|
|
Minet, J., Laloy, E., Tychon, B., & François, L. (2015). Bayesian inversions of a dynamic vegetation model at four European grassland sites. Biogeosciences, 12(9), 2809–2829.
Abstract: Eddy covariance data from four European grassland sites are used to probabilistically invert the CARAIB (CARbon Assimilation In the Biosphere) dynamic vegetation model (DVM) with 10 unknown parameters, using the DREAM((ZS)) (DiffeRential Evolution Adaptive Metropolis) Markov chain Monte Carlo (MCMC) sampler. We focus on comparing model inversions, considering both homoscedastic and heteroscedastic eddy covariance residual errors, with variances either fixed a priori or jointly inferred together with the model parameters. Agreements between measured and simulated data during calibration are comparable with previous studies, with root mean square errors (RMSEs) of simulated daily gross primary productivity (GPP), ecosystem respiration (RECO) and evapotranspiration (ET) ranging from 1.73 to 2.19, 1.04 to 1.56 g C m(-2) day(-1) and 0.50 to 1.28 mm day(-1), respectively. For the calibration period, using a homoscedastic eddy covariance residual error model resulted in a better agreement between measured and modelled data than using a heteroscedastic residual error model. However, a model validation experiment showed that CARAIB models calibrated considering heteroscedastic residual errors perform better. Posterior parameter distributions derived from using a heteroscedastic model of the residuals thus appear to be more robust. This is the case even though the classical linear heteroscedastic error model assumed herein did not fully remove heteroscedasticity of the GPP residuals. Despite the fact that the calibrated model is generally capable of fitting the data within measurement errors, systematic bias in the model simulations are observed. These are likely due to model inadequacies such as shortcomings in the photosynthesis modelling. Besides the residual error treatment, differences between model parameter posterior distributions among the four grassland sites are also investigated. It is shown that the marginal distributions of the specific leaf area and characteristic mortality time parameters can be explained by site-specific ecophysiological characteristics.
|
|
|
|
Semenov, M. A., & Stratonovitch, P. (2015). Adapting wheat ideotypes for climate change: accounting for uncertainties in CMIP5 climate projections. Clim. Res., 65, 123–139.
Abstract: This study describes integration of climate change projections from the Coupled Model Intercomparison Project Phase 5 (CMIP5) multi-model ensemble with the LARS-WG weather generator, which delivers an attractive option for the downscaling of large-scale climate projections from global climate models (GCMs) to local-scale climate scenarios for impact assessments. A subset of 18 GCMs from the CMIP5 ensemble and 2 Representative Concentration Pathways (RCPs), RCP4.5 and RCP8.5, were integrated with LARS-WG. For computationally demanding impact assessments, where it is not practical to explore all possible combinations of GCM x RCP, a climate sensitivity index could be used to select a subset of GCMs which preserves the range of uncertainty found in CMIP5. This would allow us to quantify uncertainty in predictions of impacts resulting fromthe CMIP5 ensemble by conducting fewer simulation experiments. In a case study, we describe the use of the Sirius wheat simulation model to design in silico wheat ideotypes that are optimised for future climates in Europe, sampling uncertainty in GCMs, emission scenarios, time periods and European locations with contrasting climates. Two contrasting GCMs were selected for the analysis, ‘hot’ HadGEM2-ES and ‘cool’ GISS-E2-R-CC. Despite large uncertainty in future climate projections, we were able to identify target traits for wheat improvement which may assist breeding for high-yielding wheat cultivars with increased yield stability.
|
|
|
|
Malone, R. W., Kersebaum, K. C., Kaspar, T. C., Ma, L., Jaynes, D. B., & Gillette, K. (2017). Winter rye as a cover crop reduces nitrate loss to subsurface drainage as simulated by HERMES. Agric. Water Manage., 184, 156–169.
Abstract: HERMES is a widely used agricultural system model; however, it has never been tested for simulating N loss to subsurface drainage. Here, we integrated a simple drain flbw component into HERMES. We then compared the predictions to four years of data (2002-2005) from central Iowa fields in corn-oybean with winter rye as a cover crop (CC) and without winter rye (NCC). We also compared the HERMES predictions to the more complex Root Zone Water Quality Model (RZWQM) predictions for the same dataset. The average annual observed and simulated N loss to drain flow were 43.8 and 44.4 kg N/ha (NCC) and 17.6 and 18.9 kg N/ha (CC). The slightly over predicted N loss for CC was because of over predicted nitrate concentration, which may be partly caused by slightly under predicted average annual rye shoot N (observed and simulated values were 47.8 and 46.0 kg N/ha). Also, recent research from the site suggests that the soil field capacity may be greater in CC while we used the same soil parameters for both treatments. A local sensitivity analysis suggests that increased field capacity affects HERMES simulations, which includes reduced drain flow nitrate concentrations, increased denitrification, and reduced drain flow volume. HERMES-simulated cumulative monthly drain flow and annual drain flow were reasonable compared to field data and HERMES performance was comparable to other published drainage model tests. Unlike the RZWQM simulations, however, the modified HERMES did riot accurately simulate the year to year variability in nitrate concentration difference between NCC and CC, possibly due in part to the lack of partial mixing and displacement of the soil solution. The results suggest that 1) the relatively simple model HERMES is a promising tool to estimate annual N loss to drain flow under corn-soybean rotations with winter rye as a cover crop and 2) soil field capacity is a critical parameter to investigate to more thoroughly understand and appropriately model denitrification and N losses to subsurface drainage. Published by Elsevier B.V.
|
|
|
|
Angulo, C., Rötter, R., Trnka, M., Pirttioja, N., Gaiser, T., Hlavinka, P., et al. (2013). Characteristic ‘fingerprints’ of crop model responses to weather input data at different spatial resolutions. European Journal of Agronomy, 49, 104–114.
Abstract: Crop growth simulation models are increasingly used for regionally assessing the effects of climate change and variability on crop yields. These models require spatially and temporally detailed, location-specific, environmental (weather and soil) and management data as inputs, which are often difficult to obtain consistently for larger regions. Aggregating the resolution of input data for crop model applications may increase the uncertainty of simulations to an extent that is not well understood. The present study aims to systematically analyse the effect of changes in the spatial resolution of weather input data on yields simulated by four crop models (LINTUL-SLIM, DSSAT-CSM, EPIC and WOFOST) which were utilized to test possible interactions between weather input data resolution and specific modelling approaches representing different degrees of complexity. The models were applied to simulate grain yield of spring barley in Finland for 12 years between 1994 and 2005 considering five spatial resolutions of daily weather data: weather station (point) and grid-based interpolated data at resolutions of 10 km x 10 km; 20 km x 20 km; 50 km x 50 km and 100 km x 100 km. Our results show that the differences between models were larger than the effect of the chosen spatial resolution of weather data for the considered years and region. When displaying model results graphically, each model exhibits a characteristic ‘fingerprint’ of simulated yield frequency distributions. These characteristic distributions in response to the inter-annual weather variability were independent of the spatial resolution of weather input data. Using one model (LINTUL-SLIM), we analysed how the aggregation strategy, i.e. aggregating model input versus model output data, influences the simulated yield frequency distribution. Results show that aggregating weather data has a smaller effect on the yield distribution than aggregating simulated yields which causes a deformation of the model fingerprint. We conclude that changes in the spatial resolution of weather input data introduce less uncertainty to the simulations than the use of different crop models but that more evaluation is required for other regions with a higher spatial heterogeneity in weather conditions, and for other input data related to soil and crop management to substantiate our findings. Our results provide further evidence to support other studies stressing the importance of using not just one, but different crop models in climate assessment studies. (C) 2013 Elsevier B.V. All rights reserved.
|
|
|
|
Nendel, C., Wieland, R., Mirschel, W., Specka, X., Guddat, C., & Kersebaum, K. C. (2013). Simulating regional winter wheat yields using input data of different spatial resolution. Field Crops Research, 145, 67–77.
Abstract: The success of using agro-ecosystem models for the high-resolution simulation of agricultural yields for larger areas is often hampered by a lack of input data. We investigated the effect of different spatially resolved soil and weather data used as input for the MONICA model on its ability to reproduce winter wheat yields in the Federal State of Thuringia, Germany (16,172 km(2)). The combination of one representative soil and one weather station was insufficient to reproduce the observed mean yield of 6.66 +/- 0.87 t ha(-1) for the federal state. Use of a 100 m x 100 m grid of soil and relief information combined with just one representative weather station yielded a good estimator (7.01 +/- 1.47 t ha(-1)). The soil and relief data grid used in combination with weather information from 14 weather stations in a nearest neighbour approach produced even better results (6.60 +/- 1.37 t ha(-1)); the same grid used with 39 additional rain gauges and an interpolation algorithm that included an altitude correction of temperature data slightly overpredicted the observed mean (7.36 +/- 1.17 t ha(-1)). It was concluded that the apparent success of the first two high-resolution approaches over the latter was based on two effects that cancelled each other out: the calibration of MONICA to match high-yield experimental data and the growth-defining and -limiting effect of weather data that is not representative for large parts of the region. At the county and farm level the MONICA model failed to reproduce the 1992-2010 time series of yields, which is partly explained by the fact that many growth-reducing factors were not considered in the model. (C) 2013 Elsevier B.V. All rights reserved.
|
|
