Dynamic Crop Models Research Articles

Simulation of maize yield in current and changed climatic conditions: Addressing modelling uncertainties and the importance of bias correction in climate model simulations

Appropriate knowledge and understanding of the impact of climatic variability on agricultural production is essential for devising an adaptation strategy. Climate change impact studies have to cope with the cascade of uncertainties that enter at different levels of modelling (e.g., emission scenario, climate model structure, impact assessment models). Our study aims at addressing these uncertainties through an ensemble probabilistic approach, which accounts for uncertainties in climate model simulations as well as parametric uncertainties in a dynamic crop model, when simulating maize ( Zea mays L.) growth and development. Simulations from eight regional climate models were used in combination with 10,000 different parameter sets from a dynamic crop model, reflecting biophysical uncertainties. Since regional climate model simulations can be subject to systematic biases, the use of such simulations to create impact assessment models can lead to unrealistic results. In the second phase of our study, we therefore determined the importance of bias correction of simulated meteorological variables prior to their use as input data in a dynamic crop model. The results revealed that using raw simulations from regional climate models to force a dynamic crop model produced unrealistic maize yield estimations, mainly because of underestimation of the intensity of daily precipitation. Corrected simulations from climate models significantly improved the quality of maize yield simulations, while a lower degree of improvement was observed in cases in which the frequency of wet days was underestimated in comparison to measured values. Using bias corrected climate model simulations in an ensemble probabilistic approach resulted in probability distributions of expected yield changes at three locations in Slovenia. Yield is expected to decrease on average between 10% and 16% in the 2050s and between 27% and 34% in the 2090s, while inter-annual variability is expected to increase compared to the control period between 1961 and 1990. Variance decomposition of the ensemble yield projections was performed in order to determine the RCM inter-model variability and crop model parameter uncertainty. The proportion of variance between RCMs increases during the 21st century, but never exceeds the inter-annual yield variability. Moreover, the parametric uncertainty of the WOFOST model can be regarded as negligible compared to RCM inter-model variability and yield inter-annual variability. A statistical emulator of the dynamic crop model was developed in order to analyze the impact on maize yield of weather variability within the growing season. It has been recognized that maize yield depends mostly on weather conditions during the period from 90 to 110 days after sowing, which coincides with the silking and tasseling period. High temperatures, low relative humidity and low rainfall during this period negatively affect maize growth, leading to a decrease in dry matter production. The analysis also revealed that precipitation during the growing season had a decisive impact on inter-annual yield variability at the selected locations.

Modelling climate impacts on crop yields in Belgium

CR Climate Research Contact the journal Facebook Twitter RSS Mailing List Subscribe to our mailing list via Mailchimp HomeLatest VolumeAbout the JournalEditorsSpecials CR 44:55-68 (2010) - DOI: https://doi.org/10.3354/cr00925 Modelling climate impacts on crop yields in Belgium A. Gobin* Environmental Modelling Unit, Flemish Institute for Technological Research, Boeretang 200, 2400 Mol, Belgium *Email: anne.gobin@vito.be ABSTRACT: In the last 2 decades, Belgium has experienced more monthly extremes than in any other decade since observations began in 1833. During the past 60 yr, yields have increased, on average, by 0.1 t ha–1 yr–1 for winter cereals, 0.4 t ha–1 yr–1 for potato and 0.6 t ha–1 yr–1 for sugar beet. A total of 60 to 74% of the variability in yields between 1960 and 2008 was explained by meteorological variables. Multivariate analysis showed significant statistical relationships between yield and vapour-pressure deficit, temperature and growing season length, and water logging and drought. These variables were therefore included in a new regional dynamic crop model (REGCROP) developed to assess climate impacts on regional arable crop production. Nash-Sutcliffe model efficiencies were between 0.68 and 0.84 between simulated and observed national yields for the period 1960–2008. REGCROP was subsequently run for 3 climate change scenarios and 3 typical Belgian soils (clay, loam and loamy sand) and climate impacts were compared with historical weather impacts (1960–1989). Higher temperatures increase crop development and shorten the growing season. Strong projected changes in seasonality affect cumulative drainage for both winter and summer crops and result in a drier water balance regime with climate change. Average yield losses of 12 to 27% were simulated for sugar beet and 23 to 44% for potatoes owing to drought and heat stress. Projected losses for winter cereals are 5 to 12%, mainly owing to waterlogging, whereas yield increases up to 6 or 7% are projected as a result of temperature increases and favourable vapour-pressure deficits. KEY WORDS: Arable crop yield · Climate change impact · Regional dynamic crop model · Time series · Multivariate analysis Full text in pdf format PreviousNextCite this article as: Gobin A (2010) Modelling climate impacts on crop yields in Belgium. Clim Res 44:55-68. https://doi.org/10.3354/cr00925 Export citation RSS - Facebook - Tweet - linkedIn Cited by Published in CR Vol. 44, No. 1. Online publication date: October 28, 2010 Print ISSN: 0936-577X; Online ISSN: 1616-1572 Copyright © 2010 Inter-Research.

The simulation of phenological development in dynamic crop model: The Bayesian comparison of different methods

The main objective of our study was to use Bayesian methods to quantify the uncertainties related to phenological development of maize ( Zea mays L.) under various climate conditions. For this purpose, five different phenological methods were implemented in the dynamic crop growth model, which was subsequently optimized, using the data acquired at three different locations in Slovenia. The sensitivity analysis of the crop model was performed in order to find the set of most influential physiological parameters. Subsequent Bayesian model comparison was used in order to quantify the impact of phenological method selection on the final maize yield. The results revealed the importance of using an appropriate phenological method in order to correctly estimate the duration of the growing season and yield, when used within dynamic crop model. The limitations of the phenological methods used in this study are discussed. The selection of phenological method itself did not have a significant influence on the yield estimation, except in years with high temperatures and limiting water conditions. This raises the concern that inaccurate simulation of phenological development may increase the uncertainties of impact assessment on crop yield where crop models are fed with future climate projections.

Regional simulation of maize production in tropical savanna fallow systems as affected by fallow availability

Upscaling of crop models from the field scale to the national or global scale is being used as a widespread method to make large-scale assessments of global change impacts on crop yields and agricultural production. In spite of the fact that soil fertility restoration and crop performance in many developing countries with low-input agriculture rely strongly on fallow duration and management, there are only few approaches which take into account the effect of fallowing on crop yields at the regional scale. The objectives of this study were to evaluate the sensitivity of maize yield simulations with the Environmental Policy Integrated Climate (EPIC) model to fallow availability at the field and regional scale and (2) to present a novel approach to derive a model-based estimate of the average fallow availability within a typical catchment of the sub-humid savanna zone of West Africa. Therefore, the EPIC model has been validated at the field scale and then incorporated into a spatial database covering a typical catchment within the sub-humid savanna zone of West Africa with 121 sub-basins. Maize-fallow rotations have been simulated within 2556 quasi-homogenous spatial units and then aggregated to the 10 districts within the catchment assuming three different scenarios of fallow availability: 100% of the bush-grass savanna area is available and used in fallow-crop rotations ( FU100), 50% of the bush-grass savanna area is available and used in fallow-crop rotations ( FU50) and 25% of the bush-grass savanna area is available and used in fallow-crop rotations ( FU25). A new aggregation procedure has been developed which is based on changes in the frequency of fallow-cropland classes within the sub-basins to render the simulation results in the spatial database sensitive to changes in fallow availability. Comparison of the average simulated grain yield with the mean yield over the catchment shows that the simulations overestimate maize yields by 62%, 44% and 15% for scenario FU100, FU50 and FU25, respectively. The best agreement between simulated and observed crop yields at the district scale was found when using the assumption that 25% of the savanna is available as fallow land under the present cropping patterns, which corresponds to a fallow-cropland ratio of 0.9. Comparison with farm surveys shows that the combination of remote sensing and dynamic crop modelling with yield observations provides realistic estimates of effective fallow use at the regional scale.

Assessing the vulnerability of Indian mustard to climate change

Rapeseed-mustard ( Brassica spp.) is a major group of oilseed crop in the world with India being the second largest cultivator after China. Although there has been a significant increase in oilseed production since 1960s, the demand for oilseeds production in the future is likely to go up due to population increase and their income. Mustard is much sensitive to climatic variables and hence climate could have significant effect on its production. There are very limited studies to assess the impact of climate change on oilseed crops as compared to cereals. This paper presents results of a simulation study to evaluate the impact of projected climate change on Indian mustard ( Brassica juncea) in contrasting agro-environments of the tropics. InfoCrop, a generic dynamic crop model, provides integrated assessment of the effect of weather, variety, pests and soil management practices on crop growth and yield, as well as on soil nitrogen and organic carbon dynamics in aerobic, anaerobic conditions and also greenhouse gas emissions. The validated model (InfoCrop-mustard) has reasonably predicted phenology, crop growth and yield of mustard crop. The crop was found to be sensitive to changes in carbon dioxide (CO 2) and temperature. Future climate change scenario analysis showed that mustard yields are likely to reduce in both irrigated and rainfed conditions. However, these reductions have spatial variation in different mustard growing region of India. In both irrigated and rainfed conditions, yield reduction would be higher in eastern India (67 and 57%) followed by central India (48 and 14%) and northern India (40.3 and 21.4%). This was due to maximum temperature rise in eastern part of the country, projected for 2080. In northern India, yield reduction of irrigated mustard was comparatively less due to prevailing lower temperature in this region during the crop growth period. But rainfed crop was found to be more susceptible to changing climate in north India due to projected reduction in rainfall in future scenarios. Adoption of adaptation measures like late sowing and growing long-duration varieties would be helpful in preventing yield loss of irrigated mustard in different locations of the country.

Simulation de la réponse variétale du tournesol à l’environnement à l’aide du modèle SUNFLO

Variety assessment could be supported by the use of dynamic crop modelling. The SUNFLO model was developed to simulate the achene yield and oil concentration of sunflower crop with a special attention paid to the description of varietal diversity. For that purpose, a variety was characterized in the model by 12 parameters of phenology, leaf area development, allocation and response to water stress. These parameters were measured either in field conditions (dense stands) or in greenhouse pot experiments. In 2008, two variety trials were carried out by CETIOM in non limiting conditions and a greenhouse experiment was conducted by INRA to calibrate the response of leaf expansion and plant transpiration to soil water depletion. The model parameterized with these 3 experiments on 18 commercial varieties was evaluated for yield in 42 situations of the post-registration network conducted by CETIOM in France in 2008. The yield of a given variety in a given environment was simulated with a mean error of 5 q/ha (relative error = 16%). When averaging a variety over all the environments or an environment over all the varieties, the error was of 3.5 q/ha (relative error = 11%). The model could be used to rank environments (through sunflower crop response) in a variety assessment network and to separate varieties with sufficient phenotypic differences.

Crop yield responses to climate change in the Huang-Huai-Hai Plain of China

Global climate change may impact grain production as atmospheric conditions and water supply change, particularly intensive cropping, such as double wheat–maize systems. The effects of climate change on grain production of a winter wheat–summer maize cropping system were investigated, corresponding to the temperature rising 2 and 5 °C, precipitation increasing and decreasing by 15% and 30%, and atmospheric CO 2 enriching to 500 and 700 ppmv. The study focused on two typical counties in the Huang-Huai-Hai (3H) Plain (covering most of the North China Plain), Botou in the north and Huaiyuan in the south, considering irrigated and rain-fed conditions, respectively. Climate change scenarios, derived from available ensemble outputs from general circulation models and the historical trend from 1996 to 2004, were used as atmospheric forcing to a bio-geo-physically process-based dynamic crop model, Vegetation Interface Processes (VIP). VIP simulates full coupling between photosynthesis and stomatal conductance , and other energy and water transfer processes. The projected crop yields are significantly different from the baseline yield, with the minimum, mean (±standardized deviation, SD) and maximum changes being −46%, −10.3 ± 20.3%, and 49%, respectively. The overall yield reduction of −18.5 ± 22.8% for a 5 °C increase is significantly greater than −2.3 ± 13.2% for a 2 °C increase. The negative effect of temperature rise on crop yield is partially mitigated by CO 2 fertilization. The response of a C3 crop (wheat) to the temperature rise is significantly more sensitive to CO 2 fertilization and less negative than the response of C4 (maize), implying a challenge to the present double wheat–maize systems. Increased precipitation significantly mitigated the loss and increased the projected gain of crop yield. Conversely, decreased precipitation significantly exacerbated the loss and reduced the projected gain of crop yield. Irrigation helps to mitigate the decreased crop yield, but CO 2 enrichment blurs the role of irrigation. The crops in the wetter southern 3H Plain (Huaiyuan) are significantly more sensitive to climate change than crops in the drier north (Botou). Thus CO 2 fertilization effects might be greater under drier conditions. The study provides suggestions for climate change adaptation and sound water resources management in the 3H Plain.

IRRIGATE: A dynamic integrated model combining a knowledge-based model and mechanistic biophysical models for border irrigation management

Water management practices in southern France (the Crau plain) need to be modified in order to ensure greater water use efficiency and less environmental damage while maintaining hay production levels. Farmers, water managers and policy makers have expressed the need for new methods to deal with these issues. We developed the biodecisional model IRRIGATE to test new irrigation schedules, new designs for water channels or fields and new distribution planning for a given water resource. IRRIGATE simulates the operation of a hay cropping system irrigated by flood irrigation and includes three main features: (i) border irrigation with various durations of irrigation events and various spatial orders of water distribution, (ii) species-rich grasslands highly sensitive to water deficit, (iii) interactions between irrigation and mowing. It is based on existing knowledge, adapted models and new modules based on experiments and survey data. It includes a rule-based model on the farm scale, simulating dynamically both irrigation and mowing management, and two biophysical models. The two biophysical models are a dynamic crop model on the field scale simulating plant and soil behaviour in relation to water supply, and a flood irrigation model on the border scale simulating an irrigation event according to plant and hydraulic parameters. Model outputs allow environmental (water supply, drainage), social (labour) and agronomic (yields, water productivity and irrigation efficiency) analyses of the performance of the cropping system. IRRIGATE was developed using firstly a conceptual framework describing the system modelled as three sub-systems (biophysical, technical, and decision) interacting within the farm. Then a component-based spatially explicit modelling based on the identification of the interactions between modules, the identification of temporal and spatial scales of modules and the re-use of previous models was used to develop the numerical model. An example of the use of the biodecisional model is presented showing the effects on a real farm of a severe water shortage in 2006.

Multivariate global sensitivity analysis for dynamic crop models

Dynamic crop models are frequently used in ecology, agronomy and environmental sciences for simulating crop and environmental variables at a discrete time step. They often include a large number of parameters whose values are uncertain, and it is often impossible to estimate all these parameters accurately. A common practice consists in selecting a subset of parameters by global sensitivity analysis, estimating the selected parameters from data, and setting the others to some nominal values. For a discrete-time model, global sensitivity analyses can be applied sequentially at each simulation date. In the case of dynamic crop models, simulations are usually computed at a daily time step and the sequential implementation of global sensitivity analysis at each simulation date can result in several hundreds of sensitivity indices, with one index per parameter per simulation date. It is not easy to identify the most important parameters based on such a large number of values. In this paper, an alternative method called multivariate global sensitivity analysis was investigated. More precisely, the purposes of this paper are (i) to compare the sensitivity indices and associated parameter rankings computed by the sequential and the multivariate global sensitivity analyses, (ii) to assess the value of multivariate sensitivity analysis for selecting the model parameters to estimate from data. Sequential and multivariate sensitivity analyses were compared by using two dynamic models: a model simulating wheat biomass named WWDM and a model simulating N 2O gaseous emission in crop fields named CERES-EGC. N 2O measurements collected in several experimental plots were used to evaluate how parameter selection based on multivariate sensitivity analysis can improve the CERES-EGC predictions. The results showed that sequential and multivariate sensitivity analyses provide modellers with different types of information for models which exhibit a high variability of sensitivity index values over time. Conversely, when the parameter influence is quite constant over time, the two methods give more similar results. The results also showed that the estimation of the parameters with the highest sensitivity indices led to a strong reduction of the prediction errors of the model CERES-EGC.

Leaf transmittance measurements can improve predictions of the nitrogen status for winter wheat crop

Dynamic crop models can be used to predict the occurrence of nitrogen deficiency during crop growth and optimize nitrogen fertilisation. However, prediction errors can be large and may lead to wrong recommendations. The objective of our work is to study the value of correcting the dynamic model Azodyn using transmittance measurements made with the N-Tester ® (Yara) to predict the nitrogen status of a winter wheat crop. Our approach is to use a Bayesian method called the “interacting particle filter” to fit the model's state variables to measurements obtained over the course of the season. This approach was assessed on 44 experimental plots. Predictions of crop biomass, nitrogen uptake and nitrogen nutrition index were first performed for each plot by using the model without any correction. A second series of predictions was then performed for the same variables by correcting the model with N-Tester measurements at GS 7 on Feekes’ scale. The results showed that the second series of predictions were more accurate. Depending on the prediction dates, model corrections reduced the root mean squared error by 18.1–53.2% for nitrogen nutrition index, by 9.1–10.1% for biomass, and by 17.1–45.0% for nitrogen uptake. The predictions were improved up to 52 days after the measurement but the degree of improvement was higher when the prediction date was close to the measurement date. The results also showed that, when corrected, model predictions were very sensitive to values of N-Tester measurements. It is therefore necessary to use N-Tester measurements which are as precise as possible.

Sorghum yield prediction from seasonal rainfall forecasts in Burkina Faso

The high variability of rainfall, from interannual to multi-decadal time scales, has serious impacts on food security in the West African Sahel. At five locations in Burkina Faso, we explore the potential to improve model-based prediction of sorghum yields at a range of lead-times by incorporating seasonal rainfall forecasts. Analyses considered empirical and dynamic rainfall forecasts, two methods (regression and stochastic disaggregation) for linking rainfall forecasts with crop simulation, three levels of production technology and four forecast dates (15 May, June, July and August) based on predictors observed from the preceding month, for the period of available data (1957–1998). Accuracy of yield forecasts generally decreased with lead-time. Relative to forecasts based solely on monitored weather and historic climatology, incorporating rainfall forecasts resulted in modest improvements to yield forecasts made in May or June. The benefit from seasonal rainfall forecasts tended to increase with northern latitude. Statistical and dynamic rainfall forecast systems captured much of the multi-decadal variation apparent in historic rainfall and in yields simulated with observed rainfall. This multi-decadal component of rainfall variability accounts for a portion of the apparent predictability of sorghum yields. Correlation between point-scale crop yield simulations and district-scale production statistics (1984–1998) was weakly positive late in the season, and suggest that a dynamic crop model (SARRA-H) has potential to contribute to regional yield prediction beyond what the best linear regression can provide from seasonal rainfall or its predictors. We discuss avenues for further improving crop yield forecasts during the growing season.

Potential predictability of crop yield using an ensemble climate forecast by a regional circulation model

Global/Regional Circulation Models (GCM/RCM) predict the interannual climate variability better than the absolute values of meteorological variables. Statistical bias-correction methods increase the quality of daily model predictions of incoming solar radiation, maximum and minimum temperatures and rainfall frequency and amount. However, when bias-corrected forecasts/hindcasts are used by dynamic crop models, timing of dry-spell occurrences generate the largest uncertainty during the linking process. In this study, we used 20 ensemble members of an 18-year period provided by the Florida State University/Center for Ocean-Atmospheric Prediction Studies (FSU/COAPS) regional spectral model coupled to the National Center for Atmospheric Research Community Land Model (CLM2). The daily seasonal-climate hindcast was bias-corrected and used as input to the CERES-Maize model, thus producing 20 crop yield ensemble members. Using observed weather data for the same period, a time series of simulated crop yields was produced. Finally, principal component (PC) regression analysis was used to predict this time series using the crop yield ensemble members as predictors. Between 13.7 and 28.8% of the simulated corn yield interannual variability was explained using only one principal component ( p < 0.05), and estimated yields were in the correct tercile by margins of 16.7 to 38.2% beyond chance. Predictability of simulated corn yields using principal components was improved relative to the use of bias-corrected daily hindcasts. Bias-correcting all meteorological variables used by the crop model increased predictability skills compared with use of raw hindcasts, individual bias-correction of rainfall, and climatological values.

Is it useful to combine measurements taken during the growing season with a dynamic model to predict the nitrogen status of winter wheat?

Dynamic crop models can evaluate the effects of weather, soil and farming practices and the interactions between these three sources of variability on production and the environment. However, because of uncertainty about the system equations, the parameters and the input variables of these models, their predictions are imprecise. The objective of this article is to evaluate the usefulness of measurements obtained during the growing season to improve the precision of the values of the nitrogen nutrition index predicted by a dynamic model of winter wheat crop. A Bayesian method, called “interacting particle filter” was used to correct three state variables of the model from biomass and nitrogen uptake measurements obtained between the end of winter and flowering. The value of this approach was assessed using 16 experiments carried out between 1995 and 2002. The results show that when they were used alone, the biomass measurements did not reduce the prediction errors of the model. On the other hand, the use of measurements of nitrogen uptake alone or in combination with the biomass measurements reduced the values of root mean square error by 32% in predicting the nitrogen nutrition index. The correction of the model at a single date appeared to be sufficient to improve the quality of the predictions so long as this date was late and close to the prediction date. The results also show that the use of less precise measurements than those made on our experiments would still be worthwhile in practice provided that the measurement error does not exceed 15%.

Application of an interacting particle filter to improve nitrogen nutrition index predictions for winter wheat

Dynamic crop models predict several state variables at a daily time step and thus provide useful information for optimizing agricultural techniques. But the prediction errors of these models are often large due to uncertainties in parameters, initial state values, and equations. Monte Carlo sequential methods, like the interacting particle filter [Del Moral, 1996. Nonlinear filtering: interacting particle solution. Markov Process. Relat. Fields 2, 555–580], can be used to update the state variable values predicted by nonlinear dynamic models from a set of measurements and thus reduce the prediction errors. An interesting feature of these methods is that they do not require a linearization of the original nonlinear model. Up to now, these methods have never been applied to complex dynamic crop models. In this paper, the interacting particle filter was used to update the Azodyn model, a dynamic winter wheat crop model, at 10 or 11 dates, from biomass and nitrogen uptake measurements, and to predict a variable of practical interest, the nitrogen nutrition index. We showed that the implementation of this method can reduce the root mean squared error by 66.7–79.7% for the nitrogen nutrition index, but that the filter is highly sensitive to the assumptions made about the probability distribution of the model errors. We also showed that the particle filter gives stable results with 10,000 Monte Carlo simulations and that this number of simulations can be performed in a very reasonable calculation time.

InfoCrop: A dynamic simulation model for the assessment of crop yields, losses due to pests, and environmental impact of agro-ecosystems in tropical environments. I. Model description

The problems of agriculture in many tropical countries are gradually becoming more intense due to increasing food demand led by population growth, stagnation in farm productivity, mounting yield losses due to multiple pests, increasing vulnerability to global environmental changes and the need to reduce emission of greenhouse gases. Tools and techniques are needed to assist in developing strategies that can lead to higher food production, prevent crop production losses, and ensure minimal greenhouse gas emissions while maintaining soil fertility. Several dynamic models have been developed in recent past but most of these are generally strong either in soils and crops, or in greenhouse gases (GHG) emissions. Pest induced yield losses, a critical issue in the tropics, is not addressed in most models. InfoCrop, a generic dynamic crop model, has been developed to meet these specific requirements. It provides integrated assessment of the effect of weather, variety, pests, soil and management practices on crop growth and yield, as well as on soil nitrogen and organic carbon dynamics in aerobic as well as anaerobic conditions, and greenhouse gas emissions. The model considers the key processes related to crop growth, effects of water deficit, flooding, nitrogen management, temperature and frost stresses, crop–pest interactions, soil water and nitrogen balance and (soil) organic carbon dynamics. Its general structure relating to basic crop growth and yield is largely based on several earlier models, especially SUCROS series, and is written in Fortran Simulation Environment (FSE) programming language. The model has been validated for dry matter and grain yields of several annual crops, losses due to multiple diseases and pests, and emissions of carbon dioxide, methane and nitrous oxide in a variety of agro-environments. To increase the applications of model in research and development, an extremely simple menu driven version of InfoCrop has also been developed. The users of this version do not need any background in programming.

Agrotool — A system for crop simulation

A system for simulation of crop dynamics is described. It includes a dynamic crop model for several cultures and a model shell. The model permits to simulate crop dynamics on the second level of plant productivity. The model shell serves for organization and implementation of series of computer experiments in dialogue regime. As it is known, soil and climatic conditions in the majority of Russia's regions differ from that in West Europe in a principal way. It especially concerns the continental Russia regions where deep droughts regularly occur, or Northwest regions, which are subjected to excess of water. So a special attention in model development was paid on account of the influence of water stress on plant production process.

Increased crop damage in the US from excess precipitation under climate change

Recent flooding and heavy precipitation events in the US and worldwide have caused great damage to crop production. If the frequency of these weather extremes were to increase in the near future, as recent trends for the US indicate and as projected by global climate models (e.g., US National Assessment, Overview Report, 2001, The Potential Consequences of Climate Variability and Change, National Assesment Synthesis Team, US Global Change Research Program, Washington, DC; Houghton et al., 2001, IPCC Climate Change 2001: The Scientific Basis, Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, 335pp.), the cost of crop losses in the coming decades could rise dramatically. Yet current assessments of the impacts of climate change on agriculture have not quantified the negative effects on crop production from increased heavy precipitation and flooding (Impacts of climate change and variability on agriculture, in: US National Assessment Foundation Document, 2001. National Assessment Synthesis Team, US Global Change Research Program, Washington DC.). In this work, we modify a dynamic crop model in order to simulate one important effect of heavy precipitation on crop growth, plant damage from excess soil moisture. We compute that US corn production losses due to this factor, already significant under current climate, may double during the next thirty years, causing additional damages totaling an estimated $3 billion per year. These costs may either be borne directly by those impacted or transferred to private or governmental insurance and disaster relief programs.

Interannual variability characteristics and simulated crop response of four stochastic weather generators

Many stochastic weather generators commonly used with crop models tend to under predict interannual variability of climate and, as a result, distort distributions of crop simulation results. We examine the ability of four stochastic weather generators, WeatherMan, MARKSIM, WM2 and LARS-WG, to reproduce interannual variability of monthly climate and crop simulation results. Comparisons were based on bias and RMSE of means and standard deviations of monthly precipitation totals, frequencies of wet days, mean daily temperature extremes for 12 long-term weather data sets; and yields and dates of anthesis and harvest maturity of three crops under a total of 10 scenarios simulated with dynamic crop models. Evaluation also considered statistical tests of equality of distributions of the same variables between observed and generated weather data sets. The generators generally reproduced climatic means well, except for MARKSIM which showed positive bias of mean monthly rainfall totals and frequencies. WM and LARS-WG showed substantial negative bias of interannual variability of monthly precipitation totals. As a result of stochastic resampling of wet-day frequencies, MARKSIM and WM2 showed little negative variability bias of monthly precipitation totals, but positive variability bias of monthly wet-day frequencies. WeatherMan, MARKSIM and LARS-WG under represented variability of monthly mean temperatures, whereas WM2 showed positive but smaller mean temperature variability bias. WM2 reproduced simulated distributions of yield and harvest maturity dates better than the other generators. Results for simulated anthesis dates were not consistent among generators. This study supports the need for some form of low-frequency variability correction in stochastic weather generators for applications in which reproducing interannual variability is important. Results generally favor WM2 over the other generators tested for applications in which variability of simulated yields is of primary importance.

Scaling-up crop models for climate variability applications

Although most dynamic crop models have been developed and tested for the scale of a homogeneous plot, applications related to climate variability are often at broader spatial scales that can incorporate considerable heterogeneity. This study reviews issues and approaches related to applying crop models at scales larger than the plot. Perfect aggregate prediction at larger scales requires perfect integration of a perfect model across the range of variability of perfect input data. Aggregation error results from imperfect integration of heterogeneous inputs, and includes distortions of either spatial mean values of predictions or year-to-year variability of the spatial means. Approaches for reducing aggregation error include sampling input variability in geographic or probability space, and calibration of model inputs or outputs. Implications of scale and spatial interactions for model structure and complexity are a matter of ongoing debate. Large-scale crop model applications must address limitations of soil, weather and management data. Distortion of weather sequences from spatial averaging is a particular danger. A case study of soybean in the state of Georgia, USA, illustrates several crop model scaling approaches.

ESTIMATING PARAMETERS FOR SOIL WATER BALANCE MODELS USING ADAPTIVE SIMULATED ANNEALING

A desirable feature of crop simulation models is for the same model to be applicable in any region if one provides required soil, weather, and crop input information. However, this is not an easy task. Each model requires a number of parameters that have to be measured or estimated. Moreover, many existing crop models have empirical relationships that represent important soil and crop processes, such as root growth and soil water extraction. Thus, objective methods are needed for estimating parameters for such crop models because they may not be known or easily estimated from readily available field or laboratory data. An objective procedure based on the Adaptive Simulated Annealing (ASA) optimization technique has been developed to estimate soil and root growth parameters for both the “current” (CROPGRO-Soybean v3.5) and “revised” models. We used this procedure to estimate a soil impedance factor (SIF) and a root hospitality factor (RHFAC) in the “revised” model; and the root weighting function (WR) in the “current” model. This procedure uses field measurements of volumetric soil water content at different depth increments over time during a growing season. It estimates parameters that minimize the error sum of squares between observed and simulated values. Data sets from East Campus, Nebraska (Specht et al., 1986), Gainesville, Florida (Hammond et al., 1978, unpublished), and Castana, Iowa (Mason et al., 1980), were used to evaluate the ASA optimization program and to compare the performance of the two models. The results demonstrated that the ASA program was able to fit the predicted soil water in the “revised” model. Except for the Nebraska experiment, the “current” model with the optimized WR also fit soil water content over time. Although the two models predicted reasonably well final grain yield for the Florida and Iowa experiments, the “current” model under predicted grain yield by more than 40% for the Nebraska location. The absolute percent errors in soil water contents estimated for the Nebraska experiment using the “revised” and “current” models were less than 5% and 8%, respectively. The global Adaptive Simulated Annealing (ASA) optimization method can be used successfully for estimating root growth and soil water extraction parameters for dynamic crop models.