Does increased hydrochemical model complexity decrease robustness

The aim of this study was, within a sensitivity analysis framework, to determine if additional model complexity gives a better capability to model the hydrology and nitrogen dynamics of a small Mediterranean forested catchment or if the additional parameters cause over-fitting. Three nitrogen-models of varying hydrological complexity were considered. For each model, general sensitivity analysis (GSA) and Generalized Likelihood Uncertainty Estimation (GLUE) were applied, each based on 100,000 Monte Carlo simulations. The results highlighted the most complex structure as the most appropriate, providing the best representation of the non-linear patterns observed in the flow and streamwater nitrate concentrations between 1999 and 2002. Its 5% and 95% GLUE bounds, obtained considering a multi-objective approach, provide the narrowest band for streamwater nitrogen, which suggests increased model robustness, though all models exhibit periods of inconsistent good and poor fits between simulated outcomes and observed data. The results confirm the importance of the riparian zone in controlling the short-term (daily) streamwater nitrogen dynamics in this catchment but not the overall flux of nitrogen from the catchment. It was also shown that as the complexity of a hydrological model increases over-parameterisation occurs, but the converse is true for a water quality model where additional process representation leads to additional acceptable model simulations. Water quality data help constrain the hydrological representation in process-based models. Increased complexity was justifiable for modelling river-system hydrochemistry. Increased complexity was justifiable for modelling river-system hydrochemistry.

[1]  A. Butturini,et al.  Modelling the inorganic nitrogen behaviour in a small Mediterranean forested catchment, Fuirosos (Catalonia) , 2010 .

[2]  Henrik Madsen,et al.  Sensitivity of soil parameters in unsaturated zone modelling and the relation between effective, laboratory and in situ estimates , 2005 .

[3]  Pilar Llorens,et al.  Hydrological processes and their seasonal controls in a small Mediterranean mountain catchment in the Pyrenees , 2002 .

[4]  J. Kirchner.  Getting the right answers for the right reasons: Linking measurements, analyses, and models to advance the science of hydrology , 2006 .

[5]  Jan Seibert,et al.  Reliability of Model Predictions Outside Calibration Conditions , 2003 .

[6]  Doerthe Tetzlaff,et al.  Transit time distributions of a conceptual model: their characteristics and sensitivities , 2010 .

[7]  D. Lozano,et al.  Mapa geologico de Espana , 1919 .

[8]  Ezio Todini,et al.  Recent developments in predictive uncertainty assessment based on the model conditional processor approach , 2010 .

[9]  Keith Beven,et al.  Equifinality, data assimilation, and uncertainty estimation in mechanistic modelling of complex environmental systems using the GLUE methodology , 2001 .

[10]  Soroosh Sorooshian,et al.  Toward improved calibration of hydrologic models: Multiple and noncommensurable measures of information , 1998 .

[11]  Keith Beven,et al.  Uncertainty assessment of a process-based integrated catchment model of phosphorus , 2009 .

[12]  Jeffrey L. Smith,et al.  Nitrous oxide flux from a shrub-steppe ecosystem: Sources and regulation , 1994 .

[13]  H. Birch.  Nitrification in soils after different periods of dryness , 2005, Plant and Soil.

[14]  E. Sober,et al.  The Principle of Parsimony , 1981, The British Journal for the Philosophy of Science.

[15]  Keith Beven,et al.  The future of distributed models: model calibration and uncertainty prediction. , 1992 .

[16]  G. Hornberger.  Eutrophication in peel inlet—I. The problem-defining behavior and a mathematical model for the phosphorus scenario , 1980 .

[17]  A. Butturini,et al.  Modelling the non‐linear hydrological behaviour of a small Mediterranean forested catchment , 2008 .

[18]  J. Nash,et al.  River flow forecasting through conceptual models part I — A discussion of principles☆ , 1970 .

[19]  A. Butturini,et al.  Influences of the stream groundwater hydrology on nitrate concentration in unsaturated riparian area bounded by an intermittent Mediterranean stream , 2003 .

[20]  J. Freer,et al.  Processes affecting transfer of sediment and colloids, with associated phosphorus, from intensively farmed grasslands: a critical note on modelling of phosphorus transfers , 2007 .

[21]  Francesc Sabater,et al.  Seasonal Variations of Dissolved Nitrogen and DOC:DON Ratios in an Intermittent Mediterranean Stream , 2005 .

[22]  K. Beven.  Rainfall-Runoff Modelling: The Primer , 2012 .

[23]  Soroosh Sorooshian,et al.  Sensitivity analysis of a land surface scheme using multicriteria methods , 1999 .

[24]  Michael Rode,et al.  Multi-objective calibration of a river water quality model—Information content of calibration data , 2007 .

[25]  Osvaldo E. Sala,et al.  Hierarchy of responses to resource pulses in arid and semi-arid ecosystems , 2004, Oecologia.

[26]  Xunhong Chen,et al.  Hydrologic connections of a stream–aquifer-vegetation zone in south-central Platte River valley, Nebraska , 2007 .

[27]  H. Birch.  Further observations on humus decomposition and nitrification , 1959, Plant and Soil.

[28]  Katri Rankinen,et al.  An application of the GLUE methodology for estimating the parameters of the INCA-N model. , 2006, The Science of the total environment.

[29]  George Kuczera,et al.  Assessment of hydrologic parameter uncertainty and the worth of multiresponse data , 1998 .

[30]  Roman Krzysztofowicz,et al.  Bayesian theory of probabilistic forecasting via deterministic hydrologic model , 1999 .

[31]  H. Birch.  Mineralisation of plant nitrogen following alternate wet and dry conditions , 1964, Plant and Soil.

[32]  R. Merckx,et al.  Microbial biomass responses to soil drying and rewetting: The fate of fast- and slow-growing microorganisms in soils from different climates , 1993 .

[33]  Changsheng Li,et al.  Bi-criteria evaluation of the MIKE SHE model for a forested watershed on the South Carolina coastal plain , 2010 .

[34]  A. Butturini,et al.  The influence of riparian-hyporheic zone on the hydrological responses in an intermittent stream , 2002 .

[35]  W. Schlesinger,et al.  Factors Controlling Denitrification in a Chihuahuan Desert Ecosystem , 1991 .

[36]  Félix Francés,et al.  Split-parameter structure for the automatic calibration of distributed hydrological models , 2007 .

[37]  James F. Reynolds,et al.  Modifying the ‘pulse–reserve’ paradigm for deserts of North America: precipitation pulses, soil water, and plant responses , 2004, Oecologia.

[38]  Thorsten Wagener.  Evaluation of catchment models , 2003 .

[39]  E. J. Wilson,et al.  A semi-distributed ntegrated itrogen model for multiple source assessment in tchments (INCA): Part I — model structure and process equations , 1998 .

[40]  A. Wade,et al.  Understanding hydrological and nitrogen interactions by sensitivity analysis of a catchment-scale nitrogen model , 2005 .

[41]  Thorsten Wagener,et al.  Numerical and visual evaluation of hydrological and environmental models using the Monte Carlo analysis toolbox , 2007, Environ. Model. Softw..

[42]  Osvaldo E. Sala,et al.  Thresholds, memory, and seasonality: understanding pulse dynamics in arid/semi-arid ecosystems , 2004, Oecologia.

[43]  George M. Hornberger,et al.  On modeling the mechanisms that control in‐stream phosphorus, macrophyte, and epiphyte dynamics: An assessment of a new model using general sensitivity analysis , 2001 .

[44]  A. Butturini,et al.  Calibration of the INCA model in a Mediterranean forested catchment: the effect of hydrological inter-annual variability in an intermittent stream , 2004 .