Rainfall Uncertainty Research Articles

Mulching upland rice for efficient water management: A collaborative approach in Benin

Since the 2007 food crisis, the Government of Benin has created enabling conditions that urge small scale farmers to intensify rice production in inland valleys. Unsuitable irrigation infrastructures and lack of willingness to act collectively has led farmers to complain about water shortage. Adoption of mitigating technologies offered by government (extension) and scientists (e.g. Benin Agricultural Research Institute, AfricaRice) has been low due to the institutional context within which farmers are working. A socio-technical approach which combines technical and institutional dimensions was used to identify and test mulching as a potential method for improving irrigation water efficiency in growing rice in upland parts of Benin's inland valleys. Rice farmers (from three production areas), an extension agent and a researcher formed a multi-stakeholder platform and collaborated to test the application of mulch (three doses) and the use of a lowland rice variety in replacement of an upland rice variety during two growing seasons. Multiple methods derived from researchers and farmers’ perspectives were used to evaluate trial results: quantitative scientific evidence was combined with qualitative evaluation using indicators agreed upon by the collaborative group. Results show that the lowland rice variety IR-841 with 10Mgha−1 ‘rice-straw’ mulch allows farmers to better use available water in the upland areas and increase yields. Although the preference for IR-841 over the special bred upland variety Nerica-4 is risky because of its high water demand and the uncertainty in rainfall, farmers use IR-841 for profit maximisation. Beyond its technical output, the joint experimentation facilitated the interaction of knowledge, experiences and practices among the involved actors.

Efficient Distribution of Water Between Head-Reach and Tail-End Farms in Developing Countries

The necessity of surface water for irrigation and its increasing scarcity in developing economies motivate the need for its efficient distribution. The inequity in the distribution of surface water arises because of the relative physical locations of the farms. Head-reach (primary) farms are close to the source, whereas tail-end (secondary) farms are relatively farther. The lack of physical infrastructure implies that water allocated to secondary farms must pass through primary farms. Left to their individual incentives, primary farmers use more than their fair share of water by denying its release to secondary farmers. Such an inequitable sharing results in significantly suboptimal productivity of the farming community as a whole. We propose decentralized, individually rational mechanisms to achieve socially optimal distribution of surface water for a farming community under uncertainty in rainfall, choice of multiple crops, and differing risk-bearing abilities of primary and secondary farmers. We show that the mechanisms can be efficiently computed and highlight the impact of the improved sharing of surface water. We also study the movement of the price of water with its scarcity. Ideas that can help administer the mechanisms in practice are briefly discussed.

Prospects for flash flood forecasting in mountainous regions – An investigation of Tropical Storm Fay in the Southern Appalachians

Summary The sensitivity of Quantitative flash-Flood Estimates (QFEs) and Quantitative flash-Flood Forecasts (QFFs) to Quantitative Precipitation Estimates (QPEs) and Quantitative Precipitation Forecasts (QPFs) was investigated in three headwater catchments with different topographic and hydro-geomorphic characteristics during the passage of Tropical Storm Fay, 2008 over the Southern Appalachian Mountains in North Carolina, USA. QFEs and QFFs were generated by a high-resolution hydrologic model (250 × 250 m 2 ) with coupled surface–subsurface physics and rainfall forcing from the Next Generation Multi-sensor QPE (Q2) spatial rainfall (1 × 1 km 2 ) product, and from National Digital Forecast Database (NDFD) operational QPF product (5 × 5 km 2 ). Optimal QPE products (Q2+) were derived by assimilating rainfall observations from a high density raingauge network through adaptive bias correction. Deterministic QFEs simulated by the hydrologic model agree well with streamgauge observations (15-min intervals) regarding total water volume and peak flow with Nash–Sutcliffe (NS) coefficients 0.8–0.9, thus suggesting that the model without calibration captures well the dominant flash-flood physics. The propagation of uncertainty in storm rainfall to rainfall–runoff response was subsequently evaluated through model simulations forced by Monte Carlo replications of the QPEs to generate QFE distributions. Analysis of the joint QPE–QFE distributions shows that flood response at the catchment scale is highly non-linear, and exhibits strong dependence on basin physiography, initial soil moisture conditions (transient basin storage capacity), the space–time organization of runoff generation and conveyance mechanisms, and in particular interflow dynamics, with respect to the space–time structure of rainfall. QFFs for 6- to 1-h lead times using precipitation composites of Q2 QPE and NDFD QPF to drive the hydrology model in operational mode exhibited ubiquitous lack of skill yielding consistently negative NS scores. An experiment consisting of merging satellite-like observations into operational QPE/QPF showed significant improvement in QFF performance (e.g. 5–50% relative NS increases), especially when the timing of satellite overpass is such that it captures transient episodes of heavy rainfall during the event. Future advances in QFF remain principally constrained by progress in QPE and QPF at the spatial resolution necessary to resolve rainfall-interflow dynamics in mountainous regions.

The impact of Best Management Practices on simulated streamflow and sediment load in a Central Brazilian catchment

The intense use of water for both public supply and agricultural production causes societal conflicts and environmental problems in the Brazilian Federal District. A serious consequence of this is nonpoint source pollution which leads to increasing water treatment costs. Hence, this study investigates in how far agricultural Best Management Practices (BMPs) might contribute to sustainable water resources management and soil protection in the region. The Soil and Water Assessment Tool (SWAT) was used to study the impact of those practices on streamflow and sediment load in the intensively cropped catchment of the Pipiripau River. The model was calibrated and validated against measured streamflow and turbidity-derived sediment loads. By means of scenario simulations, it was found that structural BMPs such as parallel terraces and small sediment basins (‘Barraginhas’) can lead to sediment load reductions of up to 40%. The implementation of these measures did not adversely affect the water yield. In contrast, multi-diverse crop rotations including irrigated dry season crops were found to be disadvantageous in terms of water availability by significantly reducing streamflow during low flow periods. The study considers rainfall uncertainty by using a precipitation data ensemble, but nevertheless highlights the importance of well established monitoring systems due to related shortcomings in model calibration. Despite the existing uncertainties, the model results are useful for water resource managers to develop water and soil protection strategies for the Pipiripau River Basin and for watersheds with similar characteristics.

Piecewise Disassembly of a Large-Herbivore Community across a Rainfall Gradient: The UHURU Experiment

Large mammalian herbivores (LMH) strongly influence plant communities, and these effects can propagate indirectly throughout food webs. Most existing large-scale manipulations of LMH presence/absence consist of a single exclusion treatment, and few are replicated across environmental gradients. Thus, important questions remain about the functional roles of different LMH, and how these roles depend on abiotic context. In September 2008, we constructed a series of 1-ha herbivore-exclusion plots across a 20-km rainfall gradient in central Kenya. Dubbed "UHURU" (Ungulate Herbivory Under Rainfall Uncertainty), this experiment aims to illuminate the ecological effects of three size classes of LMH, and how rainfall regimes shape the direction and magnitude of these effects. UHURU consists of four treatments: total-exclusion (all ungulate herbivores), mesoherbivore-exclusion (LMH >120-cm tall), megaherbivore-exclusion (elephants and giraffes), and unfenced open plots. Each treatment is replicated three times at three locations (“sites”) along the rainfall gradient: low (440 mm/year), intermediate (580 mm/year), and high (640 mm/year). There was limited variation across sites in soil attributes and LMH activity levels. Understory-plant cover was greater in plots without mesoherbivores, but did not respond strongly to the exclusion of megaherbivores, or to the additional exclusion of dik-dik and warthog. Eleven of the thirteen understory plant species that responded significantly to exclusion treatment were more common in exclusion plots than open ones. Significant interactions between site and treatment on plant communities, although uncommon, suggested that differences between treatments may be greater at sites with lower rainfall. Browsers reduced densities of several common overstory species, along with growth rates of the three dominant Acacia species. Small-mammal densities were 2–3 times greater in total-exclusion than in open plots at all sites. Although we expect patterns to become clearer with time, results from 2008–2012 show that the effects of excluding successively smaller-bodied subsets of the LMH community are generally non-additive for a given response variable, and inconsistent across response variables, indicating that the different LMH size classes are not functionally redundant. Several response variables showed significant treatment-by-site interactions, suggesting that the nature of plant-herbivore interactions can vary across restricted spatial scales.

Climate Variability and Durum Wheat Adaptation Using the AquaCrop Model in Southern Sardinia

Durum wheat is the most important C3 rainfed crop in southern Sardinia, Italy. Climate variability and socio- economic factors are the main sources of uncertainty and concern for farmers in this region that has experienced severe drought conditions and problems of competing water demands during the last decades. Within the framework of a European Union project, CLIMB (Climate Induced Changes in the Mediterranean Region), durum wheat production was simulated under future climate change conditions to evaluate the effects of increased maximum temperature and high rainfall variability on crop yield, and the adaptability of the crop response. The first stage of this work was devoted to properly calibrating the AquaCrop model and testing its predictive performance, prior to applying it under future climate change scenarios. Durum wheat (CV Creso) yield and daily meteorological data collected and recorded in the Agris experimental fields during the period 1995–2012 were systematically checked and analyzed before their use for calibration of the AquaCrop model. After calibration the model showed a generally good performance with a significant correlation between observed and simulated yield for durum wheat during the considered period, including the year 1995 that was characterized by a severe drought that led to water stress conditions. A future scenario of climate change was then used as input to the AquaCrop model to predict wheat yield response and to investigate the control of water availability on rainfed crop production for the period 1951–2100. The simulated future scenarios show potential improved productivity arising from the increased CO2 concentration. This encouraging result is however tempered by increased uncertainty and fluctuations in rainfall during the fall and early winter periods (September–December). The possible tradeoffs between these factors, as well as the expected negative effects of increased maximum temperatures, are being further examined.

Understanding the Dependence of Satellite Rainfall Uncertainty on Topography and Climate for Hydrologic Model Simulation

A quantitative and physical understanding of satellite rainfall uncertainties provides meaningful guidance on improving algorithms to advance hydrologic prediction. The aim of this study is to characterize satellite rainfall errors and their impact on hydrologic fluxes based on fundamental governing factors that dictate the accuracy of passive remote sensing of precipitation. These governing factors are land features-comprising topography (elevation)-and climate type, representing the average ambient atmospheric conditions. First, the study examines satellite rainfall errors of three major products, 3B42RT, Climate prediction center MORHing technique (CMORPH), and Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN), by breaking the errors down into independent components (hit, miss-rain, and false-rain biases) and investigating their contribution to runoff and soil moisture errors. The uncertainties of three satellite rainfall products are explored for five regions of the Mississippi River basin that are categorized grid cell by grid cell (at the native spatial resolution of satellite products) based on topography and climate. It is found that total and hit biases dictate the temporal trend of soil moisture and runoff errors, respectively. Miss-rain and hit biases are the leading errors in the 3B42RT and CMORPH products, respectively, whereas false-rain bias is a pervasive problem of the PERSIANN product. For 3B42RT and CMORPH, about 50%-60% of grid cells are influenced by the total bias during winter and 60%-70% of grid cells during summer. For PERSIANN, about 70%-80% of the grid cells are marked by total bias during the summer and winter seasons. False-rain bias gradually increases from lowland to highland regions universally for all three satellite rainfall products. Overall, the study reveals that satellite rainfall uncertainty is dependent more on topography than the climate of the region. This study's results indicate that it is now worthwhile to assimilate the static knowledge of topography in the satellite estimation of precipitation to minimize the uncertainty in anticipation of the Global Precipitation Measurement mission.

Quantifying the uncertainty in the Soil Conservation Service flood hydrographs: a case study in the Azores Islands

AbstractFlash flooding is characterised by a rapid flooding phenomenon caused by intense rainfall. Despite being an extreme event with high uncertainty, the rainfall‐run‐off process is often regarded as deterministic (rather than stochastic). In this paper, the Soil Conservation Service (SCS) flood hydrograph uncertainty is quantified based on the Total Error Framework (TEF), and introduced into the model by applying perturbation in the input data and model parameters. The random perturbation component is stochastically modelled. A sensitivity analysis was carried out on the stochastic model parameters, using a real case study in the Azores (Portugal). The results showed that the flood hydrograph uncertainty varies over time, with its largest deviations occurring at the beginning of the flooding because of the uncertainty associated with the SCS method curve number parameter (correlation coefficient R2 of 0.86). Rainfall uncertainty was responsible for the uncertainty in the hydrograph peaks' magnitude (R2 = 0.93) while uncertainty in the propagation velocity was responsible for the uncertainty in the peaks' time (R2 = 0.97).

Flood analysis of urban drainage systems: Probabilistic dependence structure of rainfall characteristics and fuzzy model parameters

Flood analysis of urban drainage systems plays a crucial role for flood risk management in urban areas. Rainfall characteristics, including the dependence between rainfall variables, have a significant influence on flood frequency. This paper considers the use of copulas to represent the probabilistic dependence structure between rainfall depth and duration in the synthetic rainfall generation process, and the Gumbel copula is fitted for the rainfall data in a case study of sewer networks. The probabilistic representation of rainfall uncertainty is combined with fuzzy representation of model parameters in a unified framework based on Dempster–Shafer theory of evidence. The Monte Carlo simulation method is used for uncertainty propagation to calculate the exceedance probabilities of flood quantities (depth and volume) of the case study sewer network. This study demonstrates the suitability of the Gumbel copula in simulating the dependence of rainfall depth and duration, and also shows that the unified framework can effectively integrate the copula-based probabilistic representation of random variables and fuzzy representation of model parameters for flood analysis.

Positive effect of climate change on cotton in 2050 by CO2 enrichment and conservation agriculture in Cameroon

This article predicts an unexpected positive effect of climate change on cotton production in Cameroon. Global warming could threaten cotton production in Africa due to increasing temperature and CO2, and rainfall uncertainties. This situation is worsened by the fact that most African farmers grow cotton as their cash crop and have few or no possible alternatives. Assessing the impact of climate change on cotton production is therefore critical. Here, we used CROPGRO, a process-based crop model that can simulate the main features of cotton growth and management. We applied this model to two regions in North Cameroon and a set of six regional climate projections combining general climate models and regional climate models from the ENSEMBLES project. The model was calibrated and validated with a data set of observations made in farmer fields from 2001 to 2005 and at an experimental station in 2010. Our results show unexpectedly that climate change from 2005 to 2050 in North Cameroon will have a positive effect on cotton yields with an increase of 1.3 kg ha−1 year−1 in yield, especially if conservation agriculture systems are adopted. The predicted increase of 0.05 °C year−1 in temperature will shorten crop cycles by 0.1 day year−1 with no negative effect on yields. Moreover, the fertilizing effect of CO2 enrichment will increase yields by approximately 30 kg ha−1. The rainfall pattern is likely to change, but the six regional models used to generate future weather patterns did not predict a decrease in rainfall. One model even forecast an increase in rainfall amounts. According to our findings, climate changes in North Cameroon can be anticipated by tailoring alternative cropping systems and adaptation techniques to cope with climate change.

Rainfall variability and food crop portfolio choice: evidence from Ethiopia

This paper concerns the patterns of food crop choice in a multicropping setting, in which production risk considerations and rainfall uncertainty are likely to be critical factors. The analysis employed plot level panel data from Ethiopia, combined with 30 year meteorological data corresponding to the survey villages used to construct seasonal and yearly rainfall variability. Using the single index approach, the riskiness of crop portfolio was constructed at a household level, taking into account the multicropping nature of the farming system. The combined riskiness of crops grown at a household level responded negatively to annual rainfall variability, with seasonal rainfall variability having a less consistent impact. Farmers are, therefore, more likely to select less risky crop portfolios even when intercrop interactions are taken into account.

A formal statistical approach to representing uncertainty in rainfall–runoff modelling with focus on residual analysis and probabilistic output evaluation – Distinguishing simulation and prediction

Summary While there seems to be consensus that hydrological model outputs should be accompanied with an uncertainty estimate the appropriate method for uncertainty estimation is not agreed upon and a debate is ongoing between advocators of formal statistical methods who consider errors as stochastic and GLUE advocators who consider errors as epistemic, arguing that the basis of formal statistical approaches that requires the residuals to be stationary and conform to a statistical distribution is unrealistic. In this paper we take a formal frequentist approach to parameter estimation and uncertainty evaluation of the modelled output, and we attach particular importance to inspecting the residuals of the model outputs and improving the model uncertainty description. We also introduce the probabilistic performance measures sharpness, reliability and interval skill score for model comparison and for checking the reliability of the confidence bounds. Using point rainfall and evaporation data as input and flow measurements from a sewer system for model conditioning, a state space model is formulated that accounts for three different flow contributions: wastewater from households, and fast rainfall–runoff from paved areas and slow rainfall-dependent infiltration–inflow from unknown sources. We consider two different approaches to evaluate the model output uncertainty, the output error method that lumps all uncertainty into the observation noise term, and a method based on Stochastic Differential Equations (SDEs) that separates input and model structure uncertainty from observation uncertainty and allows updating of model states in real-time. The results show that the optimal simulation (off-line) model is based on the output error method whereas the optimal prediction (on-line) model is based on the SDE method and the skill scoring criterion proved that significant predictive improvements of the output can be gained from updating the states continuously. In an effort to attain residual stationarity for both the output error method and the SDE method transformation of the observations were necessary but the statistical assumptions were nevertheless not 100% justified. The residual analysis showed that significant autocorrelation was present for all simulation models. We believe users of formal approaches to uncertainty evaluation within hydrology and within environmental modelling in general can benefit significantly from adopting the evaluation measures applied here, so the probabilistic performance of their models can be assessed properly.

Uncertainty of areal average rainfall and its effect on runoff simulation: A case study for the Chungju Dam Basin, Korea

This study investigated the relationship between the errors involved in the areal average rainfall estimation and the runoff simulation (hereafter, rainfall and runoff errors). The error statistics of the observed areal average rainfall were estimated and used for the generation of input data for the runoff simulation. The Clark instantaneous unit hydrograph was used for this runoff simulation. The runoff model parameters were estimated using several sets of rainfall-runoff data observed, whose statistics were then used for the sensitive analysis of the model simulation on the parameter sets. The rainfall error in this study was defined as the relative difference between the observed and generated areal average rainfall. On the other hand, the runoff error was designed to consider the runoff volume, peak flow and peak time, respectively. This study was applied to the Chugnju Dam Basin, Korea. The results obtained are as follows: (1) The variation of model parameters estimated in this study were about 30% of their means. Also, the effect of this error in model parameter estimation on runoff simulation was found to be maximum 15% of the peak flow. (2) The estimation error of areal average rainfall was found roughly proportional to the areal average rainfall itself. Both the rainfall and runoff errors were found to have no obvious biases. However, the variance of the peak flow error was found to be significantly higher. (3) The relationship between rainfall error and runoff volume error was roughly one to one, however, the rainfall error has become amplified by more than 50 % and transferred to the peak flow error.

Uncertainties in daily rainfall over Africa: assessment of gridded observation products and evaluation of a regional climate model simulation

ABSTRACTWe intercompare three gridded observed daily rainfall datasets over Africa (FEWS (Famine Early Warning System), GPCP (Global Precipitation Climatology Project) and TRMM (Tropical Rainfall Measuring Mission)) in order to assess uncertainties in observation products towards the evaluation of the performance of a Regional Climate Model (RegCM3) in simulating daily precipitation characteristics over a domain encompassing the whole African continent. We find that different observation products exhibit substantial systematic differences in mean rainfall, but especially in higher order daily precipitation statistics, such as frequency of wet days, precipitation intensity and extremes as well as maximum length of wet and dry spells. For example, FEWS shows mostly higher frequency and lower intensity events than TRMM and GPCP. Thus, the different datasets provide quite different representations of daily precipitation behavior. As a result, although RegCM3 captures pretty well the monsoon rainband evolution and exhibits a representation of daily precipitation statistics within the range of the observations, it performs differently with respect to the various products. For instance, it simulates more intense but less frequent events over East and Southern Africa than in FEWS and vice versa compared to TRMM. We thus highlight the uncertainty in observations as a key factor preventing a rigorous and unambiguous evaluation of climate models over Africa. Improving the quality and consistency of observation products is thus paramount for a better understanding of the response of African climate to global warming. Copyright © 2012 Royal Meteorological Society

Decision strategies for handling the uncertainty of future extreme rainfall under the influence of climate change

Several extraordinary rainfall events have occurred in Denmark within the last few years. For each event, problems in urban areas occurred as the capacity of the existing drainage systems were exceeded. Adaptation to climate change is necessary but also very challenging as urban drainage systems are characterized by long technical lifetimes and high, unrecoverable construction costs. One of the most important barriers for the initiation and implementation of the adaptation strategies is therefore the uncertainty when predicting the magnitude of the extreme rainfall in the future. This challenge is explored through the application and discussion of three different theoretical decision support strategies: the precautionary principle, the minimax strategy and Bayesian decision support. The reviewed decision support strategies all proved valuable for addressing the identified uncertainties, at best applied together as they all yield information that improved decision making and thus enabled more robust decisions.

Postcode forecasting

Postcode forecasting

Benchmarking observational uncertainties for hydrology: rainfall, river discharge and water quality

AbstractThis review and commentary sets out the need for authoritative and concise information on the expected error distributions and magnitudes in observational data. We discuss the necessary components of a benchmark of dominant data uncertainties and the recent developments in hydrology which increase the need for such guidance. We initiate the creation of a catalogue of accessible information on characteristics of data uncertainty for the key hydrological variables of rainfall, river discharge and water quality (suspended solids, phosphorus and nitrogen). This includes demonstration of how uncertainties can be quantified, summarizing current knowledge and the standard quantitative results available. In particular, synthesis of results from multiple studies allows conclusions to be drawn on factors which control the magnitude of data uncertainty and hence improves provision of prior guidance on those uncertainties. Rainfall uncertainties were found to be driven by spatial scale, whereas river discharge uncertainty was dominated by flow condition and gauging method. Water quality variables presented a more complex picture with many component errors. For all variables, it was easy to find examples where relative error magnitudes exceeded 40%. We consider how data uncertainties impact on the interpretation of catchment dynamics, model regionalization and model evaluation. In closing the review, we make recommendations for future research priorities in quantifying data uncertainty and highlight the need for an improved ‘culture of engagement’ with observational uncertainties. Copyright © 2012 John Wiley & Sons, Ltd.

Toward a reliable decomposition of predictive uncertainty in hydrological modeling: Characterizing rainfall errors using conditional simulation

This study explores the decomposition of predictive uncertainty in hydrological modeling into its contributing sources. This is pursued by developing data‐based probability models describing uncertainties in rainfall and runoff data and incorporating them into the Bayesian total error analysis methodology (BATEA). A case study based on the Yzeron catchment (France) and the conceptual rainfall‐runoff model GR4J is presented. It exploits a calibration period where dense rain gauge data are available to characterize the uncertainty in the catchment average rainfall using geostatistical conditional simulation. The inclusion of information about rainfall and runoff data uncertainties overcomes ill‐posedness problems and enables simultaneous estimation of forcing and structural errors as part of the Bayesian inference. This yields more reliable predictions than approaches that ignore or lump different sources of uncertainty in a simplistic way (e.g., standard least squares). It is shown that independently derived data quality estimates are needed to decompose the total uncertainty in the runoff predictions into the individual contributions of rainfall, runoff, and structural errors. In this case study, the total predictive uncertainty appears dominated by structural errors. Although further research is needed to interpret and verify this decomposition, it can provide strategic guidance for investments in environmental data collection and/or modeling improvement. More generally, this study demonstrates the power of the Bayesian paradigm to improve the reliability of environmental modeling using independent estimates of sampling and instrumental data uncertainties.

Investigating the similarity of satellite rainfall error metrics as a function of Köppen climate classification

This study addressed the question: How much similarity exists in uncertainty of space-borne precipitation products for similar Köppen climate zones in different and distant landmasses? Various metrics of satellite rainfall uncertainty were identified using a six year (2002–2007) archive of NASA's TRMM Multi-satellite Precipitation Analysis (TMPA) data product called 3B42V6 for two large distant landmasses that share many similar Köppen climate zones: 1) United States and 2) Australia. The level of quantitative similarity in error metrics for the same Köppen climate zones was then investigated. It was found that the bias and root mean squared error exhibited very close levels of similarity for similar Köppen climate zones in the US and Australia. However, similar inferences could not be drawn for other (higher-ordered) error metrics such as Probability of Detection (POD). The contrasting nature of the ground validation (GV) data (i.e., NEXRAD-radar in US and point gauge in Australia) for characterizing uncertainty may be one of the reasons for this observed lack of similarity. Using a dense gauge network of 42 gauges over a standard 3B42V6 grid box (~0.25°) as a ground validation benchmark, the dependence of uncertainty as a function of gauge density was quantified. These relationships were then cast in the context of our Köppen climate similarity experiment to identify the minimum level of gauge density that would be needed to resolve more accurately the actual level of similarity of error metrics for distant landmasses.

Modelling catchment inflows into Lake Victoria: uncertainties in rainfall–runoff modelling for the Nzoia River

Climate and soil characteristics vary considerably around the Lake Victoria basin resulting in high spatial and temporal variability in catchment inflows. However, data for estimating the inflows are usually sparsely distributed and error-prone. Therefore, modelled estimates of the flows are highly uncertain, which could explain early difficulties in reproducing the lake water balance. The aim of this study was to improve the estimates of catchment flow to Lake Victoria. The WASMOD model was applied to the Nzoia River, one of the major tributaries to Lake Victoria. Uncertainty was assessed within the GLUE framework. During calibration, log-transformation was performed on both simulated and observed flows. The results showed that, despite its simple structure, WASMOD produces acceptable results for the basin. For a Nash-Sutcliffe efficiency (NS) threshold of 0.6, the percentage of observations bracketed by simulations (POBS) was 74%, the average relative interval length (ARIL) was 0.93, and the maximum NS value was 0.865. The residuals were shown to be homoscedastic, normally distributed and nearly independent. When the NS threshold was increased to 0.8, POBS decreased to 54% with an improvement of ARIL to 0.49, highlighting the effect of the subjective choice of likelihood threshold. Citation Kizza, M., Rodhe, A., Xu, C.-Y. & Ntale, H. K. (2011) Modelling catchment inflows into Lake Victoria: uncertainties in rainfall–runoff modelling for the Nzoia River. Hydrol. Sci. J. 56(7), 1210–1226.