Table 2 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Optimal parameter sets of the TERRA/LM model for the different land surface sites: z 0 is the roughness length, a sw the short- wave albedo, R S,min the minimum stomatal resistance. veg is the vegetation ratio, LAI the leaf area index, c s the soil heat capacity, and n FC the field capacity.
Source publication
The turbulent sensible and latent heat fluxes simulated in the operational weather forecast model LM have been checked with data from the field experiment LITFASS 2003 (Lindenberg Inhomogeneous Terrain - Fluxes between Atmosphere and Surface: a Long-term Study) using both single site measurements and grid box aggregated fluxes. SCE-UA (single objec...
Contexts in source publication
Context 1
... SCE-UA (single objective) and the MOSCEM-UA (multi objective) approaches are applied to calibrate the land- surface scheme TERRA/LM for 11 single sites in the area of the LITFASS 2003 experiment. The parameters selected, their prescribed numerical ranges and their respective values as given in the standard TERRA/LM are listed in Table 2. With the parameters found by calibration (Table 3) TERRA/ LM is validated using an alternative data set from the same sites, using the procedure described earlier. ...
Context 2
... this study, the single-criterion SCE-UA algorithm (Duan, 1994) and the multi-criteria algorithm MOSCEM-UA have been applied consecutively. SCE-UA is used to calibrate a total of 12 parameters as listed in Table 3 while, in a second step, MOSCEM-UA is used to calibrate a subset of 6 (Table 2). The target single objective function for the SCE-UA calibration is the sum of the squared Nash- Sutcliffe measures of latent and sensible heat. ...
Context 3
... procedure was adopted because of the enormous computer resources required for the multi-objective MOSCEM-UA calibration for as many as 12 parameters in a reasonable time. The parameters listed in Tables 2 and 3 were subject to the single objective calibration. Then the parameters in Table 3 were fixed and only the parameters in Table 2 were calibrated by the multi- objective approach. ...
Context 4
... parameters listed in Tables 2 and 3 were subject to the single objective calibration. Then the parameters in Table 3 were fixed and only the parameters in Table 2 were calibrated by the multi- objective approach. The parameters selected for the final multi-objective calibration (Table 2) should characterise the surface (roughness length and albedo), the vegetation (minimum stomatal resistance and leaf area index) and the soil (soil heat capacity and field capacity). ...
Context 5
... the parameters in Table 3 were fixed and only the parameters in Table 2 were calibrated by the multi- objective approach. The parameters selected for the final multi-objective calibration (Table 2) should characterise the surface (roughness length and albedo), the vegetation (minimum stomatal resistance and leaf area index) and the soil (soil heat capacity and field capacity). Their ranges are prescribed according to measurements taken during the LITFASS 2003 experiment (Beyrich et al., 2004). ...
Context 6
... parameters in Table 2 found by the multi-objective calibration describe the objective functions closest to the lower left corner in Fig. 4. They show quite diffuse behaviour, with no clear distinction between the different land use classes and large differences occur for sites with identical vegetation types. ...
Similar publications
Wave conditions in the northern Baltic Proper during windstorm Erwin/Gudrun (January 2005) are analysed based on in situ measurements in three locations and output of operational wave models from the German Weather Forecast Service, the Danish Meteorological Institute and the Finnish Institute of Marine Research. The measured significant wave heigh...
Comparison between proxies for atmospheric circulation and temperature reveals associations over the last few decades that are inconsistent with those of the past 2000 years. Notably, patterns of middle to high latitude atmospheric circulation in both hemispheres are still within the range of variability of the last 6?10 centuries while, as demonst...
In high mountain areas, permafrost is important because it influences natural hazards and construction practices, and because it is an indicator of climate change. The modeling of its distribution and evolution over time is complicated by steep and complex topography, highly variable conditions at and below the surface, and varying climatic conditi...
Rainfall collected twice weekly from 3 nearby UK stations between November 2004 and October 2006 allows local differences in ?<sup>18</sup>O and ?<sup>2</sup>H to be observed. Local gradients in ?<sup>18</sup>O <sub>p</sub> appear to be of the same order of magnitude as national trends, suggesting the spatial ?<sup>18</sup>O <sub>p</sub> picture is...
The prediction of temporal concentration profiles of a transported pollutant in a river is still a subject of ongoing research efforts worldwide. The present paper is aimed at studying the possibility of using Multi-Layer Perceptron Neural Networks to evaluate the whole concentration versus time profile at several cross-sections of a river under va...
Citations
... It is also superior to the conventional MCMCS in its ability to estimate Pareto optimal solutions because it prevents clustering of evolved solutions in the compromized region as well as premature convergence when highly correlated performance criteria are used for model calibration (Vrugt et al., 2003). Researchers in various fields have proven its efficacy for finding Pareto optimal solutions (Vrugt et al., 2003;Johnsen et al., 2005;Schoups et al., 2005). ...
This study aims to propose a method for effectively recognising and evaluating model structural uncertainty. It began with a comparative assessment of various model structures that have differing features regarding the rainfall-runoff mechanism and DEM spatial resolution. The assessment applied a multi-objective optimisation method (MOSCEM-UA) with two objective functions (simple least-squares and the heteroscedastic maximum likelihood estimator), and focused on five historical flood events. The study was based on the assumptions that a structurally sound model assures improved prediction results (either minimized or maximized model performance measure), allows constant model performance with regard to objective functions (a small Pareto solution set), and yields good applicability of a calibrated parameter set to various events (good parameter stability). The results indicated that KWMSS, a distributed model, was superior to SFM, a simple lumped model, when estimating a Pareto solution set and assessing parameter stability for the applied events. In addition, three different spatial resolutions (250 m, 500 m, and 1 km) were compared to assess the structural uncertainty due to changes in the topographical representation in distributed rainfall-runoff modelling. The results indicated that the 250 and 500 m models were Pareto-equivalent, containing similar Pareto fronts, and both produced Pareto results superior to the 1 km model. Both models also yielded parameter stability values that were much more superior to the model based on a 1 km DEM. As the topographic representation became more detailed, the model showed a tendency to have less structural uncertainty in terms of guarantying better performance, better parameter stability, and a smaller Pareto solution set. On the other hand, the output of a spatially detailed model was likely to be insensitive to the variation of model parameters (i.e. equifinality). Copyright © 2011 John Wiley & Sons, Ltd.
... However, model calibration using other observations has received far less attention. Examples using one type of observation are the use of ground water data [42], turbulent fluxes [22], [39], or observed soil moisture data, obtained either in situ [5], [26] or through remote sensing [32], [44]. ...
In most hydrologic modeling studies, the hypothesis is made that an improvement in the modeled soil moisture leads to an improvement in the modeled surface energy balance. The objective of this paper is to assess whether this hypothesis is true. The study was performed over the winter wheat fields in the AgriSAR 2006 domain. Remotely sensed soil moisture values and latent heat fluxes were used, in combination with in situ observations. First, the land cover and saturated subsurface flow parameters were estimated using the in situ observations. A spatially distributed model simulation was then performed, for which the Brooks-Corey parameters were derived from a soil texture map, and of which the results were validated using the remote sensing data. The remotely sensed soil moisture values were then used to optimize the Brooks-Corey parameters. As expected, a better performance with respect to the soil moisture estimation was obtained. However, this did not improve the latent heat flux estimates. This can be explained by the consumption of water from the deeper soil layers by the vegetation. The overall conclusion is that, under conditions where evapotranspiration is limited by energy and not by the soil moisture content, surface soil moisture values alone are not sufficient for the optimization of hydrologic model results. More data sets are needed for this purpose.
... The parameters were optimised simultaneously for Q H,EBC and Q E,EBC with the coefficient of efficiency serving as the objective 5 function (see Eq. (10)) using a SCE-UA (Shuffled Complex Evolution University of Arizona) algorithm (Duan et al., 1992(Duan et al., , 1994 followed by a MOSCEM-UA (Multiobjective Shuffled Complex Evolution Metropolis -University of Arizona) algorithm (Vrugt et al., 2003). This procedure was applied to SEWAB just as described in detail by Johnsen et al. (2005). The performance of these runs and their influence on the uncertainty of 10 the upscaled fluxes is discussed in Sect. ...
Meteorologically measured fluxes of energy and matter between the surface and the atmosphere originate from a source area of certain extent, located in the upwind sector of the device. The spatial representativeness of such measurements is strongly influenced by the heterogeneity of the landscape. The footprint concept is capable of linking observed data with spatial heterogeneity. This study aims at upscaling eddy covariance derived fluxes to a grid size of 1 km edge length, which is typical for mesoscale models or low resolution remote sensing data.
Here an upscaling strategy is presented, utilizing footprint modelling and SVAT modelling as well as observations from a target land-use area. The general idea of this scheme is to model fluxes from adjacent land-use types and combine them with the measured flux data to yield a grid representative flux according to the land-use distribution within the grid cell. The performance of the upscaling routine is evaluated with real datasets, which are considered to be land-use specific fluxes in a grid cell. The measurements above rye and maize fields stem from the LITFASS experiment 2003 in Lindenberg, Germany and the respective modelled timeseries were derived by the SVAT model SEWAB. Contributions from each land-use type to the observations are estimated using a forward lagrangian stochastic model. A representation error is defined as the error in flux estimates made when accepting the measurements unchanged as grid representative flux and ignoring flux contributions from other land-use types within the respective grid cell.
Results show that this representation error can be reduced up to 56 % when applying the spatial integration. This shows the potential for further application of this strategy, although the absolute differences between flux observations from rye and maize were so small, that the spatial integration would be rejected in a real situation. Corresponding thresholds for this decision have been estimated as a minimum mean absolute deviation in modelled timeseries of the different land-use types with 35 W m<sup>−2</sup> for the sensible heat flux and 50 W m<sup>−2</sup> for the latent heat flux. Finally, a quality lagging scheme to classify the data with respect to representativeness for a given grid cell is proposed, based on an overall flux error estimate. This enables the data user to infer the uncertainty of mesoscale models and remote sensing products with respect to ground observations. Major uncertainty sources remaining are the lack of an adequate method for energy balance closure correction as well as model structure and parameter estimation, when applying the model for surfaces without flux measurements.
... All optimized parameter sets were provided by the GKSS Institute in Geesthacht. An exemplary description of the optimization for TERRA is given in JOHNSEN et al. (2006). ...
Commonly used measuring methods only allow for a closure of the energy balance at the Earth's surface except for a residual. Models, in contrast, inherently demand an entirely closed energy balance, which they achieve in different ways. This paper examines the model simulations depending on the methods used for closing the energy balance, the optimization of the parameters, and the equations used. The one-dimensional surface vegetation atmosphere transfer (SVAT) models REMO, SEWAB and TERRA, which were made available by the GKSS Institute in Geesthacht, are examined and compared with measured data from the LITFASS-2003 experiment. LITFASS-2003 took place during May and June 2003 in the experimental fields of the Meteorological Observatory Lindenberg. Parameter sets optimized with regard to the turbulent fluxes yield better results, but are often not consistent with site characteristics and differ within the models. We ran the models with the optimized parameter sets, and with a set of selected parameters which agree with the respective sites. SEWAB closes the energy balance by using an iteration of the surface temperature, through which the influence of the residual is distributed over all fluxes. In REMO and TERRA the soil heat flux is estimated as the residual of the other components of the energy balance equation. This results in an extreme overestimation of this heat flux. Parameter optimization has an impact on the partitioning of the available energy and its distribution to the turbulent heat fluxes. It does not influence the effect of the residual on the results. Currently we do not know how to close the measured energy balance at the surface appropriately because of the underestimation of the turbulent fluxes by eddy-covariance measurements and the unknown partitioning of the residual between the sensible and latent heat flux. Therefore, the method used in SEWAB yields better results than the one used in REMO and TERRA. The method used to close the energy balance has more impact on the simulated surface energy balance components than does their parameterization. For this reason it is not possible to attribute the differences in the calculated heat fluxes to particular differences in the equations and parameters.
German
Gewöhnlich sind unsere Messmethoden nur in der Lage, die Energiebilanz an der Erdoberfläche bis auf ein Residuum zu schließen. Modelle erfordern dagegen eine interne Schließung der Energiebilanz, welche auf verschiedene Weise erfolgen kann. In diesem Artikel wird das Energiebilanz-Schließungsproblem in Modellsimulationen untersucht in Abhängigkeit von der Schließungsmethode, der Optimierung der Parameter und der angewandten Gleichungen. Die eindimensionalen SVAT-Schemata der Modelle REMO, SEWAB und TERRA, welche durch das GKSS Institut in Geesthacht zur Verfügung gestellt wurden, wurden genutzt und verglichen mit Messdaten des LITFASS-2003 Experiment. LITFASS-2003 fand im Mai und Juni 2003 auf Messflächen des Meteorologischen Observatoriums Lindenberg statt. Die anhand der turbulenten Flüsse optimierten Parameter der drei Modelle sind oft nicht konsistent mit den Messflächeneigenschaften und unterscheiden sich für die verschiedenen Modelle. Die Modellrechnungen wurden daher mit den optimierten Parametersätzen und mit einem Satz ausgewählter Parameter, die mit den Messflächen übereinstimmen, durchgeführt. SEWAB schließt die Energiebilanz durch eine iterative Bestimmung der Oberflächentemperatur, womit der Einfluss des Residuums auf alle Flüsse verteilt wird. In REMO und TERRA wird der Bodenwärmestrom als Restglied aus allen anderen Komponenten der Energiebilanzgleichung bestimmt. Daraus folgt eine extreme Überbestimmung dieses Wärmestroms. Die Parameteroptimierung hat Einfluss auf die Verteilung der verfügbaren Energie auf die turbulenten Wärmeflüsse, beeinflusst aber nicht den Effekt des Residuums auf die Ergebnisse. Aufgrund der Unterbestimmung der turbulenten Wärmeflüsse mit der Eddy-Kovarianz Messungen und der bisher unbekannten Zusammensetzung des Residuums aus fühlbarem und latentem Wärmestrom wird die in SEWAB verwendete Methode als besser angesehen als die Methodik in TERRA und REMO. Die Methode der Schließung der Energiebilanz hat einen größeren Einfluss auf die simulierten Energiebilanzkomponenten als ihre Parametrisierung. Dadurch ist es nicht möglich die Unterschiede bei den Wärmeflüssen speziellen Unterschieden in den Gleichungen und Parametersätzen zuzuordnen.
... Hence, the objective of this paper is to develop an event based lumped conceptual rainfall-runoff model and then to compare mono-objective and multi-objective calibra- 10 tion approaches. The emphasis is put on the impact of the selected objective functions on the actual hydrograph shapes and simulation errors rather than on the calibration algorithm as several studies have already investigated this issue (Yu and Yang, 2000; Johnsen et al., 2005; Tang et al., 2006). The Gardon catchment located in southern France is used in the applications because of the recurrent flooding problems its in- 15 ...
A conceptual lumped rainfall-runoff flood event model was developed and applied on the Gardon catchment located in southern France and various mono-objective and multi-objective functions were used for its calibration. The model was calibrated on 15 events and validated on 14 others. The results of both the calibration and validation phases are compared on the basis of their performance with regards to six criteria, three global criteria and three relative criteria representing volume, peakflow, and the root mean square error. The first type of criteria gives more weight to strong events whereas the second considers all events to be of equal weight. The results show that the calibrated parameter values are dependent on the type of criteria used. Significant trade-offs are observed between the different objectives: no unique set of parameter is able to satisfy all objectives simultaneously. Instead, the solution to the calibration problem is given by a set of Pareto optimal solutions. From this set of optimal solutions, a balanced aggregated objective function is proposed, as a compromise between up to three objective functions. The mono-objective and multi-objective calibration strategies are compared both in terms of parameter variation bounds and simulation quality. The results of this study indicate that two well chosen and non-redundant objective functions are sufficient to calibrate the model and that the use of three objective functions does not necessarily yield different results. The problems of non-uniqueness in model calibration, and the choice of the adequate objective functions for flood event models, emphasise the importance of the modeller's intervention. The recent advances in automatic optimisation techniques do not minimise the user's responsibility, who has to chose multiple criteria based on the aims of the study, his appreciation on the errors induced by data and model structure and his knowledge of the catchment's hydrology.
... Neste artigo é descrito um algoritmo de calibração automática multi-objetivo de modelos hidrológicos baseado em um algoritmo evolucionário. O algoritmo original foi desenvolvido por Vrugt et al. (2003) e aplicado no modelo SAC-SMA, ou Sacramento, e outros (por exemplo, Schoups et al., 2005;Johnsen et al., 2005;Pande et al., 2005;Bos e Vreng, 2006). Esse trabalho apresenta a descrição do algoritmo bem como uma aplicação usando o modelo hidrológico IPH II (Tucci, 1998). ...
... However, it is well known that measurements usually do not show a closure of the surface energy balance (e.g., Foken and Oncley 1995;Wilson et al. 2002). When comparing the sum of the sensible and latent heat flux with the difference of net radiation and soil heat flux from 30-minmeasured averages, the rmse of all stations varies between 70 and 130 W m-2 , and the bias (mean absolute residual) varies between 57 and 107 W m-2 ( Johnsen et al. 2005). There is a year-long discussion about the source for the energy balance closure gap. ...
The representation of subgrid-scale surface heterogeneities in numerical
weather and climate models has been a challenging problem for more than
a decade. The Evaporation at Grid and Pixel Scale (EVA-GRIPS) project
adds to the numerous studies on vegetation-atmosphere interaction
processes through a comprehensive field campaign and through simulation
studies with land surface schemes and mesoscale models. The mixture of
surface types in the test area in eastern Germany is typical for larger
parts of northern Central Europe. The spatial scale considered
corresponds to the grid scale of a regional atmospheric weather
prediction or climate model and to the pixel scale of satellite images.
Area-averaged fluxes derived from point measurements, scin-tillometer
measurements, and a helicopter-borne turbulence probe were widely
consistent with respect to the sensible heat flux. The latent heat flux
from the scintillometer measurements is systematically higher than the
eddy covariance data. Fluxes derived from numerical simulations proved
the so-called mosaic approach to be an appropriate parameterization for
subgrid heterogeneity.
... However, it is well known that measurements usually do not show a closure of the surface energy balance (e.g., Foken and Oncley 1995; Wilson et al. 2002). When comparing the sum of the sensible and latent heat flux with the difference of net radiation and soil heat flux from 30-min- measured averages, the rmse of all stations varies between 70 and 130 W m –2 , and the bias (mean absolute residual) varies between 57 and 107 W m –2 (Johnsen et al. 2005). There is a year-long discussion about the source for the energy balance closure gap. ...
... However, it is well known that measurements usually do not show a closure of the surface energy balance (e.g., Foken and Oncley 1995;Wilson et al. 2002). When comparing the sum of the sensible and latent heat flux with the difference of net radiation and soil heat flux from 30-minmeasured averages, the rmse of all stations varies between 70 and 130 W m-2 , and the bias (mean absolute residual) varies between 57 and 107 W m-2 ( Johnsen et al. 2005). There is a year-long discussion about the source for the energy balance closure gap. ...
The exchange of energy, water, and momentum across the land-atmosphere interface is a vital process to be parameterized in climate and numerical weather prediction (NWP) models. Remarkable progress has been made during the last two decades in the development and validation of land-surface schemes. There is no doubt that advanced descriptions of the role of soil and vegetation for the heat and water transfer have improved our understanding of how the surface affects the atmospheric boundary layer and controls weather and climate simulations. Due to their 1-dimensional nature early land-surface schemes assumed a homogeneous vegetation cover for the model grid while today most of the models allow for a composite of different types of natural surfaces. A major focus of current land-surface model development is the complexity with which sub-grid-scale horizontal inhomogeneity should be considered. Under the auspices of the German climate research program DEKLIM, the EVA_GRIPS project (Evaporation at Grid and Pixel Scale) focuses on the determination of area-averaged surface fluxes over heterogeneous terrain at the meso- scale. The project combines experimental field work, the analysis of satellite data, and numerical modeling. A five-weeks field experiment (LITFASS-2003) was performed in the heterogeneous landscape around the Meteorological Observatory Lindenberg in the East of Germany during the growing season 2003. The experiment brought together micrometeorological, ground based remote sensing and airborne measurement systems. Comprehensive data sets from the experiment are used to force, calibrate and verify 1-dimensional land-surface schemes, non-hydrostatic mesoscale models and a large-eddy simulation model. The purpose of this article is to inform the community about the intention, methodology and results of the EVA-GRIPS project and about its data sets, which are available for model testing and validation. Consideration of sub-grid-scale inhomogeneities in land-surface schemes is recognized as an important step in model improvement, but not yet realized appropriately in many NWP and climate models. The article will inform the reader about the experimental and modelling work carried out in EVA_GRIPS. This includes a description of the measurement strategy and the presentation of some exemplary results from the LITFASS-2003 field experiment. Further, we will describe both, the implementation of sub-grid-scale information in different land-surface schemes and the impact on the boundary layer development as simulated with mesoscale models. The paper will also give some critical insights from the analysis of remotely sensed data and their application in numerical weather prediction models
In most regions of the world, and particularly in developing countries, the possibility and reliability of hydrologic predictions is severely limited, because conventional measurement networks (e.g. rain and stream gauges) are either nonexistent or sparsely located. This study, therefore, investigates various systems methods and newly available data acquisition techniques to evaluate their potential for improving hydrologic predictions in poorly gaged and ungaged watersheds. Part One of this study explores the utility of satellite-remote-sensing-based rainfall estimates for watershed-scale hydrologic modeling at watersheds in the Southeastern U.S. The results indicate that satellite-based rainfall estimates may contain significant bias which varies with watershed size and location. This bias, of course, then propagates into the hydrologic model simulations. However, model performance in large basins can be significantly improved if short-term streamflow observations are available for model calibration. Part Two of this study deals with the fact that hydrologic predictions in poorly gauged/ungauged watersheds rely strongly on a priori estimates of the model parameters derived from observable watershed characteristics. Two different investigations of the reliability of a priori parameter estimates for the distributed HL-DHMS model were conducted. First, a multi-criteria penalty function framework was formulated to assess the degree of agreement between the information content (about model parameters) contained in the precipitation-streamflow observational data set and that given by the a priori parameter estimates. The calibration includes a novel approach to handling spatially distributed parameters and streamflow measurement errors. The results indicated the existence of a significant trade-off between the ability to maintain reasonable model performance while maintaining the parameters close to their a priori values. The analysis indicates those parameters responsible for this discrepancy so that corrective measures can be devised. Second, a diagnostic approach to model performance assessment was developed based on a hierarchical conceptualization of the major functions of any watershed system. "Signature measures" are proposed that effectively extract the information about various watershed functions contained in the streamflow observations. Manual and automated approaches to the diagnostic model evaluation were explored and were found to be valuable in constraining the range of parameter sets while maintaining conceptual consistency of the model.
Streamflow depends on the soil moisture of a river catchment and can be measured with relatively high accuracy. The soil moisture in the root zone influences the latent heat flux and, hence, the quantity and spatial distribution of atmospheric water vapour and precipitation. As numerical weather forecast and climate models require a proper soil moisture initialization for their land surface models, we enhanced an Ensemble Kalman Filter to assimilate streamflow time series into the multi-layer land surface model TERRA-ML of the regional weather forecast model COSMO. The impact of streamflow assimilation was studied by an observing system simulation experiment in the Enz River catchment (located at the downwind side of the northern Black Forest in Germany). The results demonstrate a clear improvement of the soil moisture field in the catchment. We illustrate the potential of streamflow data assimilation for weather forecasting and discuss its spatial and temporal requirements for a corresponding, automated river gauging network.
