Search

We present a data‐guided framework to explore the hydrological basis of the Budyko curve, focusing on the role of soil moisture in mediating the partitioning of precipitation into runoff and evaporation and the role of climate in this partitioning. The conceptual framework for the derivation of the Budyko curve builds on empirically based power law relationships for long‐term average scaled runoff (Q/P) and scaled evaporation (E/Ep), both as functions of long‐term average (equilibrium) soil moisture. A second component of the framework is an empirical power law relationship between equilibrium soil moisture and the aridity index, Ep/P. Combinations of these empirical power law relationships with the long‐term water balance equation give rise to a form of the Budyko curve, which is a close approximation to Budyko's original relationship and is mathematically similar to the Turc‐Pike equation, thus supporting the role of equilibrium soil moisture in giving rise to the Budyko curve. In the remainder of the paper, for special cases of the parameter values of the power law relationships, we independently derive both the soil moisture dependence on Ep/P as well as the Budyko curve itself that not only reproduces a new Budyko‐type curve, but can also explain the spread of the empirical data points away from the mean curve. The paper also raises fresh questions about how the empirical power law relationships that govern the soil‐mediated precipitation partitioning may have emerged and the role of vegetation in that mediation. Key Points: Observed long‐term average (equilibrium) soil moisture has a power law relationship with the aridity indexObserved long‐term average scaled runoff and scaled evaporation have power law relationships with equilibrium soil moistureBudyko curve manifests through the soil moisture‐mediated partitioning of precipitation into runoff and evaporation [ABSTRACT FROM AUTHOR]

Despite long‐standing efforts, hydrologists still lack robust tools for calibrating land surface model (LSM) streamflow estimates within ungauged basins. Using surface soil moisture estimates from the Soil Moisture Active Passive Level 4 Soil Moisture (L4_SM) product, precipitation observations, and streamflow gauge measurements for 617 medium‐scale (200–10,000 km2) basins in the contiguous United States, we measure the temporal (Spearman) rank correlation between antecedent (i.e., pre‐storm) surface soil moisture (ASM) and the storm‐scale runoff coefficient (RC; the fraction of storm‐scale precipitation accumulation converted into streamflow). In humid and semi‐humid basins, this rank correlation is shown to be sufficiently strong to allow for the substitution of storm‐scale RC observations (available only in basins that are both lightly regulated and gauged) with high‐quality ASM values (available quasi‐globally from L4_SM) in streamflow calibration procedures. Using this principle, we define a new, basin‐wise LSM streamflow calibration approach based on L4_SM alone and successfully apply it to identify LSM configurations that produce a high rank correlation with observed RC. However, since the approach cannot detect RC bias, it is less successful in identifying LSM configurations with low mean‐absolute error. Plain Language Summary: Accurately forecasting the fraction of rainfall that runs off into streams, as opposed to infiltrates into the soil, is critical for flash‐flood prediction, water‐resource monitoring, and tracking the transport of nutrients from agricultural fields into local waterways. Such forecasting is typically performed by hydrologic models that attempt to represent the physical processes responsible for surface runoff generation. However, to provide accurate streamflow forecasts, these models typically need to be calibrated against actual streamflow observations. This is problematic given the relatively poor, and declining, global availability of stream gauges. This paper presents a novel model calibration strategy that uses soil moisture from remote sensing and numerical modeling in place of streamflow observations during calibration. This transition has significant practical advantages because, unlike streamflow observations, the soil moisture data are continuously available across space. Our results demonstrate that this new approach can significantly improve hydrologic models within humid and semi‐humid basins lacking sufficient ground‐based instrumentation for traditional streamflow calibration. Key Points: Hydrologic models still perform poorly in ungauged basinsIn relative humid areas, the correlation between pre‐storm soil moisture and the storm‐scale runoff coefficient approaches unitySuch strong correlation can be exploited to improve the precision of runoff coefficient estimates provided by models in ungaged basins [ABSTRACT FROM AUTHOR]

Lateral subsurface stormflow (SSF) is the most important runoff generation mechanism for most hillslopes in temperate climates. It is influenced by pedological, biological, and topographic factors that change during landscape evolution, but so far little is known about how SSF changes over long‐time scales. Therefore, we conducted sprinkling experiments on a silicate and carbonate moraine chronosequence in the Swiss Alps. Each chronosequence consisted of four moraines ranging between a couple of decades and ∼13,500 years in age. On each moraine, we installed three plots and measured shallow SSF in a trench. We added tracers (δ2H and NaCl) to the sprinkling water to identify mixing and flow pathways in the subsurface. The coarse and drainable sediments on the young moraines provoked more frequent and larger SSF responses than for the old moraines. There was no SSF during the sprinkling experiments on the older moraines at the calcareous study area, but SSF occurred during larger natural rainfall events. The pre‐event water fractions in SSF were higher for the old moraines than the young moraines due to the increase in silt, clay, and soil organic matter content, and subsequent increase in the amount of water stored in the soil. The results of this study suggest that soil and vegetation development affect SSF characteristics and help—together with the results for overland flow (companion paper; Maier & van Meerveld, 2021, https://doi.org/10.1029/2021WR030221)—to improve hydrological models and our understanding of the changes in near‐surface runoff generation processes during the first millennia of landscape evolution in Alpine areas. Key Points: Lateral subsurface flow (SSF) occurred most frequent for the youngest moraines at the silicate study area where runoff ratios were highestSSF occurred less frequent on the moraines in the carbonate study area than in the silicate area because the soils were less developedThe pre‐event water fractions in SSF increased with moraine age because of the increase in fine‐textured material and soil water retention [ABSTRACT FROM AUTHOR]

The physical parameterization of key processes in land surface models (LSMs) remains uncertain, and new techniques are required to evaluate LSMs accuracy over large spatial scales. Given the role of soil moisture in the partitioning of surface water fluxes (between infiltration, runoff, and evapotranspiration), surface soil moisture (SSM) estimates represent an important observational benchmark for such evaluations. Here, we apply SSM estimates from the NASA Soil Moisture Active Passive Level‐4 product (SMAP_L4) to diagnose bias in the correlation between SSM and surface runoff for multiple Noah‐Multiple Physics (Noah‐MP) LSM parameterization cases. Results demonstrate that Noah‐MP surface runoff parameterizations often underestimate the correlation between prestorm SSM and the event‐scale runoff coefficient (RC; defined as the ratio between event‐scale streamflow and precipitation volumes). This bias can be quantified against an observational benchmark calculated using streamflow observations and SMAP_L4 SSM and applied to explain a substantial fraction of the observed basin‐to‐basin (and case‐to‐case) variability in the skill of event‐scale RC estimates from Noah‐MP. Most notably, a low bias in LSM‐predicted SSM/RC correlation squanders RC information contained in prestorm SSM and reduces LSM RC estimation skill. Based on this concept, a novel case selection strategy for ungauged basins is introduced and demonstrated to successfully identify poorly performing Noah‐MP parameterization cases. Plain Language Summary: Land surface models are commonly tasked with determining what fraction of incoming rainfall infiltrates into the soil versus runs off into stream channels. The key factor determining this partitioning is the amount of water in the soil column prior to a storm event (e.g., more prestorm soil moisture is generally associated with decreased amounts of infiltration and increased surface runoff). However, due to a lack of soil moisture observations available at large scales, it has generally been difficult to assess whether existing models are accurately capturing the true relationship between prestorm soil moisture and runoff. Using newly available data from the NASA Soil Moisture Active/Passive (SMAP) mission, this paper demonstrates that land surface models often misrepresent the impact of prestorm surface soil moisture on runoff generation. This misrepresentation is shown to have a strong negative impact on the ability of models to accurately estimate runoff. A new calibration technique, based on the SMAP Level 4 soil moisture product, is introduced for eliminating this bias. Overall, results demonstrate how remotely sensed soil moisture can potentially play an important role in enhancing the operational forecasting of streamflow. Key Points: Correlation between prestorm surface soil moisture and event‐scale runoff coefficient is an important diagnostic for land surface modelsImproved model representation of this correlation is associated with more skillful estimation of storm‐to‐storm variations in runoffSatellite‐based soil moisture analysis products can be applied to detect bias in model‐based correlation and correct for its impact [ABSTRACT FROM AUTHOR]

The ever increasing pace of computational power, along with continued advances in measurement technologies and improvements in process understanding has stimulated the development of increasingly complex hydrologic models that simulate soil moisture flow, groundwater recharge, surface runoff, root water uptake, and river discharge at different spatial and temporal scales. Reconciling these high-order system models with perpetually larger volumes of field data is becoming more and more difficult, particularly because classical likelihood-based fitting methods lack the power to detect and pinpoint deficiencies in the model structure. Gupta et al. (2008) has recently proposed steps (amongst others) toward the development of a more robust and powerful method of model evaluation. Their diagnostic approach uses signature behaviors and patterns observed in the input-output data to illuminate to what degree a representation of the real world has been adequately achieved and how the model should be improved for the purpose of learning and scientific discovery. In this paper, we introduce approximate Bayesian computation (ABC) as a vehicle for diagnostic model evaluation. This statistical methodology relaxes the need for an explicit likelihood function in favor of one or multiple different summary statistics rooted in hydrologic theory that together have a clearer and more compelling diagnostic power than some average measure of the size of the error residuals. Two illustrative case studies are used to demonstrate that ABC is relatively easy to implement, and readily employs signature based indices to analyze and pinpoint which part of the model is malfunctioning and in need of further improvement. [ABSTRACT FROM AUTHOR]

At the catchment scale, soil moisture patterns have been observed to exhibit different types of dependence on topography. Some catchments have their wettest locations in the valley bottoms, while others have their wettest locations on hillslopes that are oriented away from the sun. Additionally, some catchments have soil moisture patterns that maintain a similar organization at all times (temporal stability), while other catchments have soil moisture patterns that change through time (temporal instability). Although these tendencies are well known, the reasons for their occurrence at a particular catchment are not well understood. In this paper, we investigate conditions under which different types of topographic dependence and different degrees of temporal instability can occur through the use of a new conceptual model. The type of topographic dependence and the degree of instability are quantified by two metrics that are also introduced in the paper, and the effects of soil, vegetation, and climatic parameters on these metrics are then evaluated. The evaluations indicate that saturated horizontal hydraulic conductivity, pore disconnectedness, and a vegetation evapotranspiration efficiency parameter have strong effects on the organization and instability of the soil moisture patterns. In contrast, annual potential evapotranspiration alone has less impact on the organization or its stability. [ABSTRACT FROM AUTHOR]

Predicting catchment stormflow responses after tropical deforestation remains difficult. We used 5‐min rainfall and storm runoff data for 30 events to calibrate the Green–Ampt (GA) and the Spatially Variable Infiltration (SVI) models and predict runoff responses for a small, degraded grassland catchment on Leyte Island (the Philippines), where infiltration‐excess overland flow (IOF) is considered the dominant runoff process. SVI replicated individual stormflow hydrographs better than GA, particularly for events with small runoff responses or multiple peaks. Calibrated parameter values of the SVI model (i.e., spatially averaged maximum infiltration capacity, Im and initial abstraction, F0) varied markedly between events, but were statistically significant and negatively correlated with (mid‐slope) soil moisture content at 10 cm (SWC10)—as did the "catchment‐wide effective" hydraulic conductivity (Ke) of the GA model. Using SWC10‐based estimates of F0 and Im in SVI yielded satisfactory to good simulations for 11 out of 17 events with runoff coefficients ≥15%, but failed to reproduce the hydrographs for events with very small runoff amounts (0.25–1 mm) and low runoff coefficients (3%–6%). The median field‐measured near‐surface Ksat (2 mm hr−1) was distinctly lower than the median Im (32 mm hr−1) and, to a lesser extent, Ke (∼8 mm hr−1), suggesting an underestimation of the spatially averaged Ksat by the field measurements. Application of SVI is expected to give the most realistic results for situations where IOF is dominant, that is, where surface conditions are degraded and rainfall intensities high. Plain Language Summary: It is important to be able to predict the streamflow volume and peak flow rate during intense rainfall events. We used rainfall and streamflow data for a small, degraded tropical grassland catchment on Leyte Island (the Philippines) to calibrate two rainfall infiltration models of comparable complexity: the Green–Ampt model (GA), that describes a decrease in infiltration capacity with time, and the Spatially Variable Infiltration (SVI) model for which the infiltration capacity is a function of the rainfall intensity. SVI generally performed better than GA in simulating observed streamflow responses to rainfall, especially for events with multiple rainfall peaks. Values for the two main parameters of SVI (the amount of rainfall required to initiate stormflow, and the maximum infiltration capacity of the soil) were related to the amount of moisture in the top 10 cm of the soil prior to the rainfall event. Using the measured topsoil moisture contents for 26 rainfall events to estimate the SVI parameter values and predict the stormflow response from the measured rainfall intensity produced satisfactory to good results for most of the larger events. However, it failed to reproduce the stormflow patterns for six events with mostly small to very small runoff responses. Key Points: The Spatially Variable Infiltration (SVI) model outperformed the Green–Ampt model, especially when simulating multi‐peaked storm hydrographsSVI‐model parameter values correlated negatively with antecedent topsoil moisture contentModel‐derived catchment‐scale infiltration capacities were higher than field‐measured Ksat, regardless of the model or field method used [ABSTRACT FROM AUTHOR]

It is demonstrated for the first time how model parameter, structural and data uncertainties can be accounted for explicitly and simultaneously within the Generalized Likelihood Uncertainty Estimation (GLUE) methodology. As an example application, 72 variants of a single soil moisture accounting store are tested as simplified hypotheses of runoff generation at six experimental grassland field-scale lysimeters through model rejection and a novel diagnostic scheme. The fields, designed as replicates, exhibit different hydrological behaviors which yield different model performances. For fields with low initial discharge levels at the beginning of events, the conceptual stores considered reach their limit of applicability. Conversely, one of the fields yielding more discharge than the others, but having larger data gaps, allows for greater flexibility in the choice of model structures. As a model learning exercise, the study points to a 'leaking' of the fields not evident from previous field experiments. It is discussed how understanding observational uncertainties and incorporating these into model diagnostics can help appreciate the scale of model structural error. [ABSTRACT FROM AUTHOR]

Soil moisture is a key control on runoff generation and biogeochemical processes on hillslopes. Precipitation events can evoke different soil moisture responses with depth through the soil profile, and responses can differ among landscape positions along a hillslope. We sought to elucidate the nature of these responses by estimating changes in water content, response time between peak precipitation and peak soil moisture, and wetting front velocities for 43 storms at 45 locations on three adjacent hillslopes within a headwater catchment of the southern Appalachian Mountains (NC, USA). We used a multivariate modeling approach to quantify the relative influences and the predictability of soil moisture responses by a combination of landscape and storm characteristics. We quantified the lag correlations between hillslope mean soil moisture and catchment runoff to demonstrate how storm properties and hillslope‐scale characteristics may influence runoff at the catchment outlet. Soil moisture responses varied widely, and no consistent patterns were observed among response metrics laterally or vertically along hillslopes. In contrast to other studies, we found that the relative influence of hillslope properties and storm characteristics varied with soil moisture responses and during storms. Antecedent conditions and storm depths influenced the strength of lag correlations between soil moisture and runoff, whereas storm mean intensity was correlated with the lag times. These results highlight the utility of intensive observations for characterizing heterogeneity in soil moisture responses, suggesting, among other things, a need for better representation of the subsurface processes in rainfall‐runoff models. Identifying the relative importance of drivers can be beneficial in building parsimonious hydrological models. Key Points: No consistent patterns in soil moisture responses among landscape positions laterally along hillslopesDominant controls on soil moisture responses to storms varied with individual response metrics and during stormsStorm depth, mean intensity, and antecedent conditions mediated the soil moisture‐runoff relationships in space and time [ABSTRACT FROM AUTHOR]

Microwave brightness temperature observations from the NASA Soil Moisture Active Passive (SMAP) mission and gauge‐based precipitation data over the United States are used to reconstruct the soil water loss function and then historical (1979–2019) hydrological fluxes in the form of evapotranspiration (ET) and drainage (D) are quantified. Over the period of study, with the exception of snowy and hyper‐arid regions, we observe a correlation of R2 > 0.6 between SMAP‐precipitation derived drainage estimates and streamflow measurements from the U.S. Geological Survey (USGS). There is a bias between estimated drainage and USGS streamflow with an underestimation of about 1 mm day−1 in southwest United States to 3 mm day−1 in parts of the eastern United States. SMAP‐derived sensitivities of drainage and ET partitioning with respect to precipitation anomalies are also calculated. In parts of the Great Plains the drainage partitioning exhibits a near‐linear response, while in the southeast United States, the response is nonlinear. Partitioning also is examined for 6 four‐digit hydrologic unit basins wherein year‐to‐year variations in drainage partitioning are shown to be key mediators in translating precipitation anomalies into streamflow anomalies. Observation‐driven drainage and ET estimates are obtained without relying on full hydrologic and Land Surface Models (LSMs). This independence (isolation from model parameterization assumptions) provides a path toward using satellite‐derived landscape hydrological diagnostics to assess hydrologic models and LSMs as well as to guide their further development. Key Points: A large amount of hydrologic information is encoded within satellite remote sensing observations of soil moistureSMAP‐era soil moisture and precipitation information, without reliance on models, can characterize landscape hydrological partitioningHistorical remote sensing and precipitation‐based drainage estimates closely track streamflow measurements across the United States [ABSTRACT FROM AUTHOR]

This study proposes a new process‐based framework to characterize and classify runoff events of various magnitudes occurring in a wide range of catchments. The framework uses dimensionless indicators that characterize space–time dynamics of precipitation events and their spatial interaction with antecedent catchment states, described as snow cover, distribution of frozen soils, and soil moisture content. A rigorous uncertainty analysis showed that the developed indicators are robust and regionally consistent. Relying on covariance‐ and ratio‐based indicators leads to reduced classification uncertainty compared to commonly used (event‐based) indicators based on absolute values of metrics such as duration, volume, and intensity of precipitation events. The event typology derived from the proposed framework is able to stratify events that exhibit distinct hydrograph dynamics even if streamflow is not directly used for classification. The derived typology is therefore able to capture first‐order controls of event runoff response in a wide variety of catchments. Application of this typology to about 180,000 runoff events observed in 392 German catchments revealed six distinct regions with homogeneous event type frequency that match well regions with similar behavior in terms of runoff response identified in Germany. The detected seasonal pattern of event type occurrence is regionally consistent and agrees well with the seasonality of hydroclimatic conditions. The proposed framework can be a useful tool for comparative analyses of regional differences and similarities of runoff generation processes at catchment scale and their possible spatial and temporal evolution. Key Points: A new process‐based framework for characterizing runoff events of diverse sizes is proposed and applied to a wide set of catchmentsNovel indicators combining the space–time dynamics of event precipitation and catchment state are used for classificationThe derived event typology captures distinct shapes of event hydrographs even if streamflow data are not directly used for classification [ABSTRACT FROM AUTHOR]

Long short‐term memory (LSTM) networks offer unprecedented accuracy for prediction in ungauged basins. We trained and tested several LSTMs on 531 basins from the CAMELS data set using k‐fold validation, so that predictions were made in basins that supplied no training data. The training and test data set included ∼30 years of daily rainfall‐runoff data from catchments in the United States ranging in size from 4 to 2,000 km2 with aridity index from 0.22 to 5.20, and including 12 of the 13 IGPB vegetated land cover classifications. This effectively "ungauged" model was benchmarked over a 15‐year validation period against the Sacramento Soil Moisture Accounting (SAC‐SMA) model and also against the NOAA National Water Model reanalysis. SAC‐SMA was calibrated separately for each basin using 15 years of daily data. The out‐of‐sample LSTM had higher median Nash‐Sutcliffe Efficiencies across the 531 basins (0.69) than either the calibrated SAC‐SMA (0.64) or the National Water Model (0.58). This indicates that there is (typically) sufficient information in available catchment attributes data about similarities and differences between catchment‐level rainfall‐runoff behaviors to provide out‐of‐sample simulations that are generally more accurate than current models under ideal (i.e., calibrated) conditions. We found evidence that adding physical constraints to the LSTM models might improve simulations, which we suggest motivates future research related to physics‐guided machine learning. Key Points: Overall accuracy of LSTMs in ungauged basins is comparable to standard hydrology models in gauged basinsThere is sufficient information in catchment characteristics data to differentiate between catchment‐specific rainfall‐runoff behaviors [ABSTRACT FROM AUTHOR]

The precipitation partitioning between evapotranspiration (ET) and runoff (R) at the land surface is controlled by atmospheric boundary layer and terrestrial hydrological processes. These processes in land surface models are manifested primarily as stomatal conductance, soil moisture limitation factor to transpiration (β‐factor), turbulence, and runoff generation. What are the sensitivities of precipitation partitioning to the parameterizations of these processes? To address this overarching question, the annual and seasonal means of ET and R over the conterminous United States were simulated using 48 configurations of the Noah land surface model with multiparameterization options (Noah‐MP). The Sobol' total sensitive index was used to quantify the sensitivity of ET and R to the parameterizations of the four processes mentioned above. Results show that the sensitivities of the annual means depend on climatic conditions and the interplay between ET and R plays an important role. In humid regions, precipitation is mostly partitioned into R, whereas the simulations can be more sensitive to ET's parameterizations. In arid regions, ET accounts for the major partition, whereas the simulations can be more sensitive to the runoff parameterization. Seasonal means exhibit different sensitivities from the annual means. The seasonal mean ET is more sensitive to ET's parameterizations, and R is more sensitive to the runoff parameterization. The β‐factor, which is neglectable for the annual means, is important for summer‐time ET. Mediated by the terrestrial water storage memories, ET interplays R across seasons. The winter‐time R is still sensitive to the stomatal conductance that only modulates growing‐season ET. Key Points: The sensitivity of precipitation partitioning to land surface parameterizations is quantified using the Sobol' sensitivity indexThe parameterization sensitivity depends on climatic conditions and seasons and varies with the time scaleThe influence of stomatal conductance can extend beyond growing seasons due to the terrestrial water storage memories [ABSTRACT FROM AUTHOR]

Most field-based approaches that address runoff generation questions have been conducted in steep landscapes with shallow soils. Runoff generation processes in low relief landscapes with deep soils remain less understood. We addressed this by characterizing dominant runoff generating flow paths by monitoring the timing and magnitude of precipitation, runoff, shallow soil moisture, and shallow and deep groundwater dynamics in a 3.3 ha ephemeral-to-intermittent drainage network in the Piedmont region of North Carolina, USA. This Piedmont region is gently sloped with highly weathered soils characterized by shallow impeding layers due to decreases in saturated hydraulic conductivity with depth. Our results indicated two dominant catchment storage states driven by seasonal evapotranspiration. Within these states, distinct flow paths were activated, resulting in divergent hydrograph recessions. Groundwater dynamics during precipitation events with different input characteristics and contrasting storage states showed distinct shallow and deep groundwater flow path behavior could produce similar runoff magnitudes. During an event with low antecedent storage, activation of a shallow, perched, transient water table dominated runoff production. During an event with high antecedent storage, the deeper water table activated shallow flow paths by rising into the shallow transmissive soil horizons. Despite these differing processes, the relationship between active surface drainage length (ASDL) and runoff was consistent. Hysteretic behavior between ASDL and runoff suggested that while seasonal ASDLs can be predicted based on runoff, the mechanisms and source areas producing flow can be highly variable and not easily estimated from runoff alone. These processes and flow paths have significant implications for stream chemistry across seasons and storage states. [ABSTRACT FROM AUTHOR]

Since its introduction in 1954, the Soil Conservation Service curve number (SCS-CN) method has become the standard tool, in practice, for estimating an event-based rainfall-runoff response. However, because of its empirical origins, the SCS-CN method is restricted to certain geographic regions and land use types. Moreover, it does not describe the spatial variability of runoff. To move beyond these limitations, we present a new theoretical framework for spatially lumped, event-based rainfall-runoff modeling. In this framework, we describe the spatially lumped runoff model as a point description of runoff that is upscaled to a watershed area based on probability distributions that are representative of watershed heterogeneities. The framework accommodates different runoff concepts and distributions of heterogeneities, and in doing so, it provides an implicit spatial description of runoff variability. Heterogeneity in storage capacity and soil moisture are the basis for upscaling a point runoff response and linking ecohydrological processes to runoff modeling. For the framework, we consider two different runoff responses for fractions of the watershed area: 'prethreshold' and 'threshold-excess' runoff. These occur before and after infiltration exceeds a storage capacity threshold. Our application of the framework results in a new model (called SCS-CNx) that extends the SCS-CN method with the prethreshold and threshold-excess runoff mechanisms and an implicit spatial description of runoff. We show proof of concept in four forested watersheds and further that the resulting model may better represent geographic regions and site types that previously have been beyond the scope of the traditional SCS-CN method. [ABSTRACT FROM AUTHOR]

In forested catchments, the exceedance of rainfall and antecedent water storage thresholds is often required for runoff generation, yet to our knowledge these threshold relationships remain undescribed in tropical dry forest catchments. We, therefore, identified the controls of streamflow activation and the timing and magnitude of runoff in a tropical dry forest catchment near the Pacific coast of central Mexico. During a 52 day transition phase from the dry to wet season, soil water movement was dominated by vertical flow which continued until a threshold soil moisture content of 26% was reached at 100 cm below the surface. This satisfied a 162 mm storage deficit and activated streamflow, likely through lateral subsurface flow pathways. High antecedent soil water conditions were maintained during the wet phase but had a weak influence on stormflow. We identified a threshold value of 289 mm of summed rainfall and antecedent soil water needed to generate >4 mm of stormflow per event. Above this threshold, stormflow response and magnitude was almost entirely governed by rainfall event characteristics and not antecedent soil moisture conditions. Our results show that over the course of the wet season in tropical dry forests the dominant controls on runoff generation changed from antecedent soil water and storage to the depth of rainfall. [ABSTRACT FROM AUTHOR]

We examined 3 years of measured daily values of all major water budget components (precipitation P, potential evapotranspiration PET, actual evapotranspiration ET, and runoff R) and volumetric soil water content θ of a small, forested catchment located in the west of Germany. The spatial distribution of θ was determined from a wireless sensor network of 109 points with 3 measurement depths each; ET was calculated from eddy-covariance tower measurements. The water budget was dominantly energy limited, with ET amounting to approximately 90% of PET, and a runoff ratio R/P of 56%. P, ET, and R closed the long-term water budget with a residual of 2% of precipitation. On the daily time scale, the residual of the water budget was larger than on the annual time scale, and explained to a moderate extent by θ ( R2 = 0.40). Wavelet analysis revealed subweekly time scales, presumably dominated by unaccounted fast-turnover storage terms such as interception, as a major source of uncertainty in water balance closure. At weekly resolution, soil water content explained more than half ( R2 = 0.62) of the residual. By means of combined empirical orthogonal function and cluster analysis, two slightly different spatial patterns of θ could be identified that were associated with mean θ values below and above 0.35 cm3/cm3, respectively. The timing of these patterns as well as the varying coherence between PET, ET, and soil water content responded to changes in water availability, including a moderate response to the European drought in spring 2011. [ABSTRACT FROM AUTHOR]

The computational effort associated with physically based distributed hydrological models is one of their major limitations that restrict their application in soil moisture and land surface flux simulation problems for large catchments. In this work, a new approach for reducing the computational effort associated with such models is investigated. This approach involves the formation of equivalent cross sections, designed in a manner that ensures comparable accuracy in simulating the hydrological fluxes as a fully distributed simulation. Single or multiple equivalent cross sections are formulated in each Strahler's first-order subbasin on the basis of topographic and physiographic variables representing the entire or part of the subbasin. An unsaturated soil moisture movement model based on a two-dimensional solution of the Richards' equation is used for simulating the soil moisture and hydrologic fluxes. The equivalent cross-section approach and the model are validated against observed soil moisture data in a semiarid catchment and found consistent. The results indicate that the equivalent cross-section approach is an efficient alternative for reducing the computational time of distributed hydrological modeling while maintaining reasonable accuracy in simulating hydrologic fluxes, in particular dominant fluxes such as transpiration and soil evaporation in semiarid catchments. [ABSTRACT FROM AUTHOR]

The relation between storage and discharge is an essential characteristic of many rainfall-runoff models. The simple dynamical systems approach, in which a rainfall-runoff model is constructed from a single storage-discharge relation, has been successfully applied to humid catchments. Here we investigate (1) if and when the less humid lowland Hupsel Brook catchment also behaves like a simple dynamical system by hydrograph fitting, and (2) if system parameters can be inferred from streamflow recession rates or more directly from soil moisture storage observations. Only 39% of the fitted monthly hydrographs yielded Nash-Sutcliffe efficiencies above 0.5, from which we can conclude that the Hupsel Brook catchment does not always behave like a simple dynamical system. Model results were especially poor in summer, when evapotranspiration is high and the thick unsaturated zone attenuates the rainfall input. Using soil moisture data to obtain system parameters is not trivial, mainly because there is a discrepancy between local and catchment storage. Parameters obtained with direct storage-discharge fitting led to a strong underestimation of the response of runoff to rainfall, while recession analysis leads to an overestimation. [ABSTRACT FROM AUTHOR]

An innovative algorithm, shuffled complexes with principal components analysis (SP-UCI), is developed to overcome a critical deficiency of the shuffled complex evolution scheme: population degeneration. Population degeneration means that, during the evolutionary search process, the population of search particles may degenerate into a subspace of the full parameter space, thereby missing the capacity of fully exploring the parameter space. Being confined in a subspace may even lead the particle population to converge to nonstationary points, which is a fatal malfunction. To overcome this problem, SP-UCI employs the principal components analysis to detect the occurrence of population degeneration and remedy the adverse effects. The ensemble of calibrations of the Sacramento soil moisture accounting model with the SP-UCI method over the Leaf River basin, Mississippi, retrieves the optimal parameter values with the lowest recorded root-mean-squared error of simulated daily runoff against the observation. Moreover, the result also provides consistent (narrow ranges) model parameter distribution, which results in a better understanding of the model's behavior, given the watershed's hydrologic features. [ABSTRACT FROM AUTHOR]

We run a comparative study of ecohydrological models of streamflow probability distributions (pdfs), p( Q), derived by Botter et al. (2007a, 2009), against field data gathered in different hydrological contexts. Streamflows measured in several catchments across various climatic regions of northeastern Italy and the United States are employed. The relevance of the work stems from the implied analytical predictive ability of hydrologic variability, whose role on stream and riparian ecological processes and large-scale management schemes is fundamental. The tools employed are analytical models of p( Q) (and of the related flow duration curve, D( Q)) derived by coupling suitable storage-discharge relations with a stochastic description of streamflow production through soil moisture dynamics, and are expressed as a function of few macroscopic rainfall, soil, vegetation and geomorphological parameters. In this work we compare the performances of a recent version of the model (which includes the effects of nonlinear subsurface storage-discharge relations) to those provided by the linear version through the application of the models to 13 test catchments belonging to various climatic and geomorphic contexts. A general agreement between predicted and observed daily streamflows pdfs is shown, though differences emerge between the linear and the nonlinear approaches. In particular, by including the effects of a nonlinear storage-discharge relation the model accuracy is shown to increase with respect to the linear scheme in most examined cases. We show that this is not simply attributable to the added parameter but corresponds to a proper likelihood increase. The nonlinear model is shown to exhibit three basic forms for p( Q) (monotonically decreasing with an atom of probability in Q = 0, bell-shaped with the mode close to zero, bell-shaped with the mode close to the mean), corresponding to different hydrologic regimes which are clearly detectable in field data. Inferences on the nonlinear character of the relation between subsurface storage and discharge from observed p( Q) are finally drawn. [ABSTRACT FROM AUTHOR]

Understanding the space-time variability of runoff has important implications for climate because of the linkage of runoff and evapotranspiration and is a practical concern as well for the prediction of drought and floods. In contrast to many studies investigating the space-time variability of precipitation and temperature, there has been relatively little work evaluating climate teleconnections of runoff, in part because of the absence of data sets that lend themselves to commonly used techniques in climate analysis like principal components analysis. We examine the space-time variability of runoff over North America using a 50-year retrospective spatially distributed data set of runoff and other land surface water cycle variables predicted using a calibrated macroscale hydrology model, thus avoiding some shortcomings of past studies based more directly on streamflow observations. We determine contributions to runoff variability of climatic teleconnections, soil moisture, and snow for lead times up to a year. High and low values of these sources of predictability are evaluated separately. We identify patterns of runoff variability that are not revealed by direct analysis of observations, especially in areas of sparse stream gauge coverage. The presence of nonlinear relationships between large-scale climate changes and runoff pattern variability, as positive and negative values of the large-scale climate indices rarely show opposite teleconnections with a runoff pattern. Dry soil moisture anomalies have a stronger influence on runoff variability than wet soil. Snow, and more so soil moisture, in many locations enhance the predictability due to climatic teleconnections. [ABSTRACT FROM AUTHOR]