogeneity affecting the vertical displacement of solute during snowmelt were identified: variability of soil physical properties, soil surface elevations and variability of ground frost during the melting period. To obtain accurate predictions, soil heterogeneity was the most important factor to characterize for the coarse‐textured soil under consideration. Copyright © 1999 John Wiley & Sons, Ltd.     

Precipitation induced stream flow: An event based chemical and isotopic study of a small stream in the Great Plains region of the USA

A small stream in the Great Plains of USA was sampled to understand the streamflow components following intense precipitation and the influence of water storage structures in the drainage basin. Precipitation, stream, ponds, ground-water and soil moisture were sampled for determination of isotopic (D, ¹⁸O) and chemical (Cl, SO₄) composition before and after two intense rain events. Following the first storm event, flow at the downstream locations was generated primarily through shallow subsurface flow and runoff whereas in the headwaters region – where a pond is located in the stream channel – shallow ground-water and pond outflow contributed to the flow. The distinct isotopic signatures of precipitation and the evaporated pond water allowed separation of the event water from the other sources that contributed to the flow. Similarly, variations in the Cl and SO₄ concentrations helped identify the relative contributions of ground-water and soil moisture to the streamflow. The relationship between deuterium excess and Cl or SO₄ content reveals that the early contributions from a rain event to streamflow depend upon the antecedent climatic conditions and the position along the stream channel within the watershed. The design of this study, in which data from several locations within a watershed were collected, shows that in small streams changes in relative contributions from ground water and soil moisture complicate hydrograph separation, with surface-water bodies providing additional complexity. It also demonstrates the usefulness of combined chemical and isotopic methods in hydrologic investigations, especially the utility of the deuterium excess parameter in quantifying the relative contributions of various source components to the stream flow.     

Tracer and hydrometric study of preferential flow in large undisturbed soil cores from the Georgia Piedmont,USA

We studied the temporal patterns of tracer throughput in the outflow of large (30 cm diameter by 38 cm long) undisturbed cores from the Panola Mountain Research Watershed, Georgia. Tracer breakthrough was affected by soil structure and rainfall intensity. Two rainfall intensities (20 and 40 mm hr⁻¹) for separate Cl⁻ and Br⁻ amended solutions were applied to two cores (one extracted from a hillslope soil and one extracted from a residual clay soil on the ridge). For both low and high rainfall intensity experiments, preferential flow occurred in the clay core, but not in the hillslope core. The preferential flow is attributed to well‐developed interpedal macrochannels that are commonly found in structured clay soils, characteristic of the ridge site. However, each rainfall intensity exceeded the matrix infiltration capacity at the top of the hillslope core, but did not exceed the matrix infiltration capacity at the middle and bottom of the hillslope core and at all levels in the clay core. Localized zones of saturation created when rainfall intensity exceeds the matrix infiltration capacity may cause water and tracer to overflow from the matrix into macrochannels, where preferential flow occurs to depth in otherwise unsaturated soil. Copyright © 1999 John Wiley & Sons, Ltd.     

Estimating the distribution of snow water equivalent using remotely sensed snow cover data and a spatially distributed snowmelt model: A multi-resolution,multi-sensor comparison

Time series of fractional snow covered area (SCA) estimates from Landsat Enhanced Thematic Mapper (ETM+), Moderate Resolution Imaging Spectroradiometer (MODIS), and Advanced Very High Resolution Radiometer (AVHRR) data were combined with a spatially distributed snowmelt model to reconstruct snow water equivalent (SWE) in the Rio Grande headwaters (3419 km²). In this reconstruction approach, modeled snowmelt over each pixel is integrated during the period of satellite-observed snow cover to estimate SWE. Due to underestimates in snow cover detection, maximum basin-wide mean SWE using MODIS and AVHRR were, respectively, 45% and 68% lower than SWE estimates obtained using ETM+ data. The mean absolute error (MAE) of SWE estimated at 100-m resolution using ETM+ data was 23% relative to observed SWE from intensive field campaigns. Model performance deteriorated when MODIS (MAE = 50%) and AVHRR (MAE = 89%) SCA data were used. Relative to differences in the SCA products, model output was less sensitive to spatial resolution (MAE = 39% and 73% for ETM+ and MODIS simulations run at 1 km resolution, respectively), indicating that SWE reconstructions at the scale of MODIS acquisitions may be tractable provided the SCA product is improved. When considering tradeoffs between spatial and temporal resolution of different sensors, our results indicate that higher spatial resolution products such as ETM+ remain more accurate despite the lower frequency of acquisition. This motivates continued efforts to improve MODIS snow cover products.     

Potential of soil moisture observations in flood modelling: Estimating initial conditions and correcting rainfall

Rainfall runoff (RR) models are fundamental tools for reducing flood hazards. Although several studies have highlighted the potential of soil moisture (SM) observations to improve flood modelling, much research has still to be done for fully exploiting the evident connection between SM and runoff. As a way of example, improving the quality of forcing data, i.e. rainfall observations, may have a great benefit in flood simulation. Such data are the main hydrological forcing of classical RR models but may suffer from poor quality and record interruption issues. This study explores the potential of using SM observations to improve rainfall observations and set a reliable initial wetness condition of the catchment for improving the capability in flood modelling. In particular, a RR model, which incorporates SM for its initialization, and an algorithm for rainfall estimation from SM observations are coupled using a simple integration method. The study carried out at the Valescure experimental catchment (France) demonstrates the high information content retained by SM for RR transformation, thus giving new possibilities for improving hydrological applications. Results show that an appropriate configuration of the two models allows obtaining improvement in flood simulation up to 15% in mean and 34% in median Nash Sutcliffe performances as well as a reduction of the median error in volume and on peak discharge of about 30% and 15%, respectively.     

Fingered flow: The creator of sand columns in dune and beach sands

Sand columns, sand cones, sand mushrooms and other striking sand forms are frequently observed in the Dutch and German beach and dune sands. This paper aims to clarify the mechanism of sand column formation. Recently it has become evident that homogeneous beach and dune sands often become irregularly wetted by infiltrating rainwater. In otherwise dry sandy soils, wet preferential flow paths (‘fingers’) may develop. At two test sites the volumetric soil moisture content varied between 0·2 and 12·0 per cent. The wet fingers represent the premature state of sand columns. When the dry sand in between these fingers is blown away by the wind, the more resistant wet sand of the fingers will remain in its place and appear as sand columns at the surface. As a result of wind and erosive sand drifts, striking sand forms may be formed.     

Numerical investigation of the impact of uncertainties in satellite rainfall estimation and land surface model parameters on simulation of soil moisture

This study aims at evaluating the uncertainty in the prediction of soil moisture (1D, vertical column) from an offline land surface model (LSM) forced by hydro-meteorological and radiation data. We focus on two types of uncertainty: an input error due to satellite rainfall retrieval uncertainty, and, LSM soil-parametric error. The study is facilitated by in situ and remotely sensed data-driven (precipitation, radiation, soil moisture) simulation experiments comprising a LSM and stochastic models for error characterization. The parametric uncertainty is represented by the generalized likelihood uncertainty estimation (GLUE) technique, which models the parameter non-uniqueness against direct observations. Half-hourly infra-red (IR) sensor retrievals were used as satellite rainfall estimates. The IR rain retrieval uncertainty is characterized on the basis of a satellite rainfall error model (SREM). The combined uncertainty (i.e., SREM + GLUE) is compared with the partial assessment of uncertainty. It is found that precipitation (IR) error alone may explain moderate to low proportion of the soil moisture simulation uncertainty, depending on the level of model accuracy—50–60% for high model accuracy, and 20–30% for low model accuracy. Comparisons on the basis of two different sites also yielded an increase (50–100%) in soil moisture prediction uncertainty for the more vegetated site. This study exemplified the need for detailed investigations of the rainfall retrieval-modeling parameter error interaction within a comprehensive space-time stochastic framework for achieving optimal integration of satellite rain retrievals in land data assimilation systems.     

Rainfall retention and runoff generation processes in tropical mature green roof ecosystems

Rainfall retention and runoff detention are likely the most important ecosystem services provided by extensive green roofs (EGRs) that contribute to urban stormwater mitigation and management. However, the hydrological performance and runoff generation mechanisms of mature, well-established EGRs in tropical regions remain poorly understood. This study evaluated the rainfall retention, discharge detention times and processes of runoff generation in two neighbouring 20-year-old EGRs with different slopes (2° and 14° for EGRns and EGRws, respectively) and management practices in Mexico City; results were compared with those obtained in a conventional roof (CR, 2° slope). Precipitation, substrate moisture and storm runoff were continuously measured during the 2017 and 2018 rainy seasons (May–November). Results showed spatial differences in substrate properties and moisture within and between green roofs. In general, higher bulk densities and a wide range of variation in water content characterized the bare substrate areas compared to those below vegetation. Greatest increases in substrate moisture and storm runoff were observed in the steeper green roof. Subsurface flow was the dominant process controlling the amount and timing of stormflow in the EGRs. The occurrence of saturation excess overland flow was small and detected when large rain events were preceded by high wetness conditions. The main factors influencing the hydrological responses of the green roofs were their substrate hydrophysical properties, related mostly to vegetation cover, management and age, and to much lesser extent to slope and substrate depth. On average, rainfall retention was ~60% in the EGRs with significantly longer delays and prolonged runoff times (100 and 340 min, respectively) compared to CR (3%, 20 min, and 258 min, respectively). Overall, these findings highlight the potential of EGRs in reducing stormflow and peak discharges for most rainfall in Mexico City, and thus mitigating the risk of saturation and overflow of urban drainages.     

The effect of antecedent moisture conditions on sediment and phosphorus loss during overland flow: Mahantango Creek catchment,Pennsylvania, USA

The loss of P in overland flow from most cultivated soils is controlled by erosion, and in‐turn soil moisture. We evaluated the effect of soil moisture on erosion and P transport in overland flow by applying rainfall (7 cm h^?1) to packed soil boxes (1 m long and 0·15 m wide) and field plots (1 and 10 m long by 1 m wide) of silt loams in a central Pennsylvania (USA) catchment. Flow from packed soil boxes took longer to initiate as antecedent soil moisture decreased from field capacity (2 min) to air dried (8 to 9 min). Even in the more complex field plots (i.e. soil heterogeneity and topography), the wetter site (1 by 10 m plot; 70% field capacity) produced flow more quickly (3 min) and in greater volume (439 L) than the drier site (1 by 10 m plot; 40% field capacity, 15 min, and 214 L, respectively). However, less suspended sediment was transported from wetter soil boxes (1·6 to 2·5 g L^?1) and field plots (0·9 g L^?1) than drier boxes (2·9 to 4·2 g L^?1) and plots (1·2 g L^?1). Differences are attributed to their potential for soil aggregate breakdown, slaking and dispersion, which contribute to surface soil sealing and crusting, as dry soils are subject to rapid wetting (by rainfall). During flow, selective erosion and antecedent moisture conditions affected P transport. At field capacity, DRP and PP transport varied little during overland flow. Whereas P transport from previously dry soil decreased rapidly after the initiation of flow (6 to 1·5 mg TP L^?1), owing to the greater slaking and dispersion of P‐rich particles into flow at the beginning than end of the flow event. These results indicate that soil moisture fluctuations greatly effect erosion and P transport potential and that management to decrease the potential for loss should consider practices such as conservation tillage and cover crops, particularly on areas where high soil P and erosion coincide. Copyright © 2002 John Wiley & Sons, Ltd.     

Runoff‐generating processes in hydromorphic soils on a plot scale: free gravity‐driven versus pressure‐controlled flow

Draining soil water is an important runoff generator. This study aims to describe runoff‐generating processes on a plot scale (1 m²) in hydromorphic soils with different initial soil water contents. We irrigated 16 hydromorphic soils in the northern Pre‐Alps in Switzerland and recorded the variations in water content with time domain reflectometry (TDR) at five different depths per plot. Sprinkling was repeated three times at approximately 23‐h intervals and lasted for 1 h with a volume flux density of 70 mm h^?1. The comparison between the measured water content of the drainages with two physically based models revealed which of the flow processes dominated during water recessions. We distinguished between vertical drainage, lateral outflow and infiltration without drainage. Approximately 45% of all recorded time series of soil water content did not drain within approximately 20 h after the end of irrigation, about 25% drained laterally and 10% of the outflow was vertical. The drainage of the remaining 20% was the result of both lateral and vertical water flow (≈12%), or was not interpretable with the approaches applied (≈8%). Vertical flow was only observed in layers without any or with just a few hydromorphic features. Lateral draining horizons had approximately half the storage capacity and amplitude of water recession of those with vertical flow. Vertical flow was only observed in the upper soil horizons. Thus, vertical flow transmitted water to layers with lateral outflow and did not delay runoff by deep percolation. Increasing initial soil moisture increased the percentage of water content recordings according to a lateral outflow slightly, while vertical flow was less frequent. Copyright © 2010 John Wiley & Sons, Ltd.     

Significance of sediment transport processes during piedmont floods: the 2005 flood events in Switzerland

A long‐lasting rainstorm event from 20 to 22 August 2005 affected a large part of the northern Alps and Prealps in Switzerland. It resulted in elevated discharges and flooding in many headwater catchments and mountain rivers. The associated geomorphic processes included shallow landslides, deep‐seated slope instabilities, debris flows, and fluvial sediment transport. In many parts of the affected areas human activities are important, including many buildings, traffic lines and other infrastructure. In the steeper parts, geomorphic processes were mainly responsible for flow overtopping and sediment deposition both in and outside of the channel network. In the lower parts, lateral erosion and exceedance of the channel discharge capacity were the main reasons for morphologic channel modification and flooding. Sediment‐related processes contributed a lot to the overall damage. Copyright © 2015 John Wiley & Sons, Ltd.     

A comparison of flow intensities in alluvial rivers: characteristics and implications for modelling flow processes

This paper compares flow intensity data obtained with different instruments from a variety of fluvial environments. It examines associations between the root-mean-square of longitudinal velocity fluctuations (flow intensity), local mean velocity, relative depth, and boundary resistance. Results indicate systematic differences in the behaviour of flow intensity which scale with respect to position in the boundary layer (deep sand-bedded rivers), boundary grain resistance (shallow river environments with coarse beds), and possibly form resistance (shallower sand-bedded rivers). Preliminary approaches to prediction and modelling of variations in flow intensity are suggested based upon linear regression relationships. Intensity values are also compared with theoretical and empirical limits to the use of Taylor's substitution, which allows time and frequency properties of a single-point velocity time series to be used to yield a flow length scale. In general, limits are exceeded in all environments for near-boundary flow measurements, but are met for y/d > 0·3 in most cases in sand-bed rivers, and for y/d > 0·4 in some gravel-bed environments. © 1998 John Wiley & Sons, Ltd.     

Assimilation of streamflow and in situ soil moisture data into operational distributed hydrologic models: Effects of uncertainties in the data and initial model soil moisture states

We assess the potential of updating soil moisture states of a distributed hydrologic model by assimilating streamflow and in situ soil moisture data for high-resolution analysis and prediction of streamflow and soil moisture. The model used is the gridded Sacramento (SAC) and kinematic-wave routing models of the National Weather Service (NWS) Hydrology Laboratory’s Research Distributed Hydrologic Model (HL-RDHM) operating at an hourly time step. The data assimilation (DA) technique used is variational assimilation (VAR). Assimilating streamflow and soil moisture data into distributed hydrologic models is new and particularly challenging due to the large degrees of freedom associated with the inverse problem. This paper reports findings from the first phase of the research in which we assume, among others, perfectly known hydrometeorological forcing. The motivation for the simplification is to reduce the complexity of the problem in favour of improved understanding and easier interpretation even if it may compromise the goodness of the results. To assess the potential, two types of experiments, synthetic and real-world, were carried out for Eldon (ELDO2), a 795-km² headwater catchment located near the Oklahoma (OK) and Arkansas (AR) border in the U.S. The synthetic experiment assesses the upper bound of the performance of the assimilation procedure under the idealized conditions of no structural or parametric errors in the models, a full dynamic range and no microscale variability in the in situ observations of soil moisture, and perfectly known univariate statistics of the observational errors. The results show that assimilating in situ soil moisture data in addition to streamflow data significantly improves analysis and prediction of soil moisture and streamflow, and that assimilating streamflow observations at interior locations in addition to those at the outlet improves analysis and prediction of soil moisture within the drainage areas of the interior stream gauges and of streamflow at downstream cells along the channel network. To assess performance under more realistic conditions, but still under the assumption of perfectly known hydrometeorological forcing to allow comparisons with the synthetic experiment, an exploratory real-world experiment was carried out in which all other assumptions were lifted. The results show that, expectedly, assimilating interior flows in addition to outlet flow improves analysis as well as prediction of streamflow at stream gauge locations, but that assimilating in situ soil moisture data in addition to streamflow data provides little improvement in streamflow analysis and prediction though it reduces systematic biases in soil moisture simulation.     

A modelling framework to simulate field‐scale nitrate response and transport during snowmelt: The WINTRA model

Modelling nutrient transport during snowmelt in cold regions remains a major scientific challenge. A key limitation of existing nutrient models for application in cold regions is the inadequate representation of snowmelt, including hydrological and biogeochemical processes. This brief period can account for more than 80% of the total annual surface runoff in the Canadian Prairies and Northern Canada and processes such as atmospheric deposition, overwinter redistribution of snow, ion exclusion from snow crystals, frozen soils, and snow‐covered area depletion during melt influence the distribution and release of snow and soil nutrients, thus affecting the timing and magnitude of snowmelt runoff nutrient concentrations. Research in cold regions suggests that nitrate (NO₃) runoff at the field‐scale can be divided into 5 phases during snowmelt. In the first phase, water and ions originating from ion‐rich snow layers travel and diffuse through the snowpack. This process causes ion concentrations in runoff to gradually increase. The second phase occurs when this snow ion meltwater front has reached the bottom of the snowpack and forms runoff to the edge‐of‐the‐field. During the third and fourth phases, the main source of NO₃ transitions from the snowpack to the soil. Finally, the fifth and last phase occurs when the snow has completely melted, and the thawing soil becomes the main source of NO₃ to the stream. In this research, a process‐based model was developed to simulate hourly export based on this 5‐phase approach. Results from an application in the Red River Basin of southern Manitoba, Canada, shows that the model can adequately capture the dynamics and rapid changes of NO₃ concentrations during this period at relevant temporal resolutions. This is a significant achievement to advance the current nutrient modelling paradigm in cold climates, which is generally limited to satisfactory results at monthly or annual resolutions. The approach can inform catchment‐scale nutrient models to improve simulation of this critical snowmelt period.