new approaches to hydrological prediction in data …hydrologie.org/redbooks/a333/p333 description,...

New Approaches to Hydrological Prediction in Data-sparse RegionsEdited by Koray K. Yilmaz, Ismail Yucel, Hoshin V. Gupta, Thorsten Wagener,

Dawen Yang, Hubert Savenije, Christopher Neale, Harald Kunstmann & John Pomeroy IAHS Publ. 333 (2009) ISBN 978-1-907161-04-9, 344 + x pp. Price £66.00

When data are scarce, hydrological predictions become unreliable mainly due both to the inability to specify model components and parameter values that consistently represent the dominant hydrological processes in a particular watershed, and due to the lack of high quality model forcing. This is a problem in developed and developing countries, and the focus of much research effort worldwide. This proceedings volume, from a symposium of the same name held in Hyderabad, India, September 2009, contains 40 papers from over 20 countries. They reflect differing aspects of, and approaches to, the problem and are grouped accordingly: Hydrological modelling in Poorly gauged and Ungauged basins Hydrometeorology and Climate Change Assessment Remote Sensing Applications in Hydrology Characterizing Rainfall Variability and its Impacts on Hydrological Modelling

Contents

Preface by Koray K. Yilmaz, Ismail Yucel, Hoshin V. Gupta, Thorsten Wagener, Dawen Yang, Hubert Savenije, Christopher Neale, Harald Kunstmann & John Pomeroy

v

1 Hydrological Modelling in Poorly Gauged and Ungauged Basins

Assessing parameters of physically-based models for poorly gauged basins Lev Kuchment & Alexander Gelfan

3

A universal approach to runoff processes modelling: coping with hydrological predictions in data-scarce regions O. M. Semenova & T. A. Vinogradova

11

Regularized calibration of a distributed hydrological model using available information about watershed properties and signature measures Prafulla Pokhrel & Hoshin V. Gupta

20

Landscape dependent derivation of J2000 model parameters for hydrological modelling in ungauged basins Markus Wolf, Björn Pfennig, Peter Krause & Wolfgang-Albert Flügel

26

Development of an extended spatially distributed routing scheme and its impact on process oriented hydrological modelling results Björn Pfennig, Holm Kipka, Markus Wolf, Manfred Fink, Peter Krause & Wolfgang-Albert Flügel

37

Development of a hydrological response index to represent TOPMODEL parameters Zhang Youjing, Chen Bo & He Chuan

44

Calibration of a semi-distributed hydrological model using discharge and remote sensing data Lal P. Muthuwatta, Martijn J. Booij, Tom H. M. Rientjes, M. G. Bos, A. S. M. Gieske & Mobin-ud-Din Ahmad

52

Modelling the long-term impact of climate change on rainfall–runoff processes over a large Sudano-Sahelian catchment Denis Ruelland, Vincent Guinot, Florent Levavasseur & Bernard Cappelaere

59

A probability distribution function approach to modelling rainfall–runoff response for data-sparse catchments Binquan Li & Zhongmin Liang

69

Hydrograph transposition between basins through a geomorphology-based deconvolution–reconvolution approach Houda Boudhraâ, Christophe Cudennec, Mohamed Slimani & Herve Andrieu

76

Towards the development of a consistent uncertainty framework for hydrological predictions in South Africa Evison Kapangaziwiri, Denis Hughes & Thorsten Wagener

84

Can discharge assimilation methods be used to improve flood forecasting when few data are available? Lionel Berthet, Maria-Helena Ramos, Charles Perrin, Vazken Andréassian & Cécile Loumagne

94

Modelling hydrological time series data using wavelet neural network analysis Y. R. Satyaji Rao & B. Krishna

101

An evaluation of multi-basin hydrological modelling for predictions in ungauged basins Chantal Donnelly, Joel Dahne, Göran Lindström, Jörgen Rosberg, Johan Strömqvist, Charlotta Pers, Wei Yang & Berit Arheimer

112

Using recently developed global data sets for hydrological predictions Johan Strömqvist, 121

Joel Dahné, Chantal Donnelly, Göran Lindström, Jörgen Rosberg, Charlotta Pers, Wei Yang & Berit Arheimer

Modélisation hydrologique et interrelations Climat–Homme–Environnement dans le Sahel Burkinabè (Hydrological modelling and climate–man–environment interrelationships in the Burkinabe Sahel) Jean-Emmanuel Paturel, Pierre Diello, Gil Mahé, Alain Dezetter, Hamma Yacouba, Bruno Barbier, Harouna Karambiri

128

Kriging method for estimation of groundwater resources in a basin with scarce monitoring data Chengpeng Lu, Longcang Shu, Xunhong Chen, Yuezan Tao & Ying Zhang

136

An experimental and modelling investigation of macropore dominated subsurface stormflow in vegetated hillslopes of northeast India Rupak Sarkar & Subashisa Dutta

145

2 Hydrometeorology and Climate Change Assessment

Coupled regional modelling of atmospheric–hydrologic processes for reconstruction of hy-dro-climate data and climate change assessment M. L. Kavvas, Z. Q.Chen, N. Ohara, M. L. Anderson, A. J. Shaaban & M. Z. M. Amin

155

Hydro-meteorological predictions from GCM simulations: downscaling techniques and uncertainty modelling Pradeep Mujumdar, Subimal Ghosh & Deepashree Raje

165

Coupling VIC with GCM models to predict climate change impact in the Hanjiang basin, China Shenglian Guo, Jing Guo, Jun Zhang & Hua Chen

176

Runoff modelling within the Canadian Regional Climate Model (CRCM): analysis over the Quebec/Labrador watersheds Biljana Music, Anne Frigon, Michel Slivitzky, André Musy, Daniel Caya & René Roy

183

Development of a coupled land-surface and hydrology model system for mesoscale hydro-meteorological simulations Fei Yuan, Harald Kunstmann, Chuanguo Yang, Zhongbo Yu, Liliang Ren, Benjamin Fersch & Zhenghui Xie

195

A study coupling a large-scale hydrological model with a regional climate model Bin Yong, Liliang Ren, Lihua Xiong, Xiaoli Yang, Wanchang Zhang, Xi Chen & Shanhu Jiang

203

Large-scale water balance estimations through regional atmospheric moisture flux model-ling and comparison to GRACE signals Benjamin Fersch, Harald Kunstmann, Nico Sneeuw & Balaji Devaraju

211

Assessment of flood events in the data-sparse Brahmaputra Basin in northeast India U. C. Sharma & Vikas Sharma

220

3 Remote Sensing Applications in Hydrology

Evaluation of a satellite-based near real-time global flood prediction system Koray K. Yilmaz, Robert F. Adler, Yang Hong & Harold F. Pierce

229

Estimating precipitation for poorly-gauged areas in western China Junliang Jin, Guihua Lu & Zhiyong Wu

238

TRMM rainfall data estimation over the Peruvian Amazon-Andes basin and its assimila-tion into a monthly water balance model Waldo Sven Lavado Casimiro, David Labat, Jean Loup Guyot, Josyane Ronchail & Juan Julio Ordoñez

245

Prospect for TRMM in rainfall estimates: a case study in the Laohahe basin, China 253

Hong Wang, Liliang Ren & Xiaoli Yang

Evaluation of the Hydro-Estimator satellite rainfall algorithm and its utility in hydrological prediction in a mountainous region Ismail Yucel, Robert J. Kuligowski & David Gochis

259

Spatially distributed evapotranspiration estimation using remote sensing and ground-based radiometers over cotton at Maricopa, Arizona, USA Andrew N. French, Douglas Hunsaker, Kelly Thorp & Thomas Clarke

267

Applicable algorithm to map daily evapotranspiration using MODIS images for the Laohahe River basin, northeastern China Xiuqin Fang, Liliang Ren, Qiongfang Li, Qiuan Zhu, Xiaofan Liu & Yonghua Zhu

273

4 Characterizing Rainfall Variability and Its Impacts on Hydrological Modelling

Inverting the hydrological cycle: when streamflow measurements help assess altitudinal pre-cipitation gradients in mountain areas Audrey Valery, Vazken Andréassian & Charles Perrin

281

Rainfall variability and uncertainty in water resource assessments in South Africa Tendai Sawunyama & Denis Hughes

287

Robust and flexible hydroinformatics to account for rainfall space–time variability in a data-sparse region Sameh Chargui, Christophe Cudennec, Mohamed Slimani, Jean-Christophe Pouget & Jalel Aouissi

295

Drought forecast using an artificial neural network for three hydrological zones in San Francisco River basin, Brazil Celso Augusto G. Santos, Bruno S. Morais & Gustavo B. L. Silva

302

Regional frequency analysis of annual precipitation in data-sparse regions using large-scale atmospheric variables P. Satyanarayana & V. V. Srinivas

313

Impacts of rainfall uncertainty on water resource planning models in the Upper Limpopo basin, Botswana P. K. Kenabatho, N. R. McIntyre & H. S. Wheater

320

A high-resolution hierarchical space–time framework for single storm events and its ap-plication for short-term rainfall forecasting Juan Qin, Michael Leonard, George Kuczera, Mark Thyer, Andrew Metcalfe & Martin Lambert

330

Key word index 341

New Approaches to Hydrological Prediction in Data-sparse Regions (Proc. of Symposium HS.2 at the Joint IAHS & IAH Convention, Hyderabad, India, September 2009). IAHS Publ. 333, 2009, 3-10.

Assessing parameters of physically-based models for poorly gauged basins

LEV KUCHMENT & ALEXANDER GELFANWater Problems Institute of Russian Academy of Sciences, 119991 Gubkin 3, Moscow, [email protected]

Abstract The possibility of using a priori information to reduce the amount of hydrological observation series data needed for the calibration of physically-based models of runoff generation has been studied. It is shown that by using measurements and runoff generation models in proxy-basins, the number of parameters requiring calibration can be limited to two toor three. Investigations were carried out using a physically-based model of runoff generation in the Kolyma and Seim river basins, Russia. The possibility of using observations from water-balance stations and experimental catchments as a priori data for assigning parameters of the models is demonstrated.Key words ungauged basin; physically-based model; parameters; proxy-basin


A universal approach to runoff processes modelling: coping with hydrological predictions in data-scarce regions

O. M. SEMENOVA1 & T. A. VINOGRADOVA2

1Department of Experimental Hydrology and Mathematical Modelling of Hydrological Processes, State Hydrological Institute, 23 2-ya liniya VO, 199053 St Petersburg, Russia [email protected]

2Department of Geography and Geoecology, St Petersburg State University, 31/33 10-ya liniya VO, 199178 St Petersburg, Russia

Abstract This paper discusses the features which a hydrological model should possess to be successfully applied in the task of hydrological predictions in poorly gauged regions. The Deterministic Modelling Hydrological System developed on the basis of the principle of universality is described as an example of such a model. The results of the simulations conducted across the data scarce basins of eastern Siberia are presented.Key words principle of universality; Deterministic Modelling Hydrological System, DMHS; data scarce regions; eastern Siberia


Regularized calibration of a distributed hydrological model using available information about watershed properties and signature measures

PRAFULLA POKHREL & HOSHIN V. GUPTADepartment of Hydrology & Water Resources, The University of Arizona, Tucson, Arizona 85721, USA

mailto:[email protected]

[email protected]

Abstract Physically-based distributed models are increasingly being used to predict the behaviour of hydrological processes in data-sparse regions. However, a model is a simplified representation of the real system and some form of calibration cannot be avoided. Because distributed models have large numbers of parameters to be specified, the resulting parameter estimation problem becomes ill conditioned. In this study we investigate a calibration approach that uses: (a) a simple form of spatial regularization (using scalar multipliers) to reduce the dimension of the calibration problem, and (b) signature measures targeting specific behavioural response of a watershed system to guide the parameter search towards a more “hydrologically consistent” set of parameters. Signature measures are applied as “regularization constraints”, in an approach that is functionally similar to “Tikhonov regularization”, and which results in a better-conditioned optimization problem compared to the benchmark case. The approach is demonstrated for the Blue River Basin in Oklahoma, USA.Key words regularization; multicriteria optimization; parameter estimation; distributed hydrological model


Landscape dependent derivation of J2000 model parameters for hydrological modelling in ungauged basins

MARKUS WOLF, BJÖRN PFENNIG, PETER KRAUSE & WOLFGANG-ALBERT FLÜGELDepartment of Geoinformatic, Geohydrology and Modelling, Friedrich-Schiller-University Jena, Löbdergraben 32, D-07737 Jena, Germany [email protected]

Abstract The HRU (Hydrological Response Units) regionalisation concept is realised with a GIS-based intersection of landscape parameters such as topography, soils, geology and land use. In many catchments of the world the required data are only available on a coarse spatial resolution and there is often a lack of discharge and precipitation data. But there is a demand to involve these catchments in planning of water management. The assumption of a process-driven feedback between the topography and further landscape components, as well as runoff dynamics, leads to a modified delineation of process entities by a topographic oriented HRU approach on the basis of SRTM elevation data. The approach is based on the expectation that the water balance of ungauged basins can be estimated using SRTM-based delineations of process-oriented model entities to get a suitable prediction of runoff dynamics with disposable landscape components in spite of an insufficient hydro-meteorological database. Key words SRTM; hydrological modelling; prediction in ungauged basins; hydrological response unit; Germany; South Africa


Development of an extended spatially distributed routing scheme and its impact on process oriented hydrological modelling results

BJÖRN PFENNIG, HOLM KIPKA, MARKUS WOLF, MANFRED FINK, PETER KRAUSE & WOLFGANG-ALBERT FLÜGELDepartment of Geoinformatic, Geohydrology and Modelling, Friedrich-Schiller-University Jena, Löbdergraben 32, D-07737 Jena, [email protected]

Abstract Fully spatially distributed hydrological modelling requires a topological linkage of single modelling entities (e.g. hydrological response units – HRU) in order to reproduce relevant attenuation and



translation processes within the stream, but also during the transport of water in the form of lateral surface or subsurface flow. Most often such linkage is considered by a one-dimensional (1-D) approach which links one modelling entity to only one receiver that follows the flow direction. The comparison with actual lateral water movement in catchments show that such a 1-D routing scheme is often too simple, which can lead to an overestimation of the runoff concentration along the 1-D flow paths. On the other hand, an underestimation of runoff in flow cascades that do not reside next to the main 1D flow paths can occur as the affected HRUs do not receive realistic inflow from their source entities above. As a catchment-wide consequence the 1-D routing scheme can result in a significant over- or underestimation of the contributing area for specific parts of a catchment, which can have important implications for the spatial distribution of accompanying processes such as spatial variation of soil moisture, soil erosion or solute transport. To address the problems outlined above, a new approach has been developed that allows a multi-dimensional linkage of model entities in such a way that each entity can have various receivers to which the water is passed. This extended routing scheme was implemented in the hydrological modelling system J2000 (Krause, 2001) and was used for the simulation of the hydrological processes of a number of meso-scaled catchments in Thuringia, Germany. The paper presents the most important details of the extended routing scheme, the simulation results along with the comparison of those obtained with the 1-D linkage and highlights the impacts on the hydrological process dynamics as well as on the HRU-based mass transport and balancing.Key words hydrological modelling; routing scheme; multi-dimensional linkage; HRU; model entities


Development of a hydrological response index to represent TOPMODEL parameters

ZHANG YOUJING1,2, CHEN BO2 & HE CHUAN2

1 State Key Laboratory of Hydrology-Water Resource and Hydraulic Engineering, Nanjing, China2 Department of Geographical Information Science, Hohai University, Nanjing 210098, China

[email protected]

Abstract This paper proposes an approach of constructing a hydrological response index (HRI) for description of influences of the catchment’s characteristics on hydrological simulations. HRI is composed of the topographic index of TOPMODEL and curve number (CN) of the SCS model, and can be computed from characteristics of topography and land surface components by using GIS and remote sensing techniques. Hydrological simulation is based on TOPMODEL and the parameters are calibrated. The relationship between HRI and the model’s key parameters was analysed on the basis of calibration results from 32 hydrological response units (HRU) that were derived from HRI. The study results show that correlation coefficients between HRI and the parameters m and lnT0 are 0.88 and 0.85, respectively. Furthermore, a good relationship between SR0 and NDVI by using Landsat data is also found. The validation of model parameters was carried out using those parameters. The validation results show that the correlation coefficient between observed and simulated stream discharges is 0.84. These results indicate that the proposed index can be used to represent the model parameters in the study region.Key words parameter regionalisation; hydrological response index; regression analysis; validation; Landsat data


Calibration of a semi-distributed hydrological model using discharge and remote sensing data

LAL P. MUTHUWATTA1, MARTIJN J. BOOIJ2, TOM H. M. RIENTJES3, M. G. BOS3, A. S. M. GIESKE3 & MOBIN-UD-DIN AHMAD4

1 International Water Management Institute, PO Box 2075, Colombo, Sri Lanka2 Water Engineering and Management, Faculty of Engineering Technology, University of Twente, PO Box 217,

7500 AE Enschede, The Netherlands

[email protected] International Institute for Geo-Information Science and Earth Observation (ITC), PO Box 6, 7500 AA Enschede,

The Netherlands4 CSIRO Land and Water, GPO Box 1666, Canberra ACT2601, Australia

Abstract The objective of this study is to present an approach to calibrate a semi-distributed hydrological model using observed streamflow data and actual evapotranspiration time series estimates based on remote sensing data. First, daily actual evapotranspiration is estimated using available MODIS satellite data, routinely collected meteorological data, and applying the SEBS algorithm. Second, the semi-distributed hydrological model HBV is calibrated and validated using the estimated evapotranspiration and observed discharge. This is done for multiple sub-basins of the Karkheh River basin in Iran. The Nash-Sutcliffe coefficient (NS) is calculated for each sub-basin. Maximum and minimum NS values for the calibration using observed discharge are 0.81 and 0.23, respectively, and using estimated evapotranspiration 0.61 and 0.46, respectively. The comparison of model simulations with multiple observed variables increases the probability of selecting a parameter set that represents the actual hydrological situation of the basin. The new calibration approach can be useful for further applications, especially in data-sparse river basins.Key words hydrological modelling; SEBS; remote sensing; MODIS; HBV model; actual evapotranspiration; Monte Carlo simulation; Karkheh River basin, Iran


Modelling the long-term impact of climate change on rainfall–runoff processes over a large Sudano-Sahelian catchment

DENIS RUELLAND1, VINCENT GUINOT2, FLORENT LEVAVASSEUR3 & BERNARD CAPPELAERE3

1CNRS, 2UM2, 3IRD – UMR HydroSciences Montpellier, Place E. Bataillon, F-34395 Montpellier Cedex 5, [email protected]

Abstract A conceptual hydrological model is implemented over the last 50 years on a large poorly gauged Sudano-Sahelian catchment. The purpose is to simulate the rainfall–runoff relationship over this period that has been subjected to important hydro-climatic changes. A calibration/validation exercise is performed via a lumped and semi-distributed model with a 10-day time step. Simulations based on fixed and variable parameters are compared through a multi-criteria analysis using a variety of goodness-of-fit indices. In the variable parameter option, the subsurface runoff and deep infiltration parameters are constrained by the spatio-temporal variability of a smoothed pluviometric index. The simulation results show that model efficiency is significantly improved by the integration of variable parameters. They also show a decrease of the simulated subsurface runoff and deep infiltration during the 1970s and the 1980s, which can be essentially attributed to the persistent rain deficit since the 1970s. This trend will have led to a drastic decrease of the deep water recharge and of base runoff contribution to flood composition.Key words climate change; catchment behaviour; hydrological modelling; HydroStrahler; River Bani, West Africa


A probability distribution function approach to modelling rainfall–runoff response for data-sparse catchments

BINQUAN LI & ZHONGMIN LIANGState Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing 210098, [email protected], [email protected]

Abstract A rainfall–runoff (RR) model considering the spatial variation of rainfall, soil infiltration capability and soil storage capacity over a catchment and based on probability distribution functions, is


used for rainfall–runoff modelling. The model combines infiltration excess (Horton) and saturation excess (Dunne) mechanisms. Moreover, it is applied to a data sparse catchment. Model parameters of the studied data sparse catchment are inferred from its parent gauged basin. In addition, a semi-distributed RR model called TOPMODEL is also employed in the parent gauged basin for comparison. Results show that the RR model can, to a certain extent, be applied to data sparse regions based upon hydrological similarity between the study catchment and its parent basin.Key words spatial variation; probability density functions; rainfall–runoff models; data sparse regions; Yellow River


Hydrograph transposition between basins through a geomorphology-based deconvolution–reconvolution approach

HOUDA BOUDHRAÂ1,2,3, CHRISTOPHE CUDENNEC3,4,5, MOHAMED SLIMANI2 & HERVE ANDRIEU6

1 CEMAGREF, UR HBAN, F-92163 Antony, [email protected] , [email protected]

2 INAT, Lab. STE, 1082 Tunis, Tunisia3 IRD, UMR G-EAU, 1004 Tunis, Tunisia4 Agrocampus Ouest, UMR1069, Soil Agro and hydroSystem, F-35000 Rennes, France5 INRA, UMR1069, Soil Agro and hydroSystem, F-35000 Rennes, France6 Division eau et environnement, LCPC, BP 4129, F-44341 Bouguenais Cedex, France

Abstract The aim of this study is to consider couples of basins and their respective non-calibrated robust geomorphology-based transfer functions. In the frame of discharge transposition, the two basins are respectively considered as the provider and the receiver. A discharge series of the provider basin is deconvoluted, through the inversion of its transfer function, to assess the net rainfall series. Assuming, as a first step, homogeneity between the two basins, the assessed net rainfall series is considered to be relevant for the receiver basin and convoluted with its own transfer function to simulate the discharge series at its outlet. Optimistically, the homogeneity between basins could be sufficient for nested, neighbouring and similar basins to make this approach promising when the receiver basin is ungauged. The approach is tested with simulated events for a set of four Tunisian basins (192, 180, 18.1 and 3.16 km 2). Transposition performs correctly in terms of the timing, volumes and shapes of hydrographs. Key words geomorphology-based transfer function; deconvolution; net rainfall; transposition; PUB


Towards the development of a consistent uncertainty framework for hydrological predictions in South Africa

EVISON KAPANGAZIWIRI1, DENIS HUGHES1 & THORSTEN WAGENER2

1 Institute for Water Research, Rhodes University, PO Box 94, Grahamstown 6140, South Africa [email protected], [email protected]

2 Department of Civil and Environmental Engineering, 226B Sackett Building, Pennsylvania State University, University Park, Pennsylvania 16802, USA

Abstract South Africa has a long history of using hydrological models to solve practical water resources management problems. Despite recent international advances in uncertainty analysis, uncertainty has yet to be explicitly included as part of standard modelling practice in the country. This paper reports on the initial development of a model independent framework for ensemble streamflow predictions in gauged and ungauged basins in data-poor regions. The proposed framework is largely based on existing methods and



data and includes a priori parameter estimation, a Monte Carlo framework and constraining model ensembles in ungauged basins through regional signatures of the catchment runoff response. Progress to date includes the modification of an existing a priori parameter estimation procedure that includes Monte Carlo sampling from probability distribution functions and the generation of model output ensembles. Two regional signatures have been developed, one based on the Budyko relationship and the other on the slope of the flow duration curve. A test application demonstrated that all the a priori ensembles produced behavioural flow duration curves, while only approximately 50% fell within the flow volume constraint. While the overall conclusion is that the framework is both theoretically sound as well as practical to implement, future work will focus on the development of additional regional catchment signatures and the use of the constrained ensembles in other water resources management tools, such as system yield models. Key words modelling; South Africa; uncertainty; regionalization; regional signatures; model evaluation; Budyko; parameter estimation


Can discharge assimilation methods be used to improve flood forecasting when few data are available?

LIONEL BERTHET1,2, MARIA-HELENA RAMOS1, CHARLES PERRIN1, VAZKEN ANDRÉASSIAN1 & CÉCILE LOUMAGNE1

1Cemagref, Hydrosystems and Bioprocesses Research Unit, BP 44, F-92163 Antony Cedex, France [email protected]

2AgroParisTech ENGREF, 19 Avenue du Maine, F-75732 Paris, France

Abstract Forecasting floods is a major issue for public safety all over the world. Due to the difficulties inherent in the flood forecasting exercise, data assimilation techniques have been developed to cope with model errors. Unfortunately, these techniques require recent (real or near real-time) observations which may not be readily available in regions lacking automatic measurements networks. This paper investigates the impact of data assimilation techniques on discharge forecasts and model performance when few (but not zero) discharge measurements are available for the data assimilation. A parsimonious rainfall–runoff model is applied to a set of 178 French catchments. We explore the time properties of different discharge data assimilation schemes. Life times of the updates and model performance are assessed as a function of the time between the last available discharge observation and the forecast. State updating proves to have an added value to the forecasting system, even when data availability is limited.Key words flood forecasting; data availability; rainfall–runoff modelling; data assimilation


Modelling hydrological time series data using wavelet neural network analysis

Y. R. SATYAJI RAO & B. KRISHNA Deltaic Regional centre, National Institute of Hydrology, Siddartha Nagar, Kakinada-3, Andhra Pradesh, [email protected]

Abstract Time series analysis requires mapping complex relationships between input(s) and output(s), because the forecasted values are mapped as a function of patterns observed in the past. In order to improve the precision of the forecasts, a Wavelet Neural Network (WNN) model, based on a combination of wavelet analysis and Artificial Neural Network (ANN), has been proposed. The WNN and ANN models have been applied to daily streamflow and monthly groundwater levels series where there is a scarcity of other hydrological time series data. The calibration and validation performance of the models is evaluated with appropriate statistical indices. The results of daily streamflow and monthly groundwater level series

modelling indicated that the performances of WNN models are more effective than the ANN models. This paper also highlights the capability of WNN models in estimating low and high values in the hydrological time series data. Key words time series; ANN; WNN; streamflow; groundwater levels; India


An evaluation of multi-basin hydrological modelling for predictions in ungauged basins

CHANTAL DONNELLY, JOEL DAHNE, GÖRAN LINDSTRÖM, JÖRGEN ROSBERG, JOHAN STRÖMQVIST, CHARLOTTA PERS, WEI YANG & BERIT ARHEIMERSwedish Meteorological and Hydrological Institute, SE-601 76 Norrköping, Sweden [email protected]

Abstract Availability of data is the limiting condition for reliable hydrological predictions in many regions of the world. One possible solution is to increase the scale of hydrological models so as to encompass data sparse regions within larger regions where data are available. A growing number of hydrological model studies use the spatial distribution of soils and vegetation to predict spatially-varying catchment behaviour. This introduces the potential to setup and use hydrological models over larger scales and simultaneously over several river basins. Such models can be set up easily, over very large regions, using freely available global data sets. However, the questions remain: (a) can a multi-basin hydrological model be calibrated using a uniform parameter set? and (b) can the uniformly calibrated, multi-basin hydrological model be used to make reasonable predictions in ungauged basins? Two different multi-basin regions were set up and calibrated using uniform land-use and soil-type dependent parameters in the HYPE hydrological model. Results indicate that a reasonable calibration can be achieved for a large multi-basin model set using readily available global input databases and using a homogenous parameter set.Key words distributed hydrological modelling; predictions in ungauged basins


Using recently developed global data sets for hydrological predictions

JOHAN STRÖMQVIST, JOEL DAHNE, CHANTAL DONNELLY, GÖRAN LINDSTRÖM, JÖRGEN ROSBERG, CHARLOTTA PERS, WEI YANG & BERIT ARHEIMER Swedish Meteorological and Hydrological Institute (SMHI), SE-601 76 Norrköping, [email protected]

Abstract The HYPE hydrological model was used for multi-basin applications with model input derived from global databases compiled using the World Hydrological Input Set-up Tool (WHIST). The model was applied to the La Plata Basin (3.2 million km2) in South America and to Europe (7 million km2). Water balance was modelled reasonably well, with volume errors at the gauging stations in Europe being generally <10%, whilst there were larger discrepancies in La Plata Basin. The median Nash-Sutcliffe model efficiency (NSE) was 0.27 for Europe and <0 for La Plata Basin. A simple sensitivity study shows that, for Northern Europe, the model results were most sensitive to meteorological forcing data and land cover. The results indicate that global databases can be useful for hydrological predictions in data sparse regions, although further studies are required to better distinguish between specific sources of errors and possibilities for improvements of both databases and models.Key words hydrological modelling; predictions in ungauged basins; sensitivity analysis; global databases



Modélisation hydrologique et interrelations Climat–Homme–Environnement dans le Sahel Burkinabè

JEAN-EMMANUEL PATUREL1, PIERRE DIELLO2, GIL MAHÉ3, ALAIN DEZETTER1, HAMMA YACOUBA2, BRUNO BARBIER4 & HAROUNA KARAMBIRI2 1 HSM/IRD, BP 2528, Bamako, [email protected]

2 2iE -UTER-GVEA, 01 BP 594, Ouagadougou 01, Burkina Faso3 HSM/IRD, USTL, Case MSE, Pl. E. Bataillon, F-34095 Montpellier Cedex 5, France4 CIRAD/2iE, Avenue du Président Kennedy, 01 BP 596, Ouagadougou 01, Burkina Faso

Résumé Dans le contexte sahélien, les études menées depuis une vingtaine d’années montrent que les effets conjoints du changement climatique et des activités humaines sur les états de surface sont à l’origine d’un “paradoxe hydrologique”. On observe, en effet, depuis les années 1970 que sur certains bassins versants de cette région les coefficients d’écoulement ont très fortement augmenté, entraînant parfois des écoulements plus importants qu’auparavant, en dépit d’une diminution marquée de la pluviométrie régionale. Ces modifications de la relation pluie-débit nécessitent de nouvelles approches permettant de prendre en compte conjointement la variabilité climatique et la dimension anthropique dans la modélisation hydrologique de ces bassins. Le travail mené a eu pour objectif d’intégrer dans un modèle hydrologique une composante environnementale et humaine. Cette donnée est reliée à des indicateurs de pression anthropique et climatique que sont les états de surface. Ces états conditionnent la répartition de la pluie en infiltration, évapotranspiration et écoulement. Ces indicateurs sont déterminés à partir d’images satellites LANDSAT. Pour calculer l’évolution annuelle des états de surface et des coefficients d’écoulement associés, on propose une méthodologie qui relie les différentes classes d’états de surface à l’évolution de la population. Les résultats des simulations sur un bassin de plus de 20 000 km2 au Burkina Faso montrent des améliorations très significatives dans les performances du modèle quand on utilise cette composante environnementale et humaine.Mots-clefs changement climatique; impact anthropique; modélisation hydrologique; Sahel; changements d’occupation du sol

Hydrological modelling and climate–man–environment interrelationships in the Burkinabe SahelAbstract In the Sahel, the studies conducted over the past two decades show that the combined effect of climate change and human activities on the surface conditions is the cause of a “hydrological paradox” seen since the 1970s on some watersheds in this region: the runoff coefficients have risen significantly, sometimes flows are higher than before, despite a decline in regional rainfall. These changes in rainfall–runoff relationships require new approaches to jointly take into account climate variability and the human dimension in the hydrological modelling of these basins. The study aims to integrate environmental and human elements in a hydrological model. These data are linked to indicators of human and climatic pressure: the land use in the Sahel which determines the distribution of rain between infiltration, evaporation and runoff. These indicators are characterized from Landsat satellite images at different times. To calculate the annual changes of land use and associated runoff coefficients, we propose a methodology that connects the various classes of land use and the evolution of the population, determined annually through a demographic model. The results of the simulations on a watershed of more than 20 000 km2 in Burkina Faso show very significant improvements in performance of a hydrological model when taking into account environmental and human changes.Key words anthropogenic impact; climate change; hydrological modelling; land-use/cover change; Sahel


Kriging method for estimation of groundwater resources in a

basin with scarce monitoring data

CHENGPENG LU1,2, LONGCANG SHU1,2, XUNHONG CHEN3, YUEZAN TAO4 & YING ZHANG1,2

1 State Key Laboratory of Hydrology-water Resources and Hydraulic Engineering, Hohai University, Nanjing, 210098, [email protected] College of Hydrology and Water Resource, Hohai University, Nanjing 210098, China3 School of Natural Resources, University of Nebraska-Lincoln, Lincoln, Nebraska 68583-0996, USA4 College of Civil Engineering and Hydraulic, Hefei University of Technology, Hefei 230009, China

Abstract Construction of the water table map is a key step in the assessment of water resources. However, the scarcity of groundwater monitoring data in some basins remains a problem for determination of a reliable variogram model, which is the starting point for kriging interpolation. Researchers have used the secondary variable, the sampling number of which is usually much greater than that of the primary variable, in assisting the spatial interpolation of the primary variable, e.g. by the regression kriging and co-kriging methods. These methods still require a variogram model to characterize the spatial structure of the primary variable. In this study, the authors proposed an approach that derives the variogram model of the groundwater level based on the elevation of the land surface data sets. The measurements of land surface elevation are widely available to researchers, and the density of the data locations is much larger than that of groundwater monitoring records. The land surface elevation was assumed to have a linear relationship with the groundwater level. A relationship between the variogram model for the groundwater level and the variogram model for the land surface elevation were established; the variogram model for the former can be directly inferred from the variogram model of the latter. In the derivation of the groundwater level variogram, the precipitation data can also be taken into account. This approach was implemented for the Nanjing watershed, China. A variogram model of the groundwater level was obtained from the DEM data set of 1000 m × 1000 m grid spacing.Key words kriging; regression; groundwater resources; geostatistics


An experimental and modelling investigation of macropore dominated subsurface stormflow in vegetated hillslopes of northeast India

RUPAK SARKAR1 & SUBASHISA DUTTA2

1 Faculty of Technology, Uttar Banga Krishi Viswavidyalaya (UBKV), Pundibari 736165, Cooch Behar, West Bengal, India

2 Department of Civil Engineering, Indian Institute of Technology, Guwahati 781039, North Guwahati, Assam, [email protected]

Abstract In northeast India, where the hillslopes are characterized by high degree of soil macroporosity and the area receives extreme rainfall events frequently during the monsoon season, rapid lateral preferential flow is the major source of storm runoff. Such rapid movement of storm water from the adjacent hilly areas either in the form of old or fresh water, often triggers soil erosion, landslides and flash floods in the rivers. However, due to lack of experimental data or reliable empirical models for these watersheds, very little seems to be known about infiltration behaviour, macropore connectivity, and subsurface flow pattern through soil macropores. Therefore, as a prerequisite to developing any rainfall–runoff model it is essential to characterize both surface and subsurface flow behaviour with good understanding of the critical hydrological processes prevailing in the region. The present investigation aims at characterizing the macropore flow for the region through in situ experiments and modelling. Key words macropores; hillslope; hydrology; experiments; modelling; northeast India


Coupled regional modelling of atmospheric–hydrologic processes for reconstruction of hydro-climate data and climate change assessment

M. L. KAVVAS1, Z. Q. CHEN2, N. OHARA1, M. L. ANDERSON2, A. J. SHAABAN3

& M. Z. M. AMIN3

1 Hydrologic Research Laboratory, Dept of Civil and Environmental Engineering, University of California, Davis, California 95616, USA [email protected] California Dept of Water Resources, 1416 9th St, Sacramento, California 95814, USA3 National Hydraulics Research Institute of Malaysia (NAHRIM), 43300 Seri Kembangan, Selangor, Malaysia

Abstract We describe the modelling of the Earth system over regions of varying spatial scale as a fully-coupled system of atmospheric processes aloft, coupled with the atmospheric boundary layer, land surface processes, and surface and subsurface hydrological processes. The interactions among the various component processes within the Earth system over a specified region are described, and an approach for modelling these interactions toward reconstruction of sparse hydro-climate data and assessment of climate change is discussed. The application of the resulting Regional Hydro-Climate Model (RegHCM) to several regions around the world is presented. Key words coupled atmospheric–hydrologic processes; regional hydro-climate model


Hydro-meteorological predictions from GCM simulations: downscaling techniques and uncertainty modelling

PRADEEP MUJUMDAR1, SUBIMAL GHOSH2 & DEEPASHREE RAJE1

1Department of Civil Engineering, Indian Institute of Science, Bangalore 560012, [email protected]

2Department of Civil Engineering, Indian Institute of Technology, Powai, Mumbai 400076, India

Abstract Hydrological implications of global climate change are usually assessed by downscaling appropriate predictors simulated by General Circulation Models (GCMs). Results from GCM simulations are subject to a number of uncertainties due to incomplete knowledge about the underlying geophysical processes of global change (GCM uncertainties) and uncertain future scenarios (scenario uncertainties). Disagreement between projections of regional climate change suggests that reliance on a single GCM with a few selected scenarios could lead to inappropriate planning and adaptation responses. This paper summarizes recent published work by the authors. The following methods and tools for statistical downscaling are discussed: (a) Fuzzy Clustering, (b) Relevance Vector Machine (RVM) and (c) Conditional Random Fields (CRFs). Uncertainty modelling with non-parametric methods and possibility theory are discussed. Applications of the methodologies are demonstrated by projection of the meteorological drought in the Orissa subdivision, India, and by predictions of the inflow to Hirakud Dam, Mahanadi River basin in India. Key words downscaling; uncertainty; fuzzy clustering; relevance vector machine; possibility; conditional random fields



Coupling VIC with GCM models to predict climate change impact in the Hanjiang basin, China

SHENGLIAN GUO, JING GUO, JUN ZHANG & HUA CHENState Key Laboratory of Water Resources and Hydropower Engineering Science, Wuhan University, Wuhan 430072,

China [email protected]

Abstract A Smooth Support Vector Machine (SSVM) is proposed for statistical downscaling of daily precipitation and temperature from GCM output. The Variable Infiltration Capacity (VIC) distributed hydrological model with a 9 × 9 km2 grid resolution is established and calibrated in the Hanjiang basin of China. Validation results show that SSVM can approximate observed precipitation and temperature data reasonably well, and the VIC model can simulate the runoff hydrograph with high model efficiency and low relative error. By applying the SSVM model, the trends of precipitation and temperature projected from CGCM2 under the A2 and B2 scenarios will decrease in the 2020s, and increase in the 2080s. However, in the 2050s, the precipitation will decrease under the A2 scenario and there will be no significant changes under the B2 scenario, but the temperature will be not obviously change under either scenario. Under both scenarios, the impact analysis of runoff made with the downscaled precipitation and temperature time series as input to the VIC distributed model, resulted in a decreasing trend for the 2020s and 2050s, and an overall increasing trend for the 2080s.Key words climate change; statistical downscaling; GCM; SSVM; VIC model; impact study


Runoff modelling within the Canadian Regional Climate Model (CRCM): analysis over the Quebec/Labrador watersheds

BILJANA MUSIC1, ANNE FRIGON1, MICHEL SLIVITZKY1, ANDRÉ MUSY1, DANIEL CAYA1 & RENÉ ROY2,1

1 Ouranos, Consortium sur la climatologie régionale et l’adaptation aux changements climatiques, 550 Sherbrooke Street West, 19th Floor, Montréal, Québec H3A 1B9, Canada [email protected] Hydro-Québec (IREQ), 1800 montée L. Boulet, Varennes, Québec J3X 1S1, Canada

Abstract This study focuses on evaluation of the hydrological performance of the Canadian Regional Climate Model (CRCM) coupled to the Canadian Land Surface Scheme (CLASS). The CRCM’s ability to adequately simulate annual mean runoff over 21 small watersheds in the Quebec/Labrador peninsula is assessed over the period 1961–1999. Since runoff is a spatial and temporal integrator of weather events, it represents a very useful variable for climate model validation, especially in areas where conventional surface weather observations are scarce. In addition, the sensitivity of simulated runoff to domain size and lateral boundary conditions is investigated. Results of the analysis indicate that CRCM tends to systematically underestimate observed annual mean runoff over most of the investigated watersheds. It was found that choice of simulation domain has a considerable effect on the simulated hydrological regime at the watershed scale. Different re-analyses used as driving data have less influence than domain size. However it may be important (larger than CRCM’s internal variability) when simulations are performed over a relatively small domain.Key words runoff; watershed; land-surface processes; internal variability; Canadian RCM; CLASS


Development of a coupled land-surface and hydrology model system for mesoscale hydrometeorological simulations

FEI YUAN1,2,3, HARALD KUNSTMANN3, CHUANGUO YANG1, ZHONGBO YU1, LILIANG REN1, BENJAMIN FERSCH3 & ZHENGHUI XIE2

1 State Key Laboratory of Hydrology, Water Resources and Hydraulic Engineering, Hohai University, No. 1 Xikang Road, Nanjing 20098, [email protected] LASG, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100029, China3 Forschungszentrum Karlsruhe, Institute for Meteorology and Climate Research IMK-IFU, Kreuzeckbahnstr. 19, D-82467 Garmisch-Partenkirchen, Germany

Abstract In this study, the coupled land-surface and hydrology model system (Noah LSM-HMS) was developed; it couples the Noah land-surface model (Noah LSM) with the large-scale hydrological model system (HMS). Detailed hydrological processes, such as unsaturated-zone soil moisture dynamics, river/lake–vadose and river/lake–groundwater exchange, streamflow routing, groundwater-table depth and horizontal groundwater flow are explicitly considered in this system. It is designed for interactive meteorological and hydrological simulations driven by a mesoscale meteorological model such as the Weather Research and Forecasting (WRF) system. Subsequently, Noah LSM-HMS was applied for streamflow simulations using the routine meteorological observations at 10-km resolution in the Chishui watershed in China. Results show that the streamflows calculated at the watershed outlet and two upstream hydrological stations are in reasonable agreement with those observed. Large differences between the simulated and observed streamflows still exist due to probable errors in the model structure and the meteorological forcings, especially the precipitation data.Key words land-surface model; large-scale hydrological model; runoff


A study coupling a large-scale hydrological model with a regional climate model

BIN YONG1, LILIANG REN1, LIHUA XIONG2, XIAOLI YANG1, WANCHANG ZHANG3, XI CHEN1 & SHANHU JIANG1

1 State Key Laboratory of Water Resources and Hydropower Engineering Science, Wuhan University, Wuhan 430072, China y ongbin_ [email protected] 2 State Key Laboratory of Hydrology, Water Resources and Hydraulic Engineering, Hohai University, Nanjing 210098, China3 Key Laboratory of Regional Climate-Environment Research for Temperate East Asia, Institute of Atmospheric Physics, CAS, Beijing 100029, China

Abstract On the basis of the improved SIMTOP runoff parameterization scheme and the three-layer soil moisture balance calculating method in the Xinanjiang model, this paper develops a simple, but highly-efficient, large-scale hydrological model (TOPX), which can provide the function for scaling transformation on the topographic index. Although the TOPX model has less data input and minimum parameters for calibrating, it can better describe two-dimensional hydrological processes. TOPX was coupled with the regional integrated environment modelling system (RIEMS) to use its ability of numerical simulation for the runoff in large-scale watersheds. The results of the off-line test performed at Youshui River catchment show that the TOPX model produced better simulations of daily runoff in small-sized catchments and it can capture the major characteristics of various hydrological processes. The RIEMS and TOPX coupled model was tested on-line in the Jinghe watershed. By means of the scale transformation scheme on topographic index and the yield and runoff routing theory, the coupling model used

meteorological data simulated by a regional climate model to drive the hydrological model for predicting the daily runoff at the large-scale watershed. A further analysis revealed that the accuracy of the distributed rainfall data simulated by the regional climate model (RIEMS) is the critical factor affecting the modelled runoff in the coupled model (RIEMS+TOPX). Key words topographic index; TOPMODEL; DEM; multiple flow direction algorithm


Large-scale water balance estimations through regional atmospheric moisture flux modelling and comparison to GRACE signals

BENJAMIN FERSCH1, HARALD KUNSTMANN1, NICO SNEEUW2 & BALAJI DEVARAJU2

1 Institute for Meteorology and Climate Research IMK-IFU, Forschungszentrum Karlsruhe, Kreuzeckbahnstrasse 19, D-82467 Garmisch-Partenkirchen, [email protected] Institute for Geodesy, University of Stuttgart, Geschwister-Scholl-Str. 24 D, D-70174 Stuttgart, Germany

Abstract Terrestrial water storage variations for continental-scale river catchments and basins derived from global and regional atmospheric moisture budgets modelling are evaluated and compared to GRACE satellite measurements. The regions considered in this study are the Amazon basin, the river catchments of Yenisei and Lena, the Sahara and Central Australia. If GRACE is taken as reference, the regional simulations have the potential to add value to the global moisture budgets for periods with small storage variation amplitudes. If the synoptic period is dominated by convective rainfall, the regional atmospheric model tends to overestimate precipitation. Key words joint land-surface–atmosphere modelling; WRF; GRACE; regional atmospheric modelling; continental water balance modelling; atmospheric moisture flux divergence


Assessment of flood events in the data-sparse Brahmaputra Basin in northeast India

U. C. SHARMA1 & VIKAS SHARMA2

1 Centre for Natural Resources Management, V. P. O. Tarore, District Jammu 181133 J&K, [email protected])2 S. K. University of Agricultural Sciences & Technology, Chatha, Jammu 180009 J&K, India

Abstract A preliminary quadratic relationship was developed to predict the flood events in the Brahmaputra Basin. A rainfall of 70 mm or more in a day, followed by spells of lower intensity during the following days can result in floods at the initial stages of the monsoon, while rains of even low intensity can cause floods during subsequent periods as the catchments get saturated. The slope of the catchment, amount of rainfall, vegetation cover, soil moisture and clay content of the soil were used to predict the runoff generation since rainfall and subsequent runoff generation impact the flood events. The observed and predicted values of runoff showed highly significant correlation (R = 0.8716), with a predictability of 75.9%.Key words assessment of flood events; Brahmaputra Basin, northeast India


Evaluation of a satellite-based near real-time global flood prediction system

KORAY K. YILMAZ1,2, ROBERT F. ADLER1,2, YANG HONG3 & HAROLD F. PIERCE2,4

1 Earth System Science Interdisciplinary Center, University of Maryland, College Park, Maryland 20740, [email protected] NASA Goddard Space Flight Center, Code 613.1, Greenbelt, Maryland 20771, USA3 School of Civil Engineering & Environmental Sciences, University of Oklahoma, Norman, Oklahoma 73019, USA4 Science Systems and Applications, Inc., Lanham, Maryland 20706, USA

Abstract This study provides an evaluation of a global flood prediction (GFS) system utilizing satellite-based rainfall and readily available geospatial data sets. The GFS, developed by our group, uses a relatively simple hydrological model, based on the runoff curve number method to transform rainfall into runoff. A grid-to-grid routing calculates the flow. Rainfall estimates are from TRMM Multi-satellite Precipitation Analysis (TMPA). An evaluation of the TMPA algorithm using a radar/gauge merged rainfall product over two basins in the southeast USA indicated that seasonal and regional considerations as well as basin size are important in using TMPA to drive hydrological models. GFS-based flood predictions were evaluated using observed streamflow data, MODIS-based inundation maps and a flood database. The GFS was able to simulate the onset of flood events produced by heavy rainfall; however, the prediction deteriorated in the later stages. This result points out the need for an improved routing component. The model showed dependency by the geographical region. A new hydrological model, with an improved physical represen-tation and routing component is currently under development and will likely lead to improved validation results.Key words satellite-based rainfall estimation; hydrological models; global flood prediction


Estimating precipitation for poorly-gauged areas in western China

JUNLIANG JIN1,2, GUIHUA LU1,3 & ZHIYONG WU1,3

1 State Key Laboratory of Hydrology, Water Resources and Hydraulic Engineering, Hohai University, Nanjing 210098, China [email protected]

2 College of Hydrology and Water Resources, Hohai University, Nanjing 210098, China3 Research Institute of Water Problems, Hohai University, Nanjing 210098, China

Abstract Hydrological simulations in data-sparse areas have large uncertainties. This paper proposes spatial geo-statistical interpolation algorithms based on the hydrological analogy method to estimate the spatial distribution of precipitation for data-sparse areas using Tropical Rainfall Measuring Mission (TRMM) precipitation radar (PR) data and a small number of available recorded rainfall data. Taking the Kaidu River basin in Xinjiang, China, as a case study, using to the relationship between TRMM PR data and sparsely-recorded rainfall, the spatial distribution of precipitation was estimated with the proposed method. A macro-scale land hydrological model, the Variable Infiltration Capacity (VIC) model, was then established over the study basin with the derived data. Hydrological simulation over five data-sparse basins (including Dashankou, Xining, Jiayuguan, Yingluoxia and Qingshizui) indicates that the estimated precipitation from TRMM PR data significantly improved the accuracy of hydrological simulation; the proposed method can therefore be used to estimate the spatial distribution of precipitation for sparsely-gauged areas in western China.Keywords data-sparse area; TRMM; hydrological simulation; VIC model



TRMM rainfall data estimation over the Peruvian Amazon-Andes basin and its assimilation into a monthly water balance model

WALDO SVEN LAVADO CASIMIRO1,2, DAVID LABAT2, JEAN LOUP GUYOT3, JOSYANE RONCHAIL4 & JUAN JULIO ORDOÑEZ1

1Servicio Nacional de Meteorología e Hidrología, SENAMHI, Casilla 11 1308, Lima 11, Perú [email protected]

2LMTG-Université de Toulouse-CNRS-IRD-OMP, 14 Avenue Edouard Belin, F-31400 Toulouse, France 3LMTG-Université de Toulouse-CNRS-IRD-OMP, IRD, CP 7091 Lago Sul, 71619-970 Brasília DF, Brazil4Université Paris 7 (IRD) and LOCEAN, Boite 100, 4 Place Jussieu, F-75252 Paris Cedex 05, France

Abstract The Peruvian Amazon-Andes basin corresponds to about 10% of the total Amazonian basin and is characterized by sparse rainfall data, particularly over the lowland zone (rainforest). We compare the 3B43 product of the Tropical Rainfall Measuring Mission (TRMM) with raingauge data over two sub-basins (Urubamba and Tambo) in the Ucayali basin located in the Peruvian Amazon-Andes basin. The spatial distribution of the 3B43 product is 0.25° 0.25° (approx. 27.8 27.8 km) and data are at a monthly scale. The period of comparison between on-site rainfall and 3B43 TRMM data is from January 1998 to December 2007. Comparison between on-site rainfall observations and 3B43 product is carried out using correlation coefficient and relative error. Improvement of the TRMM rainfall data is then proposed based on on-site rainfall data. After analysis of the 3B43 product, three sets of distributed rainfall data ( in situ, TRMM and on-site TRMM improved) were used as input in a GR2M monthly water balance model to simulate discharge at Urubamba and Tambo. Classical statistical overcalibration and validation procedure show a better accuracy in flow simulations, using only original TRMM data over Urubamba basin and improved data over Tambo basin. Key words Amazon basin; TRMM; Andes; Peru; monthly water balance model


Prospect for TRMM in rainfall estimates: a case study in the Laohahe basin, China

HONG WANG, LILIANG REN & XIAOLI YANGState Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, College of Hydrology and Water Resources, Hohai University, No. 1 Xikang Road, Nanjing 210098, [email protected]

Abstract Four years (2002–2005) of data from the Tropical Rainfall Measuring Mission (TRMM) 3B43 (V6) and 52 monthly raingauge measurements over the Laohahe basin in northeastern China were compared. The distribution of rainfall estimated from TRMM and raingauges showed the same spatial trend: a strong north–south gradient in rainfall. The time sequence patterns of the rainfall determined by both processes were remarkably similar (the correlation coefficient is 0.99). Analysis of rain climatology in different seasons and at 52 stations showed that the calculation period and site are important factors in estimating rainfall from TRMM. In the rainy season, the degree of accuracy of the TRMM estimate is more than 46%, while in the dry season, it is only about 12%. At the annual scale, only 23.1% of station data were accurately estimated, 0.2% underestimated and 75% overestimated. Therefore, in the Laohahe basin, TRMM is useful for estimating the average values of rainfall and good for long-term climatology with application to strategic water resource management.Key words rainfall; TRMM 3B43; raingauges; Laohahe basin, China




Evaluation of the Hydro-Estimator satellite rainfall algorithm and its utility in hydrological prediction in a mountainous region

ISMAIL YUCEL1, ROBERT J. KULIGOWSKI2 & DAVID GOCHIS3

1 Civil Engineering Department, Middle East Technical University, Ankara 06531, Turkey [email protected]

2 NOAA/NESDIS Center for Satellite Applications and Research (STAR), Camp Springs, Maryland 20746, USA3Research Applications Laboratory, National Center for Atmospheric Research, Boulder, Colorado 80305, USA

Abstract The performance of the NOAA/NESDIS operational rainfall estimation algorithm, the Hydro-Estimator (HE), is investigated with and without its orographic correction method, to assess its depiction of the timing, intensity and duration of convective rainfall, in general, and of the topography–rainfall relationship, in particular. With a few exceptions, validation of satellite rainfall estimates in complex terrain has been lacking to date, due to the paucity of pre-existing dense observation networks in mountainous areas. An event rainfall observation network in northwestern Mexico, established as part of the North American Monsoon Experiment (NAME), provides gauge-based precipitation measurements with sufficient temporal and spatial sampling characteristics to examine the climatological structure of diurnal convective activity over northwest Mexico. While the HE with orographic correction captures the spatial distribution and timing of diurnal convective events to some extent, elevation-dependent biases exist, which are characterized by underestimation of the occurrence of light precipitation at high elevations and overestimation of the occurrence of precipitation at low elevations. The potential of the HE to provide high spatial- and temporal-resolution data is tested in a hydrological application over the NAM region. The findings suggest that continued improvement of the HE orographic correction scheme is warranted, in order to advance quantitative precipitation estimation in complex terrain regions, and for use in hydrological applications.Key words rainfall algorithm; topography; satellite; observation; Hydro-Estimator (HE)


Spatially distributed evapotranspiration estimation using remote sensing and ground-based radiometers over cotton at Maricopa, Arizona, USA

ANDREW N. FRENCH, DOUGLAS HUNSAKER, KELLY THORP & THOMAS CLARKEUS Arid Land Agricultural Research Center, USDA/ARS, 21881 North Cardon Lane, Maricopa, Arizona 85238, [email protected]

Abstract Spatially distributed estimates of evapotranspiration (ET) could be valuable for monitoring croplands. Recently various ET estimation approaches have been developed that include visible and near infrared for vegetation indices and thermal infrared for surface temperatures. However, these approaches rely on high-resolution remotely sensed data (<100 m) that are too infrequent for operational use. To help fill the gap, a two-part ground and remote sensing approach is developed. The first part projects vegetation cover from the latest NDVI scenes using a simplified crop model. The second part projects spatially distributed surface temperatures using cover projections and ground-based radiometers. The projections from both parts are then input to a two-source energy balance model. The approach is demonstrated using data over a 2003 cotton experiment conducted in Maricopa, Arizona, USA. Using soil moisture depletion observations for validation, estimates were usually accurate to within 1 mm/d for two weeks beyond the


latest remote sensing acquisition.Key words evapotranspiration; surface energy balance; surface temperature; remote sensing; Maricopa, Arizona, USA


Applicable algorithm to map daily evapotranspiration using MODIS images for the Laohahe River basin, northeastern China

XIUQIN FANG1, LILIANG REN1, QIONGFANG LI1, QIUAN ZHU2, XIAOFAN LIU1 & YONGHUA ZHU1

1State Key Laboratory of Hydrology, Water Resources and Hydraulic Engineering, Hohai University, No.1 Xikang Road, Nanjing City, 210098, [email protected]

2 International Institute for Earth System Science, Nanjing University,No.22, Hankou Road, Nanjing City, 210093, China

Abstract An algorithm for mapping daily spatial actual evapotranspiration (ET) from remotely sensed MODIS data is presented. It is based on the surface energy balance scheme and the modified Priestley-Taylor equation, and has been applied to the MODIS data acquired during growing seasons over the Laohahe River basin, northeastern China. Regional daily ET values computed by the modified Xinanjiang hydrological model were used to validate ET values derived from MODIS data. The results were in good agreement, with a root mean square error of 0.3843 mm and correlation coefficient of 0.9029. It is suggested that the algorithm is applicable and operational for mapping daily actual ET over the study area.Key words daily actual evapotranspiration; MODIS; Priestley-Taylor equation; Laohahe River basin, China


Inverting the hydrological cycle: when streamflow measurements help assess altitudinal precipitation gradients in mountain areas

AUDREY VALERY, VAZKEN ANDRÉASSIAN & CHARLES PERRINCemagref, Hydrosystems and Bioprocesses Research Unit, BP 44, 92163 Antony cedex, France

[email protected]

Abstract This paper presents an attempt to “invert” the hydrological cycle and to use streamflow measurements to improve our knowledge of precipitation input in data-sparse mountainous regions. We use two data sets of 31 Swiss and 94 Swedish catchments, and three simple long-term water balance formulas. By assuming a simple two-parameter correcting model to regionalize precipitation from the too-sparse precipitation gauging network, we show that it is possible to identify, without ambiguity, the altitudinal precipitation gradient from streamflow. Although the snow undercatch parameter remains more difficult to identify, its range seems coherent with values found in the literature.Key words orographic precipitation gradient; water balance formulas; Budyko formula; Ol’dekop formula; Turc formula


Rainfall variability and uncertainty in water resource

assessments in South Africa

TENDAI SAWUNYAMA & DENIS HUGHESInstitute for Water Research, Rhodes University, Grahamstown, South Africa

[email protected]

Abstract Rainfall variability includes both spatial and temporal variability and is a potential source of uncertainty in rainfall–runoff modelling. The reliability of rainfall estimates depends on the accuracy of measurements, the number of raingauges and the spatial interpolation approach used to integrate point observations. However, degrading raingauge networks in South Africa represent a challenge to adequately account for this variability and extend rainfall records in practice. Therefore, there is need for correction procedures that address the uncertainty that exists in using such sparse data. The objective of this study is to demonstrate the importance of correcting original interpolated data sets to improve long-term estimates of spatial and temporal rainfall characteristics within South African catchments. The focus is on the generation of long time series of spatial rainfall over periods that span very different raingauge network densities. The study, through specific example sub-basins (e.g. improvements in simulation statistics such as coefficient of efficiency (CE) values of untransformed flows from 0.59 to 0.82 and 0.52 to 0.74 for 1959–1990 and 1991–2000 periods, respectively, for X31A sub-basin) demonstrated that a simple correction procedure based on rainfall frequency characteristics can be used to remove some uncertainties in spatial rainfall estimations and consequently model simulations.Key words rainfall variability; uncertainty; rainfall–runoff model; South Africa


Robust and flexible hydroinformatics to account for rainfall space–time variability in a data-sparse region

SAMEH CHARGUI1,2,3, CHRISTOPHE CUDENNEC2,3,4, MOHAMED SLIMANI1, JEAN-CHRISTOPHE POUGET5 & JALEL AOUISSI1,2,3

1 INAT, Lab. STE, 1082 Tunis, [email protected], [email protected]

2 Agrocampus Ouest, UMR1069, Soil Agro and hydroSystem, F-35000 Rennes, France3 INRA, UMR1069, Soil Agro and hydroSystem, F-35000 Rennes, France4 IRD, UMR G-EAU, 1004 Tunis, Tunisia5 IRD, UMR G-EAU, Quito, Equator

Abstract Rainfall is a major meteorological factor in water assessment in semi-arid basins. For mesoscale basins, the rainfall space–time variability is high and has a major influence on hydrological dynamics. A geomorphology-based rainfall–runoff approach is developed to account for rainfall space–time variability in a robust and flexible manner, according to changing available raingauging configurations. The approach is based on a two-step hydroinformatics protocol, firstly to analyse geomorphometry, and secondly to account for and benefit from available rainfall data. Key words rainfall–runoff; geomorphology-based transfer function; rainfall space–time variability; isochrone


Drought forecast using an artificial neural network for three hydrological zones in San Francisco River basin, Brazil

CELSO AUGUSTO G. SANTOS, BRUNO S. MORAIS & GUSTAVO B. L. SILVAFederal University of Paraíba, Department of Civil and Environmental Engineering, 58051-900 João Pessoa, Brazil


[email protected]

Abstract Three homogeneous rainfall areas were identified within San Francisco River basin, located in Northeast Brazil, by analysing the rainfall frequencies through the global wavelet power spectra that provide an unbiased and consistent estimation of the true power spectrum of the time series. Such study was accomplished using data from 248 raingauges provided by the Brazil National Water Agency (ANA), for several years between 1938 and 2005, based on their geographical distribution. For each identified region, the standardized precipitation index (SPI) was forecast using a feed-forward artificial neural network (ANN) trained by the back-propagation algorithm. The results obtained show that: the ANN is a suitable tool for this type of forecast; the accuracy is improved when the time scales of the SPI index, as well as the lead times, are increased; and the final result was not influenced by the different hydrological zones. Key words wavelet; fuzzy; ANN; drought; hydrological zones


Regional frequency analysis of annual precipitation in data-sparse regions using large-scale atmospheric variables

P. SATYANARAYANA & V. V. SRINIVASDepartment of Civil Engineering, Indian Institute of Science, 560 012 Bangalore, India

[email protected]

Abstract Regional precipitation frequency analysis (RPFA) is widely used for predicting precipitation quantiles at target sites in data-sparse areas. The RPFA involves fitting a frequency distribution to information pooled at target site from a region (group of similar sites). Conventional approaches to RPFA use precipitation statistics as attributes to form regions. Therefore, sufficient number of sites with contemporaneous data is required to form meaningful regions. This requirement cannot be met in data sparse areas. To address this issue, an approach is presented in this paper. Large-scale atmospheric variables affecting precipitation in the study area, location parameters (latitude, longitude and altitude) and seasonality of precipitation are suggested as attributes to form regions using fuzzy cluster analysis, and precipitation statistics are suggested for use in validating the delineated regions for homogeneity. Results from application to India indicate that the approach is effective for RPFA in data-sparse areas.Key words regionalization; precipitation frequency analysis; fuzzy cluster analysis; large-scale atmospheric variables; India


Impacts of rainfall uncertainty on water resource planning models in the Upper Limpopo basin, Botswana

P. K. KENABATHO, N. R. McINTYRE & H. S. WHEATERDepartment of Civil and Environmental Engineering, Imperial College London, South Kensington Campus, London SW7 2BU, [email protected]

Abstract In water resource planning in semi-arid Africa and comparable regions, uncertainty is high due to limitations in historic observations, uncertainty in hydrological models, uncertainty over future demands for water, and uncertain influences of future climate and hydrological change. The uncertainty in the future supply–demand balance should be considered in planning decisions, as it affects the risk associated with any planning option, and can help identify priorities for data collection. Focusing on rainfall and hydrological uncertainty, this paper outlines a framework of uncertainty analysis, which allows such consideration to be given. The framework consists of multi-site continuous time stochastic rainfall modelling to infill historic rainfall data. The stochastic infilling of rainfall data allows calibration of a

hydrological model under input uncertainty. The rainfall model, together with the uncertain hydrological model, is then used to generate multiple realisations of reservoir inflow over a 100-year period. This framework is applied to the Upper Limpopo basin in Botswana, using 25 years of observed daily rainfall and flow data for model calibration. A generalised linear model was used for the rainfall and a semi-distributed version of the IHACRES model was used for the hydrology. A proposed 382 106 m3 reservoir at the outlet of this catchment, which is part of Botswana’s national water resource strategy, is re-evaluated in light of the extended inflow data and the estimated uncertainty. Results show that the uncertainty has a considerable effect on the reliability of the reservoir; for example, the proportion of time for which demand for water was not met ranged from 0 to 13% over the different flow realisations. The main assumptions made, to be addressed in our future research, are stationarity of climate and that all the hydrological uncertainty arises from the historic rainfall uncertainty due to missing data.Key words IHACRES; rainfall–runoff; generalised linear models; semi-arid; Botswana; reservoir operation


A high-resolution hierarchical space–time framework for single storm events and its application for short-term rainfall forecasting

JUAN QIN1, MICHAEL LEONARD2, GEORGE KUCZERA1, MARK THYER1, ANDREW METCALFE2 & MARTIN LAMBERT2

1 School of Engineering, The University of Newcastle, Callaghan, New South Wales 2308, Australia2 School of Civil, Environmental & Mining Engineering, The University of Adelaide, Adelaide 5005, [email protected]

Abstract A new phenomenological hierarchical stochastic model is developed to robustly simulate rainfall fields consistent with 10-minute 1-km2 pixel radar images. The hierarchical framework has three levels. The first level simulates a latent Gaussian random field conditioned on the previous time step. In the second level, first-order autoregressive (AR(1)) models are used to describe the within-storm variations of the level-one parameters that control the evolution of rain fields. The third level is designed for simulation of storm sequences. Calibration and validation of the first two levels using an observed storm event (typical in Sydney, Australia) demonstrate that this two-level model produces realistic sequences of rain images which capture the physical hierarchical structure of clusters, patchiness of rain fields and the persistence exhibited during storm development. A variety of important statistics are adequately reproduced at both 10-minute and 1-hour time scales over various space scales. Application of this model to short-term rainfall forecasting is also presented.Key words stochastic space-time rainfall; hierarchical framework; high-resolution; block Toeplitz, circulant decomposition; generalized moments; parametric bootstrap; short-term forecasting

new approaches to hydrological prediction in data …hydrologie.org/redbooks/a333/p333 description,...

Documents