Including spatial distribution in a data‐driven rainfall‐runoff model to improve reservoir inflow forecasting in Taiwan

Multi‐step ahead inflow forecasting has a critical role to play in reservoir operation and management in Taiwan during typhoons as statutory legislation requires a minimum of 3‐h warning to be issued before any reservoir releases are made. However, the complex spatial and temporal heterogeneity of typhoon rainfall, coupled with a remote and mountainous physiographic context, makes the development of real‐time rainfall‐runoff models that can accurately predict reservoir inflow several hours ahead of time challenging. Consequently, there is an urgent, operational requirement for models that can enhance reservoir inflow prediction at forecast horizons of more than 3 h. In this paper, we develop a novel semi‐distributed, data‐driven, rainfall‐runoff model for the Shihmen catchment, north Taiwan. A suite of Adaptive Network‐based Fuzzy Inference System solutions is created using various combinations of autoregressive, spatially lumped radar and point‐based rain gauge predictors. Different levels of spatially aggregated radar‐derived rainfall data are used to generate 4, 8 and 12 sub‐catchment input drivers. In general, the semi‐distributed radar rainfall models outperform their less complex counterparts in predictions of reservoir inflow at lead times greater than 3 h. Performance is found to be optimal when spatial aggregation is restricted to four sub‐catchments, with up to 30% improvements in the performance over lumped and point‐based models being evident at 5‐h lead times. The potential benefits of applying semi‐distributed, data‐driven models in reservoir inflow modelling specifically, and hydrological modelling more generally, are thus demonstrated. Copyright © 2012 John Wiley & Sons, Ltd.     

A new magnitude category disaggregation approach for temporal high‐resolution rainfall intensities

Simulation of quick runoff components such as surface runoff and associated soil erosion requires temporal high‐resolution rainfall intensities. However, these data are often not available because such measurements are costly and time consuming. Current rainfall disaggregation methods have shortcomings, especially in generating the distribution of storm events. The objectives of this study were to improve point rainfall disaggregation using a new magnitude category rainfall disaggregation approach. The procedure is introduced using a coupled disaggregation approach (Hyetos and cascade) for multisite rainfall disaggregation. The new procedure was tested with ten long‐term precipitation data sets of central Germany using summer and winter precipitation to determine seasonal variability. Results showed that dividing the rainfall amount into four daily rainfall magnitude categories (1–10, 11–25, 26–50, >50 mm) improves the simulation of high rainfall intensity (convective rainfall). The Hyetos model category approach (Hyetos_Cat) with seasonal variation performs representative to observed hourly rainfall compared with without categories on each month. The mean absolute percentage accuracy of standard deviation for hourly rainfall is 89.7% in winter and 95.6% in summer. The proposed magnitude category method applied with the coupled Hyetos_Cat–cascade approach reproduces successfully the statistical behaviour of local 10‐min rainfall intensities in terms of intermittency as well as variability. The root mean square error performance statistics for disaggregated 10‐min rainfall depth ranges from 0.20 to 2.38 mm for summer and from 0.12 to 2.82 mm for the winter season in all categories. The coupled stochastic approach preserves the statistical self‐similarity and intermittency at each magnitude category with a relatively low computational burden. Copyright © 2014 John Wiley & Sons, Ltd.     

Severe soil erosion during a 3‐day exceptional rainfall event: combining modelling and field data for a fallow cereal field

Exceptional rainfall events cause significant losses of soil, although few studies have addressed the validation of model predictions at field scale during severe erosive episodes. In this study, we evaluate the predictive ability of the enhanced Soil Erosion and Redistribution Tool (SERT‐2014) model for mapping and quantifying soil erosion during the exceptional rainfall event (~235 mm) that affected the Central Spanish Pyrenees in October 2012. The capacity of the simulation model is evaluated in a fallow cereal field (1.9 ha) at a high spatial scale (1 × 1 m). Validation was performed with field‐quantified rates of soil loss in the rills and ephemeral gullies and also with a detailed map of soil redistribution. The SERT‐2014 model was run for the six rainfall sub‐events that made up the exceptional event, simulating the different hydrological responses of soils with maximum runoff depths ranging between 40 and 1017 mm. Predicted average and maximum soil erosion was 11 and 117 Mg ha^?1 event^?1, respectively. Total soil loss and sediment yield to the La Reina gully amounted to 16.3 and 9.0 Mg event^?1. These rates are in agreement with field estimations of soil loss of 20.0 Mg event^?1. Most soil loss (86%) occurred during the first sub‐event. Although soil accumulation was overestimated in the first sub‐event because of the large amount of detached soil, the enhanced SERT‐2014 model successfully predicted the different spatial patterns and values of soil redistribution for each sub‐event. Further research should focus on stream transport capacity. Copyright © 2014 John Wiley & Sons, Ltd.     

A long‐term study of stable isotopes as tracers of processes governing water flow and quality in a lowland river basin: the upper Thames,UK

A long‐term study of O, H and C stable isotopes has been undertaken on river waters across the 7000‐km² upper Thames lowland river basin in the southern UK. During the period, flow conditions ranged from drought to flood. A 10‐year monthly record (2003–2012) of the main River Thames showed a maximum variation of 3‰ (δ¹⁸O) and 20‰ (δ²H), although interannual average values varied little around a mean of –6.5‰ (δ¹⁸O) and –44‰ (δ²H). A δ²H/δ¹⁸O slope of 5.3 suggested a degree of evaporative enrichment, consistent with derivation from local rainfall with a weighted mean of –7.2‰ (δ¹⁸O) and –48‰ (δ²H) for the period. A tendency towards isotopic depletion of the river with increasing flow rate was noted, but at very high flows (>100 m³/s), a reversion to the mean was interpreted as the displacement of bank storage by rising groundwater levels (corroborated by measurements of specific electrical conductivity). A shorter quarterly study (October 2011–April 2013) of isotope variations in 15 tributaries with varying geology revealed different responses to evaporation, with a well‐correlated inverse relationship between Δ¹⁸O and baseflow index for most of the rivers. A comparison with aquifer waters in the basin showed that even at low flow, rivers rarely consist solely of isotopically unmodified groundwater. Long‐term monitoring (2003–2007) of carbon stable isotopes in dissolved inorganic carbon (DIC) in the Thames revealed a complex interplay between respiration, photosynthesis and evasion, but with a mean interannual δ¹³C‐DIC value of –14.8 ± 0.5‰, exchange with atmospheric carbon could be ruled out. Quarterly monitoring of the tributaries (October 2011–April 2013) indicated that in addition to the aforementioned factors, river flow variations and catchment characteristics were likely to affect δ¹³C‐DIC. Comparison with basin groundwaters of different alkalinity and δ¹³C‐DIC values showed that the origin of river baseflow is usually obscured. The findings show that long‐term monitoring of environmental tracers can help to improve the understanding of how lowland river catchments function. Copyright © NERC 2015. Hydrological Processes © 2015 John Wiley & Sons, Ltd.     

Seasonal variations in infiltrability of moss‐dominated biocrusts on aeolian sand and loess soil in the Chinese Loess Plateau

Biocrust effects on soil infiltration have attracted increasing attention in dryland ecosystems, but their seasonal variations in infiltrability have not yet been well understood. On the Chinese Loess Plateau, soil infiltrability indicated by saturated hydraulic conductivity (K_s) of biocrusts and bare soil, both on aeolian sand and loess soil, was determined by disc infiltrometer in late spring (SPR), midsummer (SUM), and early fall (FAL). Then their correlations with soil biological and physiochemical properties and water repellency index (RI) were analysed. The results showed that the biocrusts significantly decreased K_s both on sand during SPR, SUM, and FAL (by 43%, 66%, and 35%, respectively; P < .05) and on loess (by 42%, 92%, and 10%, respectively; P <.05). As compared with the bare soil, the decreased K_s in the biocrusted surfaces was mostly attributed to the microorganism biomass and also to the increasing content of fine particles and organic matter. Most importantly, both the biocrusts and bare soil exhibited significant (F ≥ 11.89, P ≤ .003) seasonal variations in K_s, but their patterns were quite different. Specifically, the K_s of bare soil gradually decreased from SPR to SUM (32% and 42% for sand and loess, respectively) and FAL (29% and 39%); the K_s of biocrusts also decreased from SPR to SUM (59% and 92%) but then increased in FAL (36% and 588%). Whereas the seasonal variations in K_s of the biocrusts were closely correlated with the seasonal variations in RI, the RI values were not high enough to point at hydrophobicity. Instead of that, the seasonal variations of K_s were principally explained by the changes in the crust biomass and possibly by the microbial exopolysaccharides. We conclude that the biocrusts significantly decreased soil infiltrability and exhibited a different seasonal variation pattern, which should be carefully considered in future analyses of hydropedological processes.     

Effect of rainfall uncertainty on the performance of physically based rainfall–runoff models

This paper analyses the effect of rain data uncertainty on the performance of two hydrological models with different spatial structures: a semidistributed and a fully distributed model. The study is performed on a small catchment of 19.6 km² located in the north‐west of Spain, where the arrival of low pressure fronts from the Atlantic Ocean causes highly variable rainfall events. The rainfall fields in this catchment during a series of storm events are estimated using rainfall point measurements. The uncertainty of the estimated fields is quantified using a conditional simulation technique. Discharge and rain data, including the uncertainty of the estimated rainfall fields, are then used to calibrate and validate both hydrological models following the generalized likelihood uncertainty estimation (GLUE) methodology. In the storm events analysed, the two models show similar performance. In all cases, results show that the calibrated distribution of the input parameters narrows when the rain uncertainty is included in the analysis. Otherwise, when rain uncertainty is not considered, the calibration of the input parameters must account for all uncertainty in the rainfall–runoff transformation process. Also, in both models, the uncertainty of the predicted discharges increase in similar magnitude when the uncertainty of rainfall input increase.     

Dissolved organic matter in surface runoff in the Loess Plateau of China: The role of rainfall events and land-use

Exploring the chemical characterization of dissolved organic matter (DOM) is important for understanding the fate of laterally transported organic matter in watersheds. We hypothesized that differences in water-extractable organic matter (WEOM) in soils of varying land uses and rainfall events may significantly affect the quality and the quantity of stream DOM. To test our hypotheses, characteristics of rainfall-runoff DOM and WEOM of source materials (topsoil from different land uses and gullies, as well as typical vegetation) were investigated at two adjacent catchments in the Loess Plateau of China, using ultraviolet–visible absorbance and excitation emission matrix fluorescence with parallel factor analysis (PARAFAC). Results indicated that land-use types may significantly affect the chemical composition of soil WEOM, including its aromaticity, molecular weight, and degree of humification. The PARAFAC analysis demonstrated that the soils and stream water were dominated by terrestrial/allochthonous humic-like substances and microbial transformable humic-like fluorophores. Shifts in the fluorescence properties of stream DOM suggested a pronounced change in the relative proportion of allochthonous versus autochthonous material under different rainfall patterns and land uses. For example, high proportions of forestland could provide more allochthonous DOM input. This study highlights the relevance of soils and hydrological dynamics on the composition and fluxes of DOM issuing from watersheds. The composition of DOM in soils was influenced by land-use type. Precipitation patterns influenced the proportion of terrestrial versus microbial origins of DOM in surface runoff. Contributions of allochthonous, terrestrially derived DOM inputs were highest from forested landscapes.     

Quantification of river water infiltration in shallow aquifers using acesulfame and anthropogenic gadolinium

This study has investigated the use of the artificial sweetener acesulfame and the magnetic resonance imaging contrast agent gadolinium as quantitative tracers for river water infiltration into shallow groundwater. The influence of a river on alluvial groundwater in a subalpine catchment in western Europe has been assessed using the ‘classical’ hydrochemical tracer chloride and the trace contaminants acesulfame and anthropogenic gadolinium. Mixing ratios for riverine bank filtrate with ambient groundwater and the uncertainties associated with the temporal and spatial tracer variability were calculated using acesulfame and gadolinium and compared with those obtained using chloride. The temporal variability of tracer concentrations in river water of gadolinium (standard deviation SD: 63%) and acesulfame (SD: 71%) both exceeded that of chloride (SD: 27%), and this was identified as the main source of uncertainty in the mixing analysis. Similar spatial distributions were detected in the groundwater for chloride and gadolinium, but not for acesulfame. Mixing analyses using acesulfame resulted in calculated mixing ratios that differed from those obtained using gadolinium and chloride by up to 83% and 92%, respectively. At the investigated site, which had oxic conditions and moderate temperatures, acesulfame was found to be a less reliable tracer than either gadolinium or chloride, probably because of natural attenuation and input from other sources. There was no statistically significant difference between the mixing ratios obtained using chloride or gadolinium, the mixing ratios obtained using gadolinium were 40–50% lower than those obtained using chloride. This is mainly due to a bias of the mean gadolinium concentration in river water towards higher values. In view of the uncertainties of the two tracers, neither could be preferred over the other for the quantification of bank filtrate in groundwater. At this specific site gadolinium was able to reliably identify river water infiltration and was a more precise tracer than chloride at low mixing ratios (<20%), because of the exclusive occurrence of gadolinium in river water and its high dynamic range. Copyright © 2015 John Wiley & Sons, Ltd.     

Effects of differing coverage of moss‐dominated soil crusts on hydrological processes and implications for disturbance in the Mu Us Sandland,China

To study the effects of biological soil crusts (BSCs) on hydrological processes and their implications for disturbance in the Mu Us Sandland, the water infiltration, evaporation and soil moisture of high coverage (100% BSCs), middle coverage (40% BSCs) and low coverage (0% BSCs, bare sand) of moss‐dominated crusts were conducted in this study, respectively. The conclusions are as follows: (1) the main effects of moss‐dominated crusts in the Mu Us Sandland on the infiltration of rainwater were to reduce the infiltration depths and to retain the limited rainwater in shallow soil; (2) moss‐dominated crusts have no significant effects on daily evaporation when the volumetric water content at 4 cm depth in 100% BSCs (VWC4) was over 24.7%, on enhanced daily evaporation when the VWC4 ranged from 6.5% to 24.7% and on reduced daily evaporation when the VWC4 was less than 6.5%; and (3) decreasing the coverage of moss‐dominated crusts (from 100% to 40%) did not significantly change its effects on infiltration, evaporation and soil moisture. Our results demonstrated that for the growth and regeneration of shrubs, which were dominated by Artemisia ordosica in the Mu Us Sandland, high coverage of moss‐dominated crusts has negative effects on hydrological processes, and these negative effects could not be significantly reduced by decreasing the coverage of moss‐dominated crusts from 100% to 40%. Therefore, for the sustained and healthy development of shrub communities in the Mu Us Sandland, it is necessary to take appropriate measures for the well‐developed BSCs in the sites with high vegetation coverage in the rainy season. Copyright © 2015 John Wiley & Sons, Ltd.