Choosing an arbitrary calibration period for hydrologic models: How much does it influence water balance simulations?

The selection of calibration and validation time periods in hydrologic modelling is often done arbitrarily. Nonstationarity can lead to an optimal parameter set for one period which may not accurately simulate another. However, there is still much to be learned about the responses of hydrologic models to nonstationary conditions. We investigated how the selection of calibration and validation periods can influence water balance simulations. We calibrated Soil and Water Assessment Tool hydrologic models with observed streamflow for three United States watersheds (St. Joseph River of Indiana/Michigan, Escambia River of Florida/Alabama, and Cottonwood Creek of California), using time period splits for calibration/validation. We found that the choice of calibration period (with different patterns of observed streamflow, precipitation, and air temperature) influenced the parameter sets, leading to dissimilar simulations of water balance components. In the Cottonwood Creek watershed, simulations of 50-year mean January streamflow varied by 32%, because of lower winter precipitation and air temperature in earlier calibration periods on calibrated parameters, which impaired the ability for models calibrated to earlier periods to simulate later periods. Peaks of actual evapotranspiration for this watershed also shifted from April to May due to different parameter values depending on the calibration period's winter air temperatures. In the St. Joseph and Escambia River watersheds, adjustments of the runoff curve number parameter could vary by 10.7% and 20.8%, respectively, while 50-year mean monthly surface runoff simulations could vary by 23%–37% and 169%–209%, depending on the observed streamflow and precipitation of the chosen calibration period. It is imperative that calibration and validation time periods are chosen selectively instead of arbitrarily, for instance using change point detection methods, and that the calibration periods are appropriate for the goals of the study, considering possible broad effects of nonstationary time series on water balance simulations. It is also crucial that the hydrologic modelling community improves existing calibration and validation practices to better include nonstationary processes.     

Adjusting wavelet‐based multiresolution analysis boundary conditions for long‐term streamflow forecasting

We propose a novel technique for improving a long‐term multi‐step‐ahead streamflow forecast. A model based on wavelet decomposition and a multivariate Bayesian machine learning approach is developed for forecasting the streamflow 3, 6, 9, and 12 months ahead simultaneously. The inputs of the model utilize only the past monthly streamflow records. They are decomposed into components formulated in terms of wavelet multiresolution analysis. It is shown that the model accuracy can be increased by using the wavelet boundary rule introduced in this study. A simulation study is performed to evaluate the effects of different wavelet boundary rules using synthetic and real streamflow data from the Yellowstone River in the Uinta Basin in Utah. The model based on the combination of wavelet and Bayesian machine learning regression techniques is compared with that of the wavelet and artificial neural networks‐based model. The robustness of the models is evaluated. Copyright © 2015 John Wiley & Sons, Ltd.     

Runoff evolution in the Swiss Alps: projections for selected high‐alpine catchments based on ENSEMBLES scenarios

The Alps are often referred to as the ‘water tower of Europe’. In Switzerland, many branches of the economy, especially the hydropower industry, are closely linked to and dependent on the availability of water. Assessing the impact of climate change on streamflow runoff is, thus, of great interest. Major efforts have already been made in this respect, but the analyses often focus on individual catchments and are difficult to intercompare. In this article, we analysed nine high‐alpine catchments spread over the Swiss Alps, selected for their relevance to a wide range of morphological characteristics. Runoff projections were carried out until the end of the current century by applying the Glacier Evolution Runoff Model (GERM) and climate scenarios generated in the framework of the ENSEMBLES project. We focused on assessing the uncertainty induced by the unknown climate evolution and provided general, statistically based statements, which should be useful as a ‘rule of thumb’ for analyses addressing questions related to water management. Catchments with a high degree of glacierization will undergo the largest changes. General statements about absolute variations in discharge are unreliable, but an overall pattern, with an initial phase of increased annual discharge, followed by a phase with decreasing discharge, is recognizable for all catchments with a significant degree of glacierization. In these catchments, a transition from glacial and glacio‐nival regime types to nival will occur. The timing of maximal annual runoff is projected to occur before 2050 in all basins. The time of year with maximal daily discharges is expected to occur earlier at a rate of 4·4 ± 1·7 days per decade. Compared to its present level, the contribution of snow‐ and icemelt to annual discharge is projected to drop by 15 to 25% until the year 2100. Copyright © 2011 John Wiley & Sons, Ltd.     

Spatio‐temporal patterns of major ions in urban stormwater under cold climate

Multiple natural and anthropogenic factors affect urban water chemistry. However, little is known about the abundance or temporal variation of major ions in urban runoff. This study explores the spatio‐temporal variation of major dissolved ions (Na, K, Ca, Mg, Cl, NO₃, and SO₄) and total dissolved solids (TDS) in cold climate urban stormwater. Three watersheds with varying degrees of urban land use intensity and imperviousness (from 36% to 66%) in Helsinki, Finland, were continuously monitored for 5 years using an automated sampling procedure to obtain stormwater discharge and ion concentrations and, thus, loadings. High‐resolution datasets, including long‐term continuous discharge, both measured and simulated (using Storm Water Management Model), and automatic water quality sampling enabled the accurate calculation of loads of ions and TDS. Water quality was related to explanatory watershed characteristics (e.g., watershed physiography and sampling time) using hierarchical clustering, nonmetric multidimensional scaling, and hierarchical partitioning methods. Urban land use contributed to increased ion concentrations and loads year‐round. This study highlights how stormwater ion concentrations are elevated across seasons, indicating chronic pollution phenomena. The greatest loads occurred during summer (except for Na and Cl), while the highest variation in loads was observed in autumn. Significant clusters among ions were found in the hierarchical cluster analysis, suggesting similar temporal patterns and sources for the ions in each cluster. The importance of land use was evident, though in the most urbanized watershed, concentrations were not linked to any of the investigated watershed characteristics. Based on our results, only Na and Cl are manageable by alternative winter road antiskid practices, whereas other ions resulted from diffuse pollution sources, being therefore more difficult to control. Finally, this study contributes to an increased understanding of the temporal and spatial patterns of ions in stormwater and highlights the need for consistent time series data for ion monitoring under cold climatic conditions in order to enable reliable estimates of their loads to adjacent water bodies. Finally, year‐round stormwater treatment is highly recommended.     

A combined rainfall infiltration model based on Green–Ampt and SCS‐curve number

The Green–Ampt infiltration equation is an incomplete governing equation for rainfall infiltration due to the absence of an inertia term. The estimation of the capillary pressure head at the wetting front is difficult to determine. Thus, a major limitation of the Green–Ampt model is the constant, non‐zero surface ponding depth. This paper proposes an integrated rainfall infiltration model based on the Green–Ampt model and the SCS‐CN model. It achieves a complete governing equation for rainfall infiltration by momentum balance and the water budget based on the Green–Ampt assumption, and uses the curve number from the SCS‐CN method to calculate the initial abstraction, which is used as a basic parameter for the governing equation of the intensity of rainfall loss during the runoff period. The integrated rainfall infiltration model resolves the dilemma for capillary pressure head estimation, overcomes the limitation of constant, non‐zero surface ponding depth, and facilitates the calculation of runoff for individual flood simulations. Copyright © 2014 John Wiley & Sons, Ltd.     

Smoothing of digital elevation models and the alteration of overland flow path length distributions

DEM smoothing is a common pre-processing technique used to remove undesirable roughness from a DEM. However, it is hypothesized that smoothing straightens and reduces the length of overland flow paths, which is an important factor controlling modelled time-to-peak flow. Currently, there is a lack of research comparing how different smoothing techniques alter the distribution of overland flow path length. Four low-pass filtering techniques were applied to three fine-resolution LiDAR DEMs of varying relief: the mean filter, the median filter, the Gaussian filter, and the feature-preserving DEM smoothing (FPDEMS) filter, each with different degrees of smoothing. Downslope-distance-to-stream distributions were then derived using D8 and D∞ flow directions and statistically compared to distributions derived from the unsmoothed DEM for each study site. The results indicate that the alteration of flow path length distributions as a result of smoothing is complex. Mean flow path lengths may decrease or increase in response to smoothing, depending on landscape relief and the derivation of flow directions, and generalized flow paths may become longer. The largest increase in mean flow path lengths was 19.2 m using the 21 × 21 median filter and D8 flow directions in the high-relief study site, relative to an unsmoothed mean length of 138.6 m in this site. The largest decrease in mean flow path length was 48.9 m using the 21 × 21 mean filter and D∞ flow directions in the low-relief study site, relative to an unsmoothed mean length of 290.9 m in this site. Furthermore, minimal flow path length alterations were achieved with the Gaussian filter when gentle smoothing is required, and with the FPDEMS filter when moderate to aggressive smoothing is required. These results suggest that an appropriate smoothing method should be chosen based on the relief of the landscape and the degree of smoothing required.     

Long-term research catchments to investigate shrub encroachment in the Sonoran and Chihuahuan deserts: Santa Rita and Jornada experimental ranges

Woody plant encroachment is a global phenomenon whereby shrubs or trees replace grasses. The hydrological consequences of this ecological shift are of broad interest in ecohydrology, yet little is known of how plant and intercanopy patch dynamics, distributions, and connectivity influence catchment-scale responses. To address this gap, we established research catchments in the Sonoran and Chihuahuan Deserts (near Green Valley, Arizona and near Las Cruces, New Mexico, respectively) that represent shrub encroachment in contrasting arid climates. Our main goals in the coordinated observations were to: (a) independently measure the components of the catchment water balance, (b) deploy sensors to quantify the spatial patterns of ecohydrological processes, (c) use novel methods for characterizing catchment properties, and (d) assess shrub encroachment impacts on ecohydrological processes through modelling studies. Datasets on meteorological variables; energy, radiation, and CO₂ fluxes; evapotranspiration; soil moisture and temperature; and runoff at various scales now extend to nearly 10 years of observations at each site, including both wet and dry periods. Here, we provide a brief overview of data collection efforts and offer suggestions for how the coordinated datasets can be exploited for ecohydrological inferences and modelling studies. Given the representative nature of the catchments, the available databases can be used to generalize findings to other catchments in desert landscapes.     

Assessing stand water use in four coastal wetland forests using sapflow techniques: annual estimates,errors and associated uncertainties

Forests comprise approximately 37% of the terrestrial land surface and influence global water cycling. However, very little attention has been directed towards understanding environmental impacts on stand water use (S) or in identifying rates of S from specific forested wetlands. Here, we use sapflow techniques to address two separate but linked objectives: (1) determine S in four, hydrologically distinctive South Carolina (USA) wetland forests from 2009–2010 and (2) describe potential error, uncertainty and stand‐level variation associated with these assessments. Sapflow measurements were made from a number of tree species for approximately 2–8 months over 2 years to initiate the model, which was applied to canopy trees (DBH > 10–20 cm). We determined that S in three healthy forested wetlands varied from 1.97–3.97 mm day^?1 or 355–687 mm year^?1 when scaled. In contrast, saltwater intrusion impacted individual tree physiology and size class distributions on a fourth site, which decreased S to 0.61–1.13 mm day^?1 or 110–196 mm year^?1. The primary sources of error in estimations using sapflow probes would relate to calibration of probes and standardization relative to no flow periods and accounting for accurate sapflow attenuation with radial depth into the sapwood by species and site. Such inherent variation in water use among wetland forest stands makes small differences in S (<200 mm year^?1) difficult to detect statistically through modelling, even though small differences may be important to local water cycling. These data also represent some of the first assessments of S from temperate, coastal forested wetlands along the Atlantic coast of the USA. Copyright © 2014 John Wiley & Sons, Ltd.     

Coupling the WRF model with a temperature index model based on remote sensing for snowmelt simulations in a river basin in the Altay Mountains,north‐west China

Snowmelt water is an important freshwater resource in the Altay Mountains in north‐west China; however, warming climate and rapid spring snowmelt can cause floods that endanger both public and personal property and safety. This study simulates snowmelt in the Kayiertesi River catchment using a temperature index model based on remote sensing coupled with high‐resolution meteorological data obtained from National Centers for Environmental Prediction (NCEP) reanalysis fields that were downscaled using the Weather Research Forecasting model and then bias corrected using a statistical downscaled model. Validation of the forcing data revealed that the high‐resolution meteorological fields derived from the downscaled NCEP reanalysis were reliable for driving the snowmelt model. Parameters of the temperature index model based on remote sensing were calibrated for spring 2014, and model performance was validated using Moderate Resolution Imaging Spectroradiometer snow cover and snow observations from spring 2012. The results show that the temperature index model based on remote sensing performed well, with a simulation mean relative error of 6.7% and a Nash–Sutcliffe efficiency of 0.98 in spring 2012 in the river of Altay Mountains. Based on the reliable distributed snow water equivalent simulation, daily snowmelt run‐off was calculated for spring 2012 in the basin. In the study catchment, spring snowmelt run‐off accounts for 72% of spring run‐off and 21% of annual run‐off. Snowmelt is the main source of run‐off for the catchment and should be managed and utilized effectively. The results provide a basis for snowmelt run‐off predictions, so as to prevent snowmelt‐induced floods, and also provide a generalizable approach that can be applied to other remote locations where high‐density, long‐term observational data are lacking. Copyright © 2016 John Wiley & Sons, Ltd.     

From groundwater abstraction to vegetative response in fen ecosystems

Hydrological effects of groundwater abstraction near a Danish river valley have been assessed by integrated hydrological modelling. The study site contains groundwater‐dependent terrestrial ecosystems in terms of fen and spring habitats that are highly dependent on regional and local scale hydrology. Fens are rare and threatened worldwide due to pressures from agriculture, to lack of appropriate management and to altered catchment hydrology. A solid foundation for hydrological modelling was established based on intensive monitoring at the site, combined with full‐scale pumping tests in the area. A regional groundwater model was used to describe the dynamics in groundwater recharge and the large‐scale discharge to streams. A local grid refinement approach was then applied in a detailed assessment of damage in order to balance the computational effort and the need for a high spatial resolution. A considerable flow reduction in the natural spring was monitored during a full‐scale pumping test while no significant effects on the water table in the fen habitats were observed. A modelled abstraction scenario predicted a lowering of 2–3 cm in the centre of the main fen area during summer periods. The predicted change in water table conditions in the fen habitat is compared to the variability found in 35 Danish fens, and the ecological response is discussed based on statistical water‐level vegetation relations. The results provide a rare quantitative foundation for decision making in relation to management of groundwater‐dependent terrestrial ecosystems. Copyright © 2013 John Wiley & Sons, Ltd.