Regional Regression Equations Research Articles

Performance-weighted methods for estimating monthly streamflow at ungauged sites

► We provide a systematic approach to evaluating streamflow record estimators. ► We evaluate three techniques for estimating streamflow records at ungauged sites. ► More advanced methods are shown to be hampered by parameter uncertainty. ► We present a blended technique for more-accurately estimating streamflow records. ► These results highlight the potential value of blended estimators. Records of natural streamflow are essential to nearly all water resources projects. In spite of the wide range of available techniques for estimating monthly streamflow records at ungauged sites, few studies have sought to compare their merits. Precedent has dictated the nearly ubiquitous use of the drainage-area ratio (DAR) technique. We evaluate several monthly streamflow information-transfer methods across nearly 1300 minimally-impacted streamflow gages in the continental US from the Hydroclimatic Data Network (HCDN). The DAR method is shown to be quite robust; however, other methods of standardizing streamflow, such as standardizing flows by means (SM) or normalizing flows with means and standard deviations (SMS), are shown to be superior for a subset (∼42%) of sites. The performance of SM and SMS, relative to DAR is shown to be generally correlated with the at-site aridity ratio. We introduce a performance-weighted transfer method that weights each estimator in proportion to its performance, as predicted by basin descriptors including at-site precipitation, aridity ratio, drainage area and ratios of those descriptors between the ungauged and index site. Our blended estimators showed promising results, exhibiting greater Nash–Sutcliffe efficiencies than DAR at 60% of sites considered. Our national comparative assessment demonstrates the merits and disadvantages of several techniques, including the concept of performance weighting. We also introduce regional hydroclimatic regression equations for estimating the mean and variance of monthly streamflows at ungauged sites in the conterminous US.

Annual estimates of water and solute export from 42 tributaries to the Yukon River

Annual export of 11 major and trace solutes for the Yukon River is found to be accurately determined based on summing 42 tributary contributions. These findings provide the first published estimates of tributary specific distribution of solutes within the Yukon River basin. First, we show that annual discharge of the Yukon River can be computed by summing calculated annual discharges from 42 tributaries. Annual discharge for the tributaries is calculated from the basin area and average annual precipitation over that area using a previously published regional regression equation. Based on tributary inputs, we estimate an average annual discharge for the Yukon River of 210 km3 year–1. This value is within 1% of the average measured annual discharge at the U.S. Geological Survey gaging station near the river terminus at Pilot Station, AK, for water years 2001 through 2005. Next, annual loads for 11 solutes are determined by combining annual discharge with point measurements of solute concentrations in tributary river water. Based on the sum of solutes in tributary water, we find that the Yukon River discharges approximately 33 million metric tons of dissolved solids each year at Pilot Station. Discharged solutes are dominated by cations calcium and magnesium (5.65 × 109 and 1.42 × 109 g year–1) and anions bicarbonate and sulphate (17.3 × 109 and 5.40 × 109 g year–1). These loads compare well with loads calculated independently at the three continuous gaging stations along the Yukon River. These findings show how annual solute yields vary throughout a major subarctic river basin and that accurate estimates of total river export can be determined from calculated tributary contributions. Copyright © 2011. This article is a U.S. Government work and is in the public domain in the USA.

Generalized Mediterranean Annual Water Yield Model: Grunsky’s Equation and Long-Term Average Temperature

To estimate mean annual water yield on ungauged watersheds, regional regression equations have been developed that are a function of upstream area, precipitation, and temperature, and are commonly used. In this study, the 100-year old Grunsky equation is used as the foundation for a fully implemented generalized Mediterranean water yield model that is based on long-term annual precipitation and temperature. The work here (1) extends and generalizes Grunsky’s equation beyond its original conditions, (2) establishes an underlying hydro-climate generalized model for average annual Mediterranean watershed water yield and loss calculation, and (3) applies the validated model to estimate watershed water yield and predict potential water yield response to watershed temperature and precipitation. The water yield model was applied to watersheds in coastal California, southern France, and Portugal, but might be applied to other Mediterranean sites for which precipitation and temperature data are available. It also ...

Regional flow duration curves for ungauged sites in Sicily

Abstract. Flow duration curves are simple and powerful tools to deal with many hydrological and environmental problems related to water quality assessment, water-use assessment and water allocation. Unfortunately the scarcity of streamflow data enables the use of these instruments only for gauged basins. A regional model is developed here for estimating flow duration curves at ungauged basins in Sicily, Italy. Due to the complex ephemeral behavior of the examined region, this study distinguishes dry periods, when flows are zero, from wet periods using a three parameters power law to describe the frequency distribution of flows. A large dataset of streamflows has been analyzed and the parameters of flow duration curves have been derived for about fifty basins. Regional regression equations have been developed to derive flow duration curves starting from morphological basin characteristics.

Hydrographs and Estimates of Scour Depth Excess for Pier Scour Prediction

A practical procedure for using the Scour Rate in Cohesive Soils (SRICOS) method to predict bridge pier scour depth in ungauged streams is presented. The procedure is illustrated by the Interstate 90 westbound bridge over Split Rock Creek in South Dakota. Instead of generating a continuous hydrograph for the design service life of a bridge, the simplified approach considered a series of maximum annual floods. The parameters of peak flow distribution were calculated with estimates of peak flow magnitude obtained from regional regression equations. Flood duration was estimated with a rainfall atlas and a Soil Conservation Service runoff curve number based on a design storm. A set of equally probable future hydrographs was generated and used with the SRICOS method to compute the final scour depth. The predicted final scour depth followed the normal distribution. The risk values associated with different design scour depths were calculated with the mean and standard deviation of the normal distribution. The scour predictions were consistent with a bridge founded on highly erosion-resistant cohesive soils. The cumulative effect of scour from the small floods in a long sequence of floods could be represented by an equivalent time; the scour depth produced by the whole sequence could be modeled by using only the large floods if the equivalent time were small compared with the combined duration of the large floods.

A Fully Distributed Implementation of Mean Annual StreamFlow Regional Regression Equations1

Abstract: Estimates of mean annual streamflow are needed for a variety of hydrologic assessments. Away from gage locations, regional regression equations that are a function of upstream area, precipitation, and temperature are commonly used. Geographic information systems technology has facilitated their use for projects, but traditional approaches using the polygon overlay operator have been too inefficient for national scale applications. As an alternative, the Elevation Derivatives for National Applications (EDNA) database was used as a framework for a fully distributed implementation of mean annual streamflow regional regression equations. The raster “flow accumulation” operator was used to efficiently achieve spatially continuous parameterization of the equations for every 30 m grid cell of the conterminous United States (U.S.). Results were confirmed by comparing with measured flows at stations of the Hydro‐Climatic Data Network, and their applications value demonstrated in the development of a national geospatial hydropower assessment. Interactive tools at the EDNA website make possible the fast and efficient query of mean annual streamflow for any location in the conterminous U.S., providing a valuable complement to other national initiatives (StreamStats and the National Hydrography Dataset Plus).

Baseflow index regionalization analysis in a mediterranean area and data scarcity context: Role of the catchment permeability index

Low flow characteristics are affected by different physiographic factors such as climate, topography, geology and soils, and regional regression prediction models, to estimate low flow indexes at ungauged sites, mainly rely on these factors. The paper focuses on the baseflow index, one of the most important low flow characteristics for a catchment, and presents: (i) the analysis of baseflow separation algorithms for BFI evaluation and (ii) a regional approach to predict the BFI at ungauged sites in a Mediterranean region, for which only very poor data are available. The prediction of baseflow contribution to total streamflow is based on the introduction of a permeability index, at the catchment scale, and regional linear regression equations simply relate the latter to the BFI. For the studied area geological features have been found to be the major factor affecting baseflow and the permeability index estimation for a particular catchment, in an apparently over-simplified schematization, essentially reflects catchment lithology. As a matter of fact, an accurate catchment geology spatial variability description reduces the average long-term BFI index prediction error from 23% to 14% and above all increases the explained variance from 23% to 68%.

Regularized joint inverse estimation of extreme rainfall amounts in ungauged coastal basins of El Salvador

A regularized joint inverse procedure is presented and used to estimate the magnitude of extreme rainfall events in ungauged coastal river basins of El Salvador: Paz, Jiboa, Grande de San Miguel, and Goascoran. Since streamflow measurements reflect temporal and spatial rainfall information, peak-flow discharge is hypothesized to represent a similarity measure suitable for regionalization. To test this hypothesis, peak-flow discharge values determined from streamflow recurrence information (10-year, 25-year, and 100-year) collected outside the study basins are used to develop regional (country-wide) regression equations. Peak-flow discharge derived from these equations together with preferred spatial parameter relations as soft prior information are used to constrain the simultaneous calibration of 20 tributary basin models. The nonlinear range of uncertainty in estimated parameter values (1 curve number and 3 recurrent rainfall amounts for each model) is determined using an inverse calibration-constrained Monte Carlo approach. Cumulative probability distributions for rainfall amounts indicate differences among basins for a given return period and an increase in magnitude and range among basins with increasing return interval. Comparison of the estimated median rainfall amounts for all return periods were reasonable but larger (3.2–26%) than rainfall estimates computed using the frequency-duration (traditional) approach and individual rain gauge data. The observed 25-year recurrence rainfall amount at La Hachadura in the Paz River basin during Hurricane Mitch (1998) is similar in value to, but outside and slightly less than, the estimated rainfall confidence limits. The similarity in joint inverse and traditionally computed rainfall events, however, suggests that the rainfall observation may likely be due to under-catch and not model bias.

Enhancing Hyrdological Simulation Program – FORTRAN Model Channel Hydraulic Representation1

Abstract: The hydrological simulation program – FORTRAN (HSPF) is a comprehensive watershed model that employs depth‐area‐volume‐flow relationships known as the hydraulic function table (FTABLE) to represent the hydraulic characteristics of stream channel cross‐sections and reservoirs. An accurate FTABLE determination for a stream cross‐section site requires an accurate determination of mean flow depth, mean flow width, roughness coefficient, longitudinal bed slope, and length of stream reach. A method that uses regional regression equations to estimate mean flow depth, mean flow width, and roughness coefficient is presented herein. FTABLES generated by the proposed method (Alternative Method) and FTABLES generated by Better Assessment Science Integrating Point and Nonpoint Sources (BASINS) were compared. As a result, the Alternative Method was judged to be an enhancement over the BASINS method. First, the Alternative Method employs a spatially variable roughness coefficient, whereas BASINS employs an arbitrarily selected spatially uniform roughness coefficient. Second, the Alternative Method uses mean flow width and mean flow depth estimated from regional regression equations whereas BASINS uses mean flow width and depth extracted from the National Hydrography Dataset (NHD). Third, the Alternative Method offers an option to use separate roughness coefficients for the in‐channel and floodplain sections of compound channels. Fourth, the Alternative Method has higher resolution in the sense that area, volume, and flow data are calculated at smaller depth intervals than the BASINS method. To test whether the Alternative Method enhances channel hydraulic representation over the BASINS method, comparisons of observed and simulated streamflow, flow velocity, and suspended sediment were made for four test watersheds. These comparisons revealed that the method used to estimate the FTABLE has little influence on hydrologic calibration, but greatly influences hydraulic and suspended sediment calibration. The hydrologic calibration results showed that observed versus simulated daily streamflow comparisons had Nash‐Sutcliffe efficiencies ranging from 0.50 to 0.61 and monthly comparisons had efficiencies ranging from 0.61 to 0.84. Comparisons of observed and simulated suspended sediments concentrations had model efficiencies ranging from 0.48 to 0.56 for the daily, and 0.28 to 0.70 for the monthly comparisons. The overall results of the hydrological, hydraulic, and suspended sediment concentration comparisons show that the Alternative Method yielded a relatively more accurate FTABLE than the BASINS method. This study concludes that hydraulic calibration enhances suspended sediment simulation performance, but even greater improvement in suspended sediment calibration can be achieved when hydrological simulation performance is improved. Any improvements in hydrological simulation performance are subject to improvements in the temporal and spatial representation of the precipitation data.

Evolution of tidal creek networks in a high sedimentation environment: A 5-year experiment at Tijuana Estuary, California

In a large (8 ha) salt marsh restoration site, we tested the effects of excavating tidal creeks patterned after reference systems. Our purposes were to enhance understanding of tidal creek networks and to test the need to excavate creeks during salt marsh restoration. We compared geomorphic changes in areas with and without creek networks (n 5 3; each area 1.3 ha) and monitored creek cross-sectional areas, creek lengths, vertical accretion, and marsh surface elevations for 5 yr that included multiple sedimentation events. We hypothesized that cells with creeks would develop different marsh surface and creek network characteristics (i.e., surface elevation change, sedimentation rate, creek cross-sectional area, length, and drainage density). Marsh surface vertical accretion averaged 1.3 cm yr 21 with large storm inputs, providing the opportunity to assess the response of the drainage network to extreme sedimentation rates. The constructed creeks initially filled due to high accretion rates but stabilized at cross-sectional areas matching, or on a trajectory toward, equilibrium values predicted by regional regression equations. Sedimentation on the marsh surface was greatest in low elevation areas and was not directly influenced by creeks. Time required for cross-sectional area stabilization ranged from 0 to . 5 yr, depending on creek order. First-order constructed creeks lengthened rapidly (mean rate of 1.3 m yr 21 ) in areas of low elevation and low vegetation cover. New (volunteer) creeks formed rapidly in cells without creeks in areas with low elevation, low vegetation cover, and high elevation gradient (mean rate of 6.2 m yr 21 ). After 5 yr, volunteer creeks were, at most, one-fourth the area of constructed creeks and had not yet reached the upper marsh plain. In just 4 yr, the site's drainage density expanded from 0.018 to reference levels of 0.022 m m 22 . Pools also formed on the marsh plain due to sediment resuspension associated with wind-driven waves. We conclude that excavated creeks jump-started the development of drainage density and creek and channel dimensions, and that the tidal prism became similar to those of the reference site in 4-5 yr.

Geração de modelos de regionalização de vazões máximas, médias de longo período e mínimas de sete dias para a bacia do rio do Carmo, Minas Gerais

O fato de ser limitada a disponibilidade hídrica de uma bacia hidrográfica, aliado ao crescente aumento da demanda motivado pelos mais variados usos e impulsionado pelo desenvolvimento econômico, exige a adoção de critérios de gestão capazes de minimizar os inevitáveis conflitos pelo uso da água. Para que a gestão seja adequada, requer-se o conhecimento dos diferentes segmentos do ciclo hidrológico, notadamente da precipitação e da vazão. Com isso, e considerando-se a importância de se verem atendidas as regiões com ausência ou escassez de dados fluviométricos, desenvolveu-se o presente estudo de construção de modelos matemáticos de regionalização de vazões na bacia do rio do Carmo. Para isso, foram construídas as séries de chuvas totais anuais, totais do semestre mais chuvoso e máximas diárias, e as séries de vazões máximas, médias e mínimas. As séries de chuvas no ponto foram convertidas em chuvas na área com o uso do método de Thiessen. Com base nas séries anuais das vazões máximas, médias e mínimas, ajustaram-se modelos de probabilidade que permitiram a extrapolação dos dados históricos. Ainda, para pesquisar a influência do relevo sobre as variáveis hidrológicas foram feitas as quantificações de diferentes parâmetros que caracterizam a bacia. Deste modo, ao final foram construídos os modelos de regionalização das vazões.

PREDICTION OF STREAM TEMPERATURE IN FORESTED WATERSHEDS1

ABSTRACT: Removal of streamside vegetation changes the energy balance of a stream, and hence its temperature. A common approach to mitigating the effects of logging on stream temperature is to require establishment of buffer zones along stream corridors. A simple energy balance model is described for prediction of stream temperature in forested headwater watersheds that allows evaluation of the performance of such measures. The model is designed for application to “worst case” or maximum annual stream temperature, under low flow conditions with maximum annual solar radiation and air temperature. Low flows are estimated via a regional regression equation with independent variables readily accessible from GIS databases. Testing of the energy balance model was performed using field data for mostly forested basins on both the west and east slopes of the Cascade Mountains, and was then evaluated using the regional equations for low flow and observed maximum reach temperatures in three different east slope Cascades catchments. A series of sensitivity analyses showed that increasing the buffer width beyond 30 meters did not significantly decrease stream temperatures, and that other vegetation parameters such as leaf area index, average tree height, and to a lesser extent streamside vegetation buffer width, more strongly affected maximum stream temperatures.

Time Domain Reflectometry Field Calibration in the Little Washita River Experimental Watershed

Accurate measurement of profile soil water content is essential to many areas of environmental and agricultural research. In this study five methods were evaluated for determining volumetric profile soil water content from time domain reflectometry (TDR) data at nine locations within the Little Washita River Experimental Watershed (LWREW) in south central Oklahoma. Soil compositions for the sites ranged between 24.6 to 86.4% for sand and 5.2 to 29.2% for clay. Comparisons were made between gravimetric soil sample data and soil water content as determined by the TDR factory instrument calibration, two methods of site-specific calibration, a regional calibration technique, and an empirically derived universal approach. Method 1 is the factory calibration, which uses average values for model coefficients that were derived from extensive laboratory work and theoretical analysis. Method 2 fits a site-specific linear regression of TDR time delay on measured soil-core water content. Method 3 uses the factory calibration equation and site-specific values for the ratio of TDR time delay in dry soil, to that in air (Ts/Ta). In Method 4, a regional linear regression equation was developed from an analysis combining data from all study sites. Method 5 applies a universal equation based on the linear relationship between soil water content and the apparent dielectric constant of soil (Ka) measured by TDR. Statistical analysis of the data showed that of the five methods, only the mean root mean square error (RMSE) value for Method 2 was significantly different from all other methods (α = 0.05). Method 2 also had the smallest standard deviation and standard error, and the narrowest range of RMSE values. All field calibration methods show that it is necessary to include very low water content data in determining absolute water content. When compared with the factory calibration, all other methods improved the measurement of soil water content, with Method 2 providing the most accurate results at the site-specific level of analysis.

Time Domain Reflectometry Field Calibration in the Little Washita River Experimental Watershed

Accurate measurement of profile soil water content is essential to many areas of environmental and agricultural research. In this study five methods were evaluated for determining volumetric profile soil water content from time domain reflectometry (TDR) data at nine locations within the Little Washita River Experimental Watershed (LWREW) in south central Oklahoma. Soil compositions for the sites ranged between 24.6 to 86.4% for sand and 5.2 to 29.2% for clay. Comparisons were made between gravimetric soil sample data and soil water content as determined by the TDR factory instrument calibration, two methods of site‐specific calibration, a regional calibration technique, and an empirically derived universal approach. Method 1 is the factory calibration, which uses average values for model coefficients that were derived from extensive laboratory work and theoretical analysis. Method 2 fits a site‐specific linear regression of TDR time delay on measured soil‐core water content. Method 3 uses the factory calibration equation and site‐specific values for the ratio of TDR time delay in dry soil, to that in air (Ts/Ta). In Method 4, a regional linear regression equation was developed from an analysis combining data from all study sites. Method 5 applies a universal equation based on the linear relationship between soil water content and the apparent dielectric constant of soil (Ka) measured by TDR. Statistical analysis of the data showed that of the five methods, only the mean root mean square error (RMSE) value for Method 2 was significantly different from all other methods (α = 0.05). Method 2 also had the smallest standard deviation and standard error, and the narrowest range of RMSE values. All field calibration methods show that it is necessary to include very low water content data in determining absolute water content. When compared with the factory calibration, all other methods improved the measurement of soil water content, with Method 2 providing the most accurate results at the site‐specific level of analysis.

EVALUATION OF THE TRANSFER METHOD FOR ESTIMATING RIVER FLOWS

EVALUATION OF THE TRANSFER METHOD FOR ESTIMATING RIVER FLOWSSeveral methods exist for estimating river flows at ungaged sites, including regional regression equations and hydrologic modeling. The transfer method is a simple method of using gage data from one site to estimate flows at an ungaged site, based on the proportionality of contributing drainage areas. This paper evaluates the ability of the transfer method to estimate flow at gaged sites, thereby...Author(s)Scott B. BellR. Scott WadeSourceProceedings of the Water Environment FederationSubjectSession 2: Watershed Model DevelopmentDocument typeConference PaperPublisherWater Environment FederationPrint publication date Jan, 2002ISSN1938-6478SICI1938-6478(20020101)2002:2L.252;1-DOI10.2175/193864702785664923Volume / Issue2002 / 2Content sourceWatershed ConferenceFirst / last page(s)252 - 266Copyright2002Word count80

Regional calibration of a watershed model

As watershed models become increasingly sophisticated and useful, there is a need to extend their applicability to locations where they cannot be calibrated or validated. A new methodology for the regionalization of a watershed model is introduced and evaluated. The approach involves calibration of a watershed model to many sites in a region, concurrently. Previous research that has sought to relate the parameters of monthly water balance models to physical drainage basin characteristics in a region has met with limited success. Previous studies have taken the two-step approach: (a) estimation of watershed model parameters at each site, followed by (b) attempts to relate model parameters to drainage basin characteristics. Instead of treating these two steps as independent, both steps are implemented concurrently. All watershed models in a region are calibrated simultaneously, with the dual objective of reproducing the behaviour of observed monthly streamflows and, additionally, to obtain good relationships between watershed model parameters and basin characteristics. The approach is evaluated using 33 basins in the southeastern region of the United States by comparing simulations using the regional models for three catchments which were not used to develop the regional regression equations. Although the regional calibration approach led to nearly perfect regional relationships between watershed model parameters and basin characteristics, these “improved” regional relationships did not result in improvements in the ability to model streamflow at ungauged sites. This experiment reveals that improvements in regional relationships between watershed model parameters and basin characteristics will not necessarily lead to improvements in the ability to calibrate a watershed model at an ungauged site.

Estimation of Time of Concentration for Maryland Streams

The time of concentration (TC) is an important input to most hydrologic models and is usually estimated by travel-time computations or by using rainfall-runoff data. Average TCs were determined for 78 rural and urban watersheds in Maryland and related to watershed characteristics using regression analysis. The regression equation is based on the channel length and slope; the percentage of the watershed covered with forests, lakes, and ponds; and the percentage of the watershed with impervious areas. The equation is applicable for estimating TCs for rural and urban watersheds in Maryland with watershed characteristics similar to the gauging station data. TC values computed at Maryland gauging stations were compared with estimates from an equation developed by Kirpich and the Soil Conservation Service (SCS) lag equation, and with basin lag times determined by the U.S. Geological Survey (USGS). The average TC values computed in this analysis were about 5 percent higher than the basin lag-time estimates, which is consistent with the USGS’s definition of lag time. TC estimates from the Kirpich or SCS equations were consistently lower than the values computed from gauging station data. The tendency to underestimate TCs is a major reason why hydrologic models often provide conservative estimates of design discharges compared with regional regression equations and gauging station data.

Estimation of geomorphically significant flows in alpine streams of the Rocky Mountains, Colorado (USA)

Streamflows recorded at 24 gauging stations in the Rocky Mountains of Colorado were analyzed to derive regional regression equations for estimating the natural flow duration and flood frequency in reaches where the natural flows are unknown or have been altered by diversion or regulation. The principal objective of this analysis is to determine whether the relatively high, infrequent, but geomorphically and ecologically important flows in the Rocky Mountains can be accurately estimated by regional flow duration equations. The region considered in this study is an area of relatively abundant runoff, and, consequently, intense water resources development. The specific streams analyzed here, however, are unaltered and remain nearly pristine. Regional flow duration equations are derived for two situations. When the mean annual discharge is known, flows ≧10% of the time can be estimated with an uncertainty of ±9% for the 10% exceedance flow, to ±11% for the 1.0% exceedance flow. When the mean annual discharge is unknown, the relatively high, infrequent flow can be estimated using the mean basin precipitation rate (in m3/s), and basin relief with an uncertainty of ±23% for the 10% exceedance flow to ±21% for the 1.0% exeedance flow. The uncertainty in estimated discharges using the equations derived in this analysis is substantially smaller than has been previously reported, especially for the geomorphically significant flows which are relatively large and infrequent. The improvement is due primarily to the quality of streamflow records analyzed and a well-defined hydrologic region.

Regional Regression Models of Annual Streamflow for the United States

Estimates of annual streamflow volumes are needed in many different types of hydrologic studies. Usually a streamgauge is unavailable at the location of interest, hence regional methods that relate streamflow to readily measured geomorphic and climate characteristics provide a practical solution. Hydrologic, geomorphic, and climatic characteristics of 1,553 undeveloped watersheds across the United States are used to develop regional regression equations that relate the first two moments of annual streamflow to readily measured basin and climate characteristics. These relations are summarized for each of 18 major U.S. water resource regions. The relationships are remarkably precise, with adjusted R2 values ranging from 90.2–99.8% and an average value of 96.2% across the continent. The usefulness of these relationships is evaluated by deriving their information content in terms of equivalent record length. These results indicate that regional models of annual streamflow, including runoff maps, are less accurate than suggested by traditional goodness-of-fit statistics. We also provide estimates of precipitation and temperature elasticity of streamflow, by region.

Development of regional parameter estimation equations for a macroscale hydrologic model

A methodology for developing regional parameter estimation equations, designed for application to continental scale river basins, is described. The approach, which is applied to the two-layer Variable Infiltration Capacity (VIC-2L) land surface hydrologic model, uses a set of 34 unregulated calibration or “training” catchments (drainage areas 10 2–10 4 km 2) distributed throughout the Arkansas–Red River basin of the south central U.S. For each of these catchments, parameters were determined by: a) prior estimation of two of the model parameters (saturated hydraulic conductivity and pore size distribution index) from the U.S. Soil Conservation Service State Soil Geographic Data Base (STATSGO) data base; and b) estimation of the remaining seven parameters via a search procedure that minimizes the sum of squares of differences between predicted and observed streamflow. The catchment parameters were then related to 11 ancillary distributed land surface characteristics extracted from STATSGO, and 17 variables derived from station meteorological data. The seven regression equations explained from 54 to 76% of the variance of the parameters. The most frequently occurring ancillary variables were the average permeability, saturated hydraulic conductivity, and SCS hydrologic Group B (typically soils with moderately high infiltration rates) fraction derived from STATSGO, and the average temperature and standard deviation of fall precipitation. The method was tested by comparing simulations using the regional (regression equation) parameters for six unregulated catchments not in the parameter estimation set. The model performance using the regional parameters was quite good for most of the calibration and validation catchments, which were humid and semi-humid. The model did not perform as well for the smaller number of arid to semi-arid catchments.