Quick Flow Component Research Articles

Strength and Memory of Precipitation's Control Over Streamflow Across the Conterminous United States

AbstractHow precipitation (P) is translated into streamflow (Q) and over what timescales (i.e., “memory”) is difficult to predict without calibration of site‐specific models or using geochemical approaches, posing barriers to prediction in ungauged basins or advancement of general theories. Here, we used a data‐driven approach to identify regional patterns and exogenous controls on P–Q interactions. We applied an information flow analysis, which quantifies uncertainty reduction, to a daily time series of P and Q from 671 watersheds across the conterminous United States. We first demonstrated that information transfer from P to Q primarily reflects the quickflow component of water‐budgets, based on a watershed model. Readily quantifiable information flows show a functional relationship with model parameters, suggesting utility for model calibration. Second, applied to real watersheds, P–Q information flows exhibit seasonally varying behavior within regions in a manner consistent with dominant runoff generation mechanisms. However, the timing and the magnitude of information flows also reflect considerable subregional heterogeneity, likely attributable to differences in watershed size, baseflow contributions, and variation in aerial coverage of preferential flow paths. A regression analysis showed that a combination of climate and watershed characteristics are predictive of P–Q information flows. Though information flows cannot, in most cases, uniquely determine dominant runoff mechanisms, they provide a means to quantify the heterogeneous outcomes of those mechanisms within regions, thereby serving as a benchmarking tool for models developed at the regional scale. Last, information flows characterize regionally specific ways in which catchment connectivity changes from the wet to dry season.

Comparison of Bayesian, k-Nearest Neighbor and Gaussian process regression methods for quantifying uncertainty of suspended sediment concentration prediction

Suspended sediment transport in river system is a complex process influenced by many factors that their interactions lead to nonlinear and high scatter of concentration-discharge relationships. This makes the model prediction subject to high uncertainty and providing one value as the model prediction is somehow useless and cannot provide adequate information about the model accuracy and associated uncertainty. Current study compares the efficiency of Bayesian (i.e. Bayesian segmented linear regression (BSLR) and Bayesian linear model (BLR)), Gaussian Process Regression (GPR) and k-Nearest Neighbor (k-NN) in quantifying uncertainty of the suspended sediment concentration prediction in three watersheds namely Arazkoseh, Oghan and Jajrood located in Iran. Three input combinations including, contemporary discharge, slow and quick flow components and contemporary, one and two antecedent days discharge, were used. The BSLR model was able to identify threshold value, furthermore, pre-threshold and post-threshold slopes of BSLR model indicated that for Arazkoseh watershed channel and for Oghan and Jajrood watersheds, upland area are dominate sediment sources. In all three studied cases, given prediction interval width and the percent of enclosed observed data by prediction interval, k-NN model provided more reliable prediction interval. Moreover, separation stream flow into slow and quick flow components lead to improved performance of GPR and k-NN models in the studied watersheds, and the best results for Arazkoseh and Oghan watersheds were obtained when slow and quick flow components were used as the model input.

Magnesium as Environmental Tracer for Karst Spring Baseflow/Overflow Assessment—A Case Study of the Pertuso Karst Spring (Latium Region, Italy)

Following a previous research carried out on the same site, this paper presents the update of the Mg2+ based method for the estimation of Pertuso Spring discharge, located in Central Italy. New collected data confirmed the validity of the proposed model and the conservative behaviour of Mg2+ for groundwater related to the Pertuso Spring aquifer. Further analysis allowed to obtain a local linear relationship between magnesium concentration and total spring discharge (including exploitation rate), regardless of the mixing model proposed with the Aniene River. As regards two samples which fall out of the linear relationship and could have been detected as “outliers”, more in-depth data processing and sensitivity analyses revealed that the lowering in magnesium, at equal discharges, is determined by the appearance of the quick-flow component, less mineralized and related to storm events. Results showed that under specific conditions, related to the absence or presence of previous intense rainfall events, Mg2+ could be effectively a useful tracer for separating spring conduit flow (overflow) from diffuse flow (baseflow) within the karst aquifer.

Recursive Bayesian Estimation of Conceptual Rainfall‐Runoff Model Errors in Real‐Time Prediction of Streamflow

AbstractConceptual rainfall‐runoff models account for the spatial dynamics of hydrological processes in a basin using simple spatially lumped storage‐flow relations. Such rough approximations introduce model errors that are often difficult to characterize. Here, we develop and apply a methodology that recursively estimates and accounts for model errors in real‐time streamflow prediction settings by adding time‐dependent random noise to the internal states (storages) of the hydrological model. Magnitude of the added noise depends on a precision (inverse variance) parameter that is estimated from rainfall‐runoff data. A recursive Bayesian technique is used for estimation: posteriors of hydrological parameters and states are updated through time with an ensemble Kalman filter, whereas the posterior of the precision parameter is updated recursively using a novel gamma density approximation technique. Applying this algorithm to different model error scenarios allows identification of the main source of model errors. The methodology is applied to short‐term streamflow prediction with the Hymod rainfall‐runoff model in a semi‐cold, semi‐humid basin in Iran. Results show that (i) streamflow prediction in this snow‐dominated basin is more affected by model errors in the slow flow than the quick flow component of the model, (ii) accounting for model errors in the slow flow component improves both low and high flow predictions, and (iii) predictive performance further improves by accounting for Hymod parameter uncertainty in addition to model errors. Overall, accounting for model errors increased Nash‐Sutcliffe efficiency (by 1–5%), reduced mean absolute error (by 2–43%), and improved probabilistic predictive performance (by 50–80%).

Hydrogeological Investigation of a Karst Aquifer System

This study aims to present the hydrogeological and hydrochemical characteristics of Paradisos Karst aquifer system (PAS), Northeast Greece. The average area of the PAS is estimated according to Thornthwaite and Mather (1957) (T-M) procedure at 74.2 km2. The karst system, which is largely fault-controlled, discharges through three permanent karst springs: KS15, KS20 and KS21. The average discharge rate was calculated for two of them at 0.726 m3/s. The time series analysis shows that: (a) the memory effect for the PAS is extremely high; and (b) the karst system is a poorly developed with high storage capacity and absence of a quick flow component. The chemical type for the majority of the groundwater samples is of Ca-HCO3 type. The chemical properties of the groundwaters are strongly influenced by alkaline earth metals and weak acids. The presence (sporadically) of high NH4 + content and slightly increased nitrate, ammonium and phosphate values in springs KS20 and KS21 is attributed to the presence of anthropogenic sources, especially septic tanks and agriculture. All studied groundwaters are oversaturated in calcite and aragonite, which suggests a mechanism of diffuse flow. The morphology and geology of the PAS catchment area, and data from hydrographs and chemographs, show that the hydrologic system is dominantly of diffuse flow. The use of time series analysis combined with classic hydrogeological techniques (i.e., degree of karstification, T-M procedure, chemographs and saturation indices) can be used in order to successfully characterize a karst system and achieve a sustainable integrated water management in it.

Non-free ionic transport of sodium, magnesium, and calcium in streams of two adjacent headwater catchments with different vegetation types in Japan

Sodium (Na), magnesium (Mg), calcium (Ca) are usually believed to occur mostly as free ions in the fresh water and consequently little is known about their chemical species. To understand the importance of non-free ionic fractions (NIF) of major metals in freshwater streams, Na, Mg, Ca, silicon (Si), and fulvic acid-like materials (FAM) were measured in streams of mountainous adjacent headwater catchments dominated by different vegetation types (planted evergreen coniferous forest and natural deciduous broadleaf forest). During both no rainfall periods and rainstorms, the proportion of NIF relative to total elements was lower in the coniferous catchment than in the deciduous catchment, although it sometimes accounted for half or more of the total concentrations of Na, Mg, and Ca in both catchments. The solubility of metal compounds was higher than the measured maximum concentrations of Na+, Mg2+, and Ca2+ to the extent that inorganic bonding was hardly possible. During no rainfall periods when FAM was slightly produced into the streams, the fluxes of NIF and Si were highly correlated (r>0.92, p<0.0001, n=30) in both catchments. During a small rainstorm, the flux of NIF correlated weakly with that of Si but did not correlate with that of FAM in both catchments. In contrast, during a heavy rainstorm, the flux of NIF correlated strongly (r⩾0.83, p<0.0001, n=26) with that of FAM in the deciduous catchment where relatively deep soil water compared to near-surface water was the predominant component of stream water. However, during the heavy rainstorm in the coniferous catchment, only the flux of NIF originated in the quick-flow component (i.e., surface or near-surface water) in stream water (ΔNIF) correlated strongly (r⩾0.81, p<0.0001, n=22) with that of FAM. These findings imply that heavy rainstorms may enhance the bonding of the major metals with humic substances mainly in the deciduous catchment; and also exhibit that, in the headwater catchments, both water flow pathways resulted from the different vegetation types play a very important role to promote the bonding of major metals with humic substances in stream water.

Development of a modular streamflow model to quantify runoff contributions from different land uses in tropical urban environments using Genetic Programming

The decrease of pervious areas during urbanization has severely altered the hydrological cycle, diminishing infiltration and therefore sub-surface flows during rainfall events, and further increasing peak discharges in urban drainage infrastructure. Designing appropriate waster sensitive infrastructure that reduces peak discharges requires a better understanding of land use specific contributions towards surface and sub-surface processes. However, to date, such understanding in tropical urban environments is still limited. On the other hand, the rainfall–runoff process in tropical urban systems experiences a high degree of non-linearity and heterogeneity. Therefore, this study used Genetic Programming to establish a physically interpretable modular model consisting of two sub-models: (i) a baseflow module and (ii) a quick flow module to simulate the two hydrograph flow components. The relationship between the input variables in the model (i.e. meteorological data and catchment initial conditions) and its overall structure can be explained in terms of catchment hydrological processes. Therefore, the model is a partial greying of what is often a black-box approach in catchment modelling. The model was further generalized to the sub-catchments of the main catchment, extending the potential for more widespread applications. Subsequently, this study used the modular model to predict both flow components of events as well as time series, and applied optimization techniques to estimate the contributions of various land uses (i.e. impervious, steep grassland, grassland on mild slope, mixed grasses and trees and relatively natural vegetation) towards baseflow and quickflow in tropical urban systems. The sub-catchment containing the highest portion of impervious surfaces (40% of the area) contributed the least towards the baseflow (6.3%) while the sub-catchment covered with 87% of relatively natural vegetation contributed the most (34.9%). The results from the quickflow module revealed average runoff coefficients between 0.12 and 0.80 for the various land uses and decreased from impervious (0.80), grass on steep slopes (0.56), grass on mild slopes (0.48), mixed grasses and trees (0.42) to relatively natural vegetation (0.12). The established modular model, reflecting the driving hydrological processes, enables the quantification of land use specific contributions towards the baseflow and quickflow components. This quantification facilitates the integration of water sensitive urban infrastructure for the sustainable development of water in tropical megacities.

Quantifying faecal indicator organism hydrological transfer pathways and phases in agricultural catchments

Faecal indicator organisms (FIOs) can impact on water quality and pose a health and environmental risk. The transfer of FIOs, such as Escherichia coli (E. coli), from land to water is driven by hydrological connectivity and may follow the same flowpaths as nutrients, from agricultural and human sources. This study investigated E. coli transfer in two catchment areas with high source and transport pressures. These pressures were: organic phosphorus (P) loading; human settlement; conduits and fissures in a grassland karst area; and clay rich and impermeable soils in a mixed arable area. The occurrence of E. coli and its transport pathways, along with the pathways of nutrients, were studied using a combination of targeted FIO sampling, during different hydrological phases and events, and high resolution nutrient analysis. The quick flow component in both catchments was found to be a more potent vector for E. coli, and was coincident with the total P flowpaths using a P Loadograph Recession Analysis (LRA). The karst grassland catchment was found to be a transport limited system and the mixed arable catchment a source limited system. Hence, despite the grassland catchment being a potentially higher FIO source, the E. coli loads leaving the catchment were low compared to the mixed arable catchment. E. coli load whole-event comparisons also indicated that the grassland karst transfers tended to be much lower on falling phases of runoff, while the arable catchment, over greywacke and mudstone geology, showed little change between the phases. Furthermore, the arable catchment showed asymptotic decline of sustained E. coli loads towards low flows, which may be indicative of chronic point sources. These results indicate the dominance of transport mechanisms over source mechanisms for mass E. coli loads and also chronic loads during low flow. These will be important considerations for risk assessment and mitigation.

Flow paths and phosphorus transfer pathways in two agricultural streams with contrasting flow controls

AbstractIn this paper, we analyse 4 years of data from simultaneous high‐frequency monitoring of streamflow and phosphorus (P) concentration. This was carried out to investigate hydrological flow paths and P transfer pathways from diffuse sources in two intensively farmed river catchments (~10 km2) with contrasting flow controls and dominating flow paths. Catchment scale P loss was viewed on an annual and event flow basis and related to hydrological flow paths. A grassland catchment with mostly poorly drained soils, and a higher Q10:Q90 ratio (60 compared with 24), had three times higher annual P loss than an arable catchment with mostly well‐drained soils (1.04 compared with 0.34 kg TP ha−1) despite the arable catchment having larger areas with high soil P status and more discharge. Neither of the catchments indicated supply limitations. The magnitude of the P losses from the two catchments was not defined by land use, source pressure or discharge volume but rather by more basic rainfall‐to‐runoff partitioning influences that determine proportions of quickflow and slowflow. There were larger differences between the years than between the catchments, and the P loss of the arable catchment appeared more sensitive to climate. The results confirmed the need to manage the quickflow components of runoff to moderate P transfers. Therefore, in order to further reduce diffuse pollution it may be necessary to account for the contrast in hydrological function before or in addition to any of the other factors known to influence P losses from catchments (such as soil P and land use). Schemes designed to attenuate diffuse P after mobilization from soil surfaces can then be targeted (and resourced) more effectively. Copyright © 2014 John Wiley &amp; Sons, Ltd.

Use of hydrologic time-series data for identification of hydrodynamic function and behavior in a karstic water system in China

Karst aquifers are characterized by high heterogeneity and spatial variability of hydrogeological parameters. Time-series analysis of rainfall and discharge (as input and output functions), including correlation, cross-amplitude, phase and coherency, was applied to the Houzhai karstic water system in Guizhou Province, southwest China, in order to study the function, hydrodynamic behavior and hydraulic properties of the aquifer system. Autocorrelation and cross-correlation of the variables showed that the degree of sensitivity of the system to the rainfall input signal decreased gradually from the upstream to the downstream sections, but the memorizing action increased gradually. Analysis of the phase function showed that there is a lag in the discharge response to the rainfall input signal. The lag time increased gradually from the upstream to the downstream sections, and the degree of the linearity upstream is higher than downstream. The quick-flow and slow-flow components were also divided in the cross-amplitude analysis. The results showed that about 60% of spring discharge in the upstream section was quick flow. The quick-flow component downstream reduced to only 5%. The main control factors of the karst-system response are the karst geomorphology and the inner origination structure of the karstic multi-medium.

Spring hydrographs as indicators of droughts in a karst environment

Summary Spring discharges of a large karst system (Campania, Southern Italy) have been analysed to determine if they were related to rainfall and periods of poor recharge. Due to the Mediterranean climate and hydrogeological conditions of the aquifers, the spring hydrographs are generally characterised by one annual smoothed peak during spring season and a negligible quick flow component. Insufficient recharge due to poor annual rainfall results in flat spring hydrographs (with no peak) that indicate a continuously decreasing discharge. Flat spring hydrographs reveal a drought, which is characterised by a prolonged shortage of water that induces a reduction in discharge during the following year as well. Droughts also appear to be induced by consecutive years with lower than average rainfall. The historical data have shown that each hydrological year depends on the previous year because annual rainfall series do not have a random character and aquifers have a “memory effect”, which results in spring discharges amplifying the effect of poor rainfall. Due to a long historical series and the specific karst spring regime, a flat hydrograph can be forecast as early as winter, thereby providing a useful tool for water management. The time-lag between rainfall and spring discharges has been utilised to determine the amount of rainfall required to avoid a drought.

Using Nitrate to Quantify Quick Flow in a Karst Aquifer

In karst aquifers, contaminated recharge can degrade spring water quality, but quantifying the rapid recharge (quick flow) component of spring flow is challenging because of its temporal variability. Here, we investigate the use of nitrate in a two-endmember mixing model to quantify quick flow in Barton Springs, Austin, Texas. Historical nitrate data from recharging creeks and Barton Springs were evaluated to determine a representative nitrate concentration for the aquifer water endmember (1.5 mg/L) and the quick flow endmember (0.17 mg/L for nonstormflow conditions and 0.25 mg/L for stormflow conditions). Under nonstormflow conditions for 1990 to 2005, model results indicated that quick flow contributed from 0% to 55% of spring flow. The nitrate-based two-endmember model was applied to the response of Barton Springs to a storm and results compared to those produced using the same model with delta(18)O and specific conductance (SC) as tracers. Additionally, the mixing model was modified to allow endmember quick flow values to vary over time. Of the three tracers, nitrate appears to be the most advantageous because it is conservative and because the difference between the concentrations in the two endmembers is large relative to their variance. The delta(18)O-based model was very sensitive to variability within the quick flow endmember, and SC was not conservative over the timescale of the storm response. We conclude that a nitrate-based two-endmember mixing model might provide a useful approach for quantifying the temporally variable quick flow component of spring flow in some karst systems.

Parameter estimation in semi‐distributed hydrological catchment modelling using a multi‐criteria objective function

AbstractOutput generated by hydrologic simulation models is traditionally calibrated and validated using split‐samples of observed time series of total water flow, measured at the drainage outlet of the river basin. Although this approach might yield an optimal set of model parameters, capable of reproducing the total flow, it has been observed that the flow components making up the total flow are often poorly reproduced. Previous research suggests that notwithstanding the underlying physical processes are often poorly mimicked through calibration of a set of parameters hydrologic models most of the time acceptably estimates the total flow. The objective of this study was to calibrate and validate a computer‐based hydrologic model with respect to the total and slow flow. The quick flow component used in this study was taken as the difference between the total and slow flow. Model calibrations were pursued on the basis of comparing the simulated output with the observed total and slow flow using qualitative (graphical) assessments and quantitative (statistical) indicators. The study was conducted using the Soil and Water Assessment Tool (SWAT) model and a 10‐year historical record (1986–1995) of the daily flow components of the Grote Nete River basin (Belgium). The data of the period 1986–1989 were used for model calibration and data of the period 1990–1995 for model validation. The predicted daily average total flow matched the observed values with a Nash–Sutcliff coefficient of 0·67 during calibration and 0·66 during validation. The Nash–Sutcliff coefficient for slow flow was 0·72 during calibration and 0·61 during validation. Analysis of high and low flows indicated that the model is unbiased. A sensitivity analysis revealed that for the modelling of the daily total flow, accurate estimation of all 10 calibration parameters in the SWAT model is justified, while for the slow flow processes only 4 out of the set of 10 parameters were identified as most sensitive. Copyright © 2007 John Wiley &amp; Sons, Ltd.

A technique for deriving an average event unit hydrograph from streamflow—only data for ephemeral quick-flow-dominant catchments

A technique is presented for empirical and parametric estimation of an average event unit hydrograph response curve solely from measured streamflow data for use in rainfall-runoff models with a focus on modelling flow for natural resource management. As the technique does not require rainfall data, the unit hydrograph can be derived solely from streamflow data at a temporal resolution appropriate for the response of the catchment. The response curves derived for a number of stations show a power law relationship of decay in flow after peak that can be described using a three-parameter function. The approach is best suited to ephemeral streamflow regimes dominated by surface and near-surface runoff (quick-flow component) where flow events are well-separated and largely independent of each other. Analysis of the derived unit hydrograph for 28 quick-flow-dominated catchments illustrates the range of parameter values obtained in fitting the power law.

Epikarst storage in a karst aquifer: a conceptual model based on isotopic data, Milandre test site, Switzerland

The Milandre test site is a karst aquifer characterized by diffuse infiltration, a well developed conduit network, and several tributaries feeding an underground river. Field data include discharge rate measurements, stable isotopes, weekly rainfall and spring-water isotope sampling, and detailed isotope sampling during three flood events. Flood sampling was carried out at several tributaries corresponding to conduit flow, vadose flow and seepage flow. Weekly sampling showed a strong buffering of the rainfall isotopic signal at the spring. This attenuation suggests an important mixing reservoir in the system. Flood events showed highly peaking hydraulic responses but buffered rain isotope responses. These results indicate that the soil and epikarst sub-systems have an important storage capacity. A conceptual model of flow and transport in the soil and epikarst zone is proposed: Soil plays an important role in mixing due to the presence of capillary water storage. Consequently dampened concentrations reach the epikarst despite a rapid hydraulic response. The epikarst acts as the storage element and distributes water as either a base flow component or a quick flow component. When recharge exceeds a given threshold, excess infiltrated water bypasses the soil and epikarst and reaches the saturated zone as fresh flow. Based on this model, the significance of phreatic storage is thought to be limited, at least in Milandre test site. Hence the saturated zone is seen mainly as a transmissive zone through its well developed conduit network.

Composite transfer functions for karst aquifers

Linear transfer functions have been extensively used in hydrological studies. Generally, we support this conclusion: rainfall-runoff models based on the convolution between rainfall rates and a nonparametric transfer function (NTF) are not successful at simulating karst spring discharges during long recession periods. The tails of identified transfer functions have irregular shapes and they are not accurate physical representation of the transport through a karst system. Irregularities are the result of unavoidable errors in input and output time series and simplifications made by considering the system as linear and time invariant. This paper deals with a new form of the transfer functions for karst aquifers, the so-called composite transfer function (CTF). The CTF simulates discharges by two transfer functions adapted for the quick flow and the slow flow hydrograph component modeling. NTF is responsible for the quick flow component. The slow flow component is modeled by a parametric transfer function that is an instantaneous unit hydrograph mathematically formulated and defined from a conceptual model. By using the CTF, the irregular shape of the tail of the identified transfer function can be avoided, and the simulation of long recession periods as well as the simulation of a complete hydrograph becomes more successful. The NTF, the Nash model, the Zoch model and other similar conceptual models can be considered separately as simplified forms of the CTF. The rainfall-runoff model based on the convolution between rainfall rates and the CTF was tested on the Jadro Spring in Croatia. The results of the application are compared with the results obtained by applying NTFs independently.

Understanding runoff processes using a watershed model—a case study in the Western Ghats in South India

The wet tropical Western Ghat Mountain ranges in South India present an interesting combination of meteorological and physical characteristics. The results of a watershed model analysis carried out to understand the catchment response and the relative importance of different runoff processes in the region are reported in this paper. A lumped parameter model simulating saturated source area runoff, lateral flow through pipes and the saturated zone groundwater flow, has been developed assuming that source area runoff is the only quickflow component. The model has been calibrated on seven catchments using sufficiently long records of daily data. A wide range of tests has been used to show that the model performs reliably. The influence of catchment characteristics on the relative importance of the flow components and the catchment response has been studied. The model simulations have been interpreted to infer that the pipeflow contributions augment the contributions of source area runoff to stream quickflow. Suggestions for further research in the area are given, based on the inferences drawn.

Clearcutting effects on streamflow in coppiced Eucalyptus globulus stands in Portugal

Two Eucalyptus globulus catchments in Central Portugal were instrumented to gain some understanding of the hydrological effects of coppicing forest stands of this fast-growing species. Suspected increases in total flow, quick flow and delayed flow after clearcutting were tested statistically using the R-Student statistic for outlier detection. Under normal climatic conditions, significant annual flow increases were found only during the water-year immediately after clearcutting. However, when the water-year following harvesting was extremely dry, no flow increase occurred during that year although flow increases were found during the second water-year. Nevertheless, owing to the rapid regrowth of the stands, the hydrological effects of clearcutting were always short lived. Depending on the dominant runoff-generating processes, increases in total flow were due to increases in both the quick and delayed flow components or to increases only in the quick flow component.

Darcian flood generation modelling

This paper discusses experience of the use of Darcy's law for porous media flow in the context of practical modelling of flood generation. Specific drawbacks to the straightforward use of single porosity Darcian formulations for flood generation are discussed with reference to modelled examples. A quick flow component is frequently found to be needed to supplement modelled porous medium flow to match flashy stream hydrographs and observed rates of change of flow. This can in cases be justified in the field with reference to underdrainage, surface flow, natural piping and the occurrence of macropores. Rough limits are given for pure Darcian hydrograph rises under specified conditions: a range of simple methods is suggested for modelling the addition of a fast flow component where it is appropriate.

Hydrologic modeling Of New England river basins using radar rainfall data

Quantitative weather radar measurements of rainfall provide the input to a hydrologic forecast model designed to use the full spatial resolution of the radar data. The gridded model, which incorporates a detailed map of the stream network, is based on a simple kinematic representation of the river basin response. Only two parameters control the shape of the hydrograph: the velocity characteristic of subsurface flow to the nearest stream, and the streamflow velocity itself. Comparisons are made between model hydrograph forecasts and observed streamflow records for the Souhegan (440 km2), and the Squannacook (160 km2) river basins. A single Z‐R relation was used for all storms (Z = 230R1.4), except for one case with strong convection (Z = 400R1.3). A linear scaling of the volume of the radar‐derived storm rainfall produces reasonable agreement between the predicted and observed hydrographs. The volume scale factor, which varies from 20% or less in the summer to 100% in the spring, is consistent with the climatological mean monthly rainfall‐runoff ratio. In the two basins studied, for hydrograph peaks of moderate amplitude, overland flow and other quickflow components of the hydrograph are not generally observed. The hillslope response is modeled by a single characteristic subsurface flow velocity (2 × 10−3 m s−1), with a streamflow velocity of 0.6 m s−1 in the Souhegan, (0.3 m s−1 in the Squannacook). The results suggest that models which utilize the basin geometry and rainfall data to a maximum, but which otherwise contain few parameters, can be very successful.