Agricultural Production Systems Simulator Research Articles

Seasonal climate forecast-an important tool in managing the risk of extreme weather events in Australia's wheat industry

This study aims to investigate the benefits of using the Australian Community Climate and Earth-System Simulator-Seasonal Version 2 (ACCESS-S2), a state-of-the-art seasonal climate forecast (SCF) system, to improve contingent decision-making in the Australian wheat industry. Six locations in eastern Australia, three cultivars with varying maturities, and 17 times of sowing (TOS) were considered. Seasonal hindcasts from ACCESS-S2, initialized on the 1st May for the period 1981–2018 were linked with the Agricultural Production System sIMulator (APSIM)-Wheat model (v7.10) to (1) assess the hindcast skill of ACCESS-S2 in predicting the occurrence of extreme weather events and in predicting above/below long-term median wheat yield; (2) identify optimal management strategies in mitigating the impact of extreme weather events; and (3) quantify the yield benefit of using ACCESS-S2 at the key decision-making point of sowing against a baseline management situation (i.e., without the use of SCF). Research results showed that (1) ACCESS-S2 has varying skills in predicting extreme climate indices and wheat yield across the six locations at a seasonal lead time; (2) 68 % of cases had a yield gain using SCF information; (3) across all cases, there was an average yield gain of 281 kg/ha representing an increase of 13 %; (4) benefits of using SCF were seen across predicted wet years, neutral years and dry years accounting for 69 %, 65 % and 72 % respectively, and (5) the identified optimal management options varied from location to location with earlier sowing associated with either suntop or gregory or both being the optimal strategies at most of the locations considered. Overall, there is a demonstrated benefit in utilizing ACCESS-S2 forecasts in the Australian wheat industry to improve farm management decision-making. The benefit can occur in any climate state but with dry years being more likely and significant (with an average yield increase of 97 % across the locations).

Integrating Crop Modeling and Machine Learning for the Improved Prediction of Dryland Wheat Yield

One of the crucial research areas in agricultural decision-making processes is crop yield prediction. This study leverages the advantages of hybrid models to address the complex interplay of genetic, environmental, and management factors to achieve more accurate crop yield forecasts. Therefore, this study used the data of wheat growth environment, crop management, and historical yield in experimental fields in Anding District, Dingxi City, Gansu Province from 1984 to 2021 to construct eight machine learning models and ensemble models. Furthermore, Agricultural Production Systems sIMulator (APSIM), machine learning (ML), and APSIM combined with machine learning (APSIM-ML) were employed to predict wheat yields in 2012, 2016, and 2021. The results show that the APSIM-ML weighted ensemble prediction model, optimized to minimize the MSE, performed the best. Compared to the optimized ML and APSIM models, the average improvements in the RMSE, RRMSE, and MBE for the test years were 43.54 kg/ha, 3.55%, and 15.54 kg/ha, and 93.96 kg/ha, 7.55%, and 104.21 kg/ha, respectively. At the same time, we found that the dynamic flow of water and nitrogen between the soil and crops had the greatest impact on wheat yield prediction. This study improved the accuracy of dryland wheat yield prediction in Gansu Province and provides technical support for the intelligent production of dryland wheat in the loess hilly area.

Sustainable management strategies for balancing crop yield, water use efficiency and greenhouse gas emissions

CONTEXTTo ensure food security in the face of climate change, it is crucial to use effective agricultural management practices that produce the triangular win-win for water, crops, and environments. OBJECTIVEThis study aimed to explore the sustainable management measures for balancing crop yield, water use efficiency (WUE) and greenhouse gas (GHG) emissions in China during 1981–2020. METHODSThe study selected the Beijing-Tianjin-Hebei (BTH) region of China as the research area, where there were sharp contradictions among water shortage, food security and polluted environment. Four irrigation treatments and four fertilization treatments were set referring to the local cultivated experiences. The daily dynamic changes in crop yield, WUE and GHG emissions were simulated by optimized Agricultural Production Systems sIMulator (APSIM) model. RESULTS AND CONCLUSIONSThe irrigations had positive impacts on crop yield and WUE in BTH region. The irrigation amount's rise caused a gradual augmentation in the total yield of winter wheat and summer maize from 1981 to 2020.The highest total yield was achieved under W4 irrigation treatment (320 mm water for winter wheat and 60 mm water for summer maize), with an annual mean of 24,579.2 kg·ha−1. The annual average WUE under the W3 irrigation treatment (240 mm water for winter wheat) was the highest, with an average of 13.9 kg·ha−1·mm−1. However, extensive fertilization exacerbated soil GHG emissions. The soil GHG emissions were found to be significant increase as the input amount of fertilization. The FI fertilization measure (no fertilization) was assessed to have the least impact on the environment, with an average GHG emissions of 0.4 Mg·CO2-eq·ha−1. The optimal goals of high yield, high WUE and low GHG emissions were achieved by choosing effective combination practices of water and nitrogen management. Overall, the W3 irrigation and F3 fertilization (180 kg/ha fertilizer for winter wheat and 120 kg/ha fertilizer for summer maize.) treatments had the maximum water-fertilizer production indexes in the BTH region. SIGNIFICANCEThis study highlights the importance of agricultural management measures for guaranteeing food security and environmental sustainability at the same time under climate change.

Evaluating the Agricultural Production Systems sIMulator (APSIM) wheat module for California

Context Computer-based crop simulation models are important tools for agricultural research and management. APSIM (Agricultural Production Systems sIMulator) is commonly used around the world but has not been widely validated in North America. Aims The objective of this work was to evaluate the reliability of APSIM for simulating wheat production in California, with the aim of providing guidance for future field research aimed at model calibration and validation. Methods Environmental and management data from state-wide wheat variety trials of common wheat (Triticum aestivum L.) were used to parameterise the APSIM-Wheat module (ver. 7.10 r4220). Simulated yield and protein data were compared with observed field trial results to test the reliability of APSIM simulations. Key results The most reliable simulation of grain yield had a root-mean-square error of 1040 kg/ha and normalised root-mean-square error of 16% relative to actual field data. Preliminary calibration of the model for Californian wheat varieties did not improve simulation accuracy or precision. Conclusions The accuracy or precision of the simulations was comparable to that of other tests of the APSIM-Wheat module in environments where it has not been previously calibrated but was considered too low to be reliable. The lack of reliability was due to the poor representation of local Californian wheat genotypes by existing APSIM cultivars, as well as possible lack of precision and accuracy of field data. Implications APSIM could be a valuable tool for wheat research and management in California; however, further research is needed to generate suitable field data for model calibration and validation.

Modeling long-term nitrogen utilization under alfalfa–corn rotation in Northeast China

ContextNortheast China is the main commodity grain base in China owing to its intensive corn (Zea mays L.) planting and heavy nitrogen (N) application. Alfalfa (Medicago sativa L.) and corn rotation have the potential to decrease N application and enable efficient N utilization, but its long-term impacts remain unclear. ObjectiveThis study utilized Agricultural Production Systems sIMulator (APSIM) to evaluate the N process of long-term alfalfa–corn rotation, including the soil N status and N legacy benefits of subsequent corn crops, and to optimize alfalfa–corn rotation. MethodsAPSIM was calibrated and validated utilizing datasets of crop N uptake (Nupt), soil inorganic N, and soil total nitrogen (TN) under continuous corn, continuous alfalfa, and rotations of three, four, and five years of alfalfa followed by two years of corn, defined as CC, CA, and 3A2C, 4A2C, and 5A2C treatments, respectively, which were collected from a field experiment in Quanyangou–Lishu, Jilin Province, China during 2014–2022. The validated APSIM was then utilized to simulate long-term (1981–2020) effects on N utilization under CC, 3A2C, 4A2C, and 5A2C treatments at Quanyangou–Lishu, Fujia–Lishu, Chaoyang, and Hailun in Northeast China. ResultsAPSIM effectively simulated the Nupt, inorganic N, and TN, with normalized root mean square errors of 5%–19%, 20%–30%, and 2.0%–4.2%, respectively, between the observed and simulated values. Compared with the CC treatment, alfalfa–corn rotation treatments reduced the N fertilizer application by 143 kg N ha−1 and significantly improved the net N mineralization and Nupt by 41 and 107 kg N ha−1, respectively, achieving increases of 50% and 517% in N use efficiency and N partial factor productivity based on corn equivalent yield, respectively. Alfalfa–corn rotation also consumed 94% of 200–500 cm inorganic N through alfalfa roots absorbing N after 9–14 years, which decreased the risk of soil inorganic N accumulation and N leaching. After 40 years, SON values of alfalfa–corn rotation treatments were 8% and 4% higher in the 0–20 and 20–60 cm soil layers and 1.6% lower in the 60–200 cm soil layer than those of the CC treatment at Quanyangou–Lishu, Fujia–Lishu and Hailun. The SON values of alfalfa–corn rotation treatments were 2.9% lower in three soil layers at 0–200 cm deep relative to those of the CC treatment at Chaoyang, with low precipitation of only 478 mm. Alfalfa–corn rotation also improved the first-year corn yield following alfalfa by 2.3%–8.8% at Quanyangou–Lishu and Fujia–Lishu. ConclusionsThe 4A2C treatment at Quanyangou–Lishu, Fujia–Lishu, and Hailun and the 3A2C treatment at Chaoyang are recommended in Northeast China to increase N use efficiency and N partial factor productivity by decreasing N fertilizer application, increasing SON, and absorbing inorganic N from the 200–500 cm soil layer during the alfalfa growing period and increasing mineral N during the corn growing period. ImplicationsOptimal alfalfa–corn rotations show great potential to reduce soil inorganic N residue and increase N use efficiency and SON in Northeast China.

A framework for assessing the impacts of land-use/cover change and climate change on wheat productivity under 1.5 and 2.0 °C warming at watershed scale.

Irrigation is used extensively to enhance grain production and ensure food security. Many studies have used crop models and global climate models to study the variation of irrigated crop yield in the context of climate change. But most considered the influence of direct climate change but neglected the influence of irrigation water availability, which is affected by land-use/cover change (LUCC) and indirect climate change, on irrigated crop yield. This study therefore developed a framework including Patch-generating Land Use Simulation model, Soil and Water Assessment Tool, Agricultural Production Systems sImulator Model, and global climate models for exploring the impacts of LUCC, direct climate change, and indirect climate change on wheat yield in a typical watershed. Both LUCC and climate change caused increased runoff from October to May, and thus increased the irrigation water availability, by 51.6 and 30.7 mm per growing season under 1.5 and 2.0 °C warming, respectively. The combined influence of LUCC, direct, and indirect climate change increased wheat yield by about 18.5% and 15.5% in the context of 1.5 and 2.0 °C warming, respectively. The relative contribution of LUCC, indirect climate change and direct climate change to yield was 4.7%, 41.2%, and 54.1% under 1.5 °C warming, and 13.1%, 28.7%, and 58.2% under 2.0 °C warming, respectively. We suggest that changes in irrigation water availability should be considered from a watershed perspective when simulating the influence of climate change on crop yield, especially regional crop production estimation. © 2023 Society of Chemical Industry.

Unveiling the resilience of smallholder farmers in Senegal amidst extreme climate conditions

AbstractIn Senegal, agriculture is an important sector underpinning the socioeconomic fabric of the populace. Notably, the agricultural production in this region exhibits heightened sensitivity to climatic perturbations, particularly droughts and heat waves. This study aims to determine the resilience of different agronomic interventions for farmers practicing mixed farming that produce both crops (i.e., groundnut (Arachis hypogaea L.) and pearl millet (Pennisetum glaucum (L.) R. Br.)) and raise animals in the Groundnut Basin in Senegal, which holds historical and socioeconomic significance. To understand the current situation regarding demographics, economics, consumption behavior, and farm operations for smallholder farmers, data were comprehensively collected from government and nongovernment organizations (NGO) reports, scientific papers, organization databases, and surveys. Additionally, the Agricultural Production Systems sIMulator (APSIM) was used to understand how combinations of three planting dates, three plant densities, and six urea nitrogen (N) fertilizer rates affected the yield of pearl millet, which were used as the alternative scenarios to the baseline in the farm modeling and analyses. All the collected and generated data were used as inputs into the Farm Simulation Model (FARMSIM) to generate economic, nutritional, and risk data associated with mixed farming systems. The generated data were then used to determine the resilience of the alternative scenarios against the baseline. Initially, a multi‐objective optimization was employed to meet nutritional needs while maintaining a healthy diet at the lowest cost. Then, the scenarios that met the population's nutritional requirements were evaluated based on four economic indicators: net cash farm income (NCFI), ending cash reserves (EC), net present value (NPV), and internal rate of return (IRR). Lastly, those that passed the economic feasibility test were ranked based on risk criteria certainty equivalent (CE) and risk premium (RP). The analyses found N fertilizer rates of 0, 20, and 100 kg N ha−1 were generally economically not feasible. Additionally, medium (early‐July to late‐August) and late (late‐July to mid‐September) planting dates generally performed better than early (early‐June to late‐July) planting dates, while plant densities of 3.3 and 6.6 pL m−2 performed better than 1.1. The robust resilience approach introduced in this study is easily transferable to other regions.

Effects of climate factors on wheat and maize under different crop rotations and irrigation strategies in the North China Plain

The North China Plain (NCP) is a crucial agricultural region for grain production in China, primarily focusing on wheat and maize cultivation. However, these crop yields are highly vulnerable to fluctuations in temperature and precipitation, with climate change being a significant factor. This study investigates the impact of climate factors on wheat and maize yields in the NCP under various crop rotations and water supply strategies. Using the Agricultural Production Systems sIMulator crop mechanism model, we evaluated the effects of different crop rotations and water supply strategies on wheat and maize yields. A comprehensive analysis of the simulated results determined the yield variation trends and their correlation and sensitivity to different climate factors. The findings revealed that precipitation levels over the past 40 years showed no significant trend, while radiation levels showed a significant decreasing trend, and annual mean maximum and minimum temperatures displayed significant increasing trends. Furthermore, the study found that irrigation practices and crop rotations substantially impact grain yield in the study area, with average yields ranging from 8105.5 kg ha−1 under rainfed conditions to 13 088.8 kg ha−1 under fully irrigated conditions. There was a statistically significant trend of increasing yields for fully irrigated Monocrop-Wheat and decreasing yields for fully irrigated Monocrop-Maize over the past 40 years. Sensitivity analysis also showed that rational rotation and irrigation can reduce grain yield sensitivity to climate change. In conclusion, the prudent use of rotation and irrigation can enhance food production resilience to climate change. These findings have significant implications for developing strategies to optimize crop yields and adapt to climate change in the NCP while considering trade-offs.

Component-wise iterative ensemble Kalman inversion for static Bayesian models with unknown measurement error covariance

The ensemble Kalman filter (EnKF) is a Monte Carlo approximation of the Kalman filter for high dimensional linear Gaussian state space models. EnKF methods have also been developed for parameter inference of static Bayesian models with a Gaussian likelihood, in a way that is analogous to likelihood tempering sequential Monte Carlo (SMC). These methods are commonly referred to as ensemble Kalman inversion (EKI). Unlike SMC, the inference from EKI is asymptotically biased if the likelihood is non-linear and/or non-Gaussian and if the priors are non-Gaussian. However, it is significantly faster to run. Currently, a large limitation of EKI methods is that the covariance of the measurement error is assumed to be fully known. We develop a new method, which we call component-wise iterative EKI (CW-IEKI), that allows elements of the covariance matrix to be inferred alongside the model parameters at negligible extra cost. This novel method is compared to SMC on a linear Gaussian example as well as four examples with non-linear dynamics (i.e. non-linear function of the model parameters). The non-linear examples include a set of population models applied to synthetic data, a model of nitrogen mineralisation in soil that is based on the Agricultural Production Systems Simulator, a model predicting seagrass decline due to stress from water temperature and light, and a model predicting coral calcification rates. On our examples, we find that CW-IEKI has relatively similar predictive performance to SMC, albeit with greater uncertainty, and it has a significantly faster run time.

Global sensitivity analysis of APSIM-wheat yield predictions to model parameters and inputs

The performance of cropping system models is often limited by uncertainties in model parameters and inputs. This study aims to explore how model yield prediction responds to these uncertainties under various environmental and management conditions. The Sobol’ method was used to investigate the sensitivity of the Agricultural Production Systems SIMulator (APSIM)-Wheat yield prediction to the following factors: air temperature (maximum and minimum), precipitation, initial soil nitrogen content, nitrogen fertilization amount, and soil hydraulic parameters. Eighteen scenarios were defined to consider a combination of three weather conditions (wet, normal, and dry growing season, based on the historical climatology of the Wimmera district of Victoria, Australia), three soil types (sandy, loamy, and clayey soils), and two nitrogen fertilization applications (50 and 100 kg N/ha). Eight thousand APSIM simulations were used to calculate the first order and total sensitivity indices for each scenario. The effects of weather, soil water characteristics and nitrogen availability were estimated by measuring their impacts on sensitivity indices. Our results show that yield was more sensitive either to variables that control water availability (precipitation and soil type) or variables that control nitrogen availability, depending on which was more limiting to wheat growth under the scenario. For example, in case of low nitrogen fertilization scenario, the initial nitrogen content sensitivity ranked first. Variation of this factor contributed 64% to 79% to the variance in simulated yield for clayey soils, nearly four times higher than the second-ranked factor (nitrogen fertilization amount). Soil hydraulic parameters and precipitation were most important when the crop growth was more constrained by water availability than by nitrogen availability. In the case of sandy soils in dry years with high nitrogen fertilization level, soil parameters and precipitation accounted for 83% of the yield variability. Minimum temperature consistently ranked as the least important factor under all scenarios. This work will help researchers better understand the model inputs’ impacts on the simulated yield variability under various soil, management, and weather conditions. The results also support agronomists and practitioners in the optimal use of the APSIM-Wheat model as a management tool.

I-RAT: A discussion support system to rapidly assess economic and environmental impacts of different sugarcane irrigation practices

Water pollution and climate change are among the greatest threats to the iconic Great Barrier Reef (GBR). To improve the GBR’s long-term outlook, making improvements on these fronts is required. Owing to the complex interactions between soils, climate, and farm management, a tool is needed to guide agricultural land managers about which management practice changes are more likely to deliver improved water quality and climate resilience outcomes whilst maintaining profitability. Using concepts drawn from social studies of science and technology, a ‘discussion’ support system (DSS) named Irrigation Rapid Assessment Tool (i-RAT) was developed through participatory processes designed to enhance co-learning from the development of this DSS. i-RAT is a rapid assessment-visualisation tool in an interactive web application (https://i-rat.net) to visualise the impacts of changed irrigation practices for sugarcane farmers and extension staff. Specifically, i-RAT compares these impacts regarding farm economics, water quality, productivity, carbon and nitrogen cycles, and greenhouse gas emission. The core of i-RAT is the sugarcane module from the Agricultural Production Systems sIMulator (APSIM) modelling platform. i-RAT was first developed for the Burdekin sugarcane growing region in Queensland (Australia’s largest sugarcane producing region). Various soil types and management scenarios representing farming practices in the Burdekin region were used for simulations (1971–2021) to generate a datacube. In this paper, we describe how fundamental learnings about more (drought) resilient farming systems and more sustainable irrigation practices can be extracted from i-RAT. Details of the participatory approach with research and industry partners that informed the design and function of i-RAT and how APSIM was set up and parameterised are described. A ‘what-if’ analysis demonstrated the i-RAT features and application.

Climate Change and Soil Dynamics: A Crop Modelling Approach

The impact of global climate change is a challenge to the sustainability of many ecosystems, including soil systems. However, the performance of soil properties under future climate was rarely assessed. Therefore, this study was carried out to evaluate selected soil processes under climate change using an agri-environmental modeling approach to Sri Lanka. The Agricultural Production Systems Simulator (APSIM) model was used to simulate soil and plant-related processes using recent past (1990–2019) and future (2041–2070) climates. Future climate data were obtained for a regional climate model (RCM) under representative concentrations pathway 4.5 scenarios. Rainfalls are going to be decreased in all the tested locations under future climate scenarios while the maximum temperature showcased rises. According to simulated results, the average yield reduction under climate change was 7.4%. The simulated nitrogen content in the storage organs of paddy declined in the locations (by 6.4–25.5%) as a reason for climate change. In general, extractable soil water relative to the permanent wilting point (total available water), infiltration, and biomass carbon lost to the atmosphere decreased while soil temperature increased in the future climate. This modeling approach provides a primary-level prediction of soil dynamics under climate change, which needs to be tested using fieldwork.

Genotype × environment × agronomic management interaction to enhance wheat yield in the Mediterranean rainfed environments of Morocco: II. Process based modeling

Durum wheat (Triticum turgidum subsp. durum) is the oldest and most cultivated cereal crop in Middle East and North Africa (MENA) region and under Mediterranean climatic conditions. Morocco is one of the largest producer of durum wheat in MENA region, cultivated in more than 1 million ha area produced 2.5 million tons in 2020, which accounts for 17% of the total production in the region. In the region, rainfed production system is predominant, and with declining rainfall amounts with high variability, increasing water scarcity, and suboptimal input application, its productivity growth is low and needs to be increased to fulfill the growing demand. Developing context-specific management advisory is needed to improve productivity and resilience under such variable rainfed production environments. Agricultural Production Systems sIMulator (APSIM) model was calibrated and evaluated using four years (2015–2019) of on-station experimental data from genotype × seeding time × water management experiment conducted at International Center for Agricultural Research in the Dry Areas (ICARDA) research station, Morocco. Long-term (1984–2021) simulation was carried out to determine the contribution of Genotype × Environment × Management components for sustainably improving crop productivity. The results showed rainfall or supplementary irrigation (23–36%) followed by N fertilizer (28–38%), cultivar (9–14%), and seeding date (7–14%) have the largest contribution to the yield variance of durum wheat in Merchouch, Meknes, and Sidi El Aidi regions of Morocco. Under rainfed conditions, wheat yield was highest in Merchouch (4.5 t ha−1) and lowest (1.8 t ha−1) in Sidi El Aidi. Due to significant rainfall variability, the seeding date varies across year and location; however, generally, it is between 2nd week of November to 1st week of December. Under rainfed conditions, seeding after 1st week of December caused the average yield reduction of 120, 81, and 31 kg ha−1 d−1 in Merchouch, Meknes, and Sidi El Aidi, respectively. In all locations, short-duration varieties provided higher averaged yields with better resilience than medium and long-duration varieties. Decomposing yield variance caused by Genotype × Environment × Management provides the opportunity for risk reduction, improvement of wheat yield and resilience, and designing climate-smart adaptation strategies in rainfed Mediterranean conditions. Our findings highlight one-size-fits-all approach is inadequate and context-specific tailored agronomic practices and suitable genotypes is crucial for achieving sustainability and resilience of wheat production in variable climatic condition in Morocco and similar production environment.

Modeling the effect of soil fertility management options on maize yield stability under variable climate in a sub-humid zone in Ghana

IntroductionUncertainty in the yield of maize due to variability in weather is a major challenge to smallholder farmers in Sub Sahara Africa. This study explores the potential of combining locally available organic resources and inorganic fertilizer to increase grain yield and reduce variability in yields associated with variations in rainfall distribution.MethodsTo assess the effectiveness of this practice, the Agricultural Production Systems sIMulator (APSIM) crop model was calibrated and evaluated using maize experiments on nutrient management options. The evaluated model was used to simulate maize growth and yield using multiple-year data (1984–2018) under different planting windows for two growing seasons. The treatments were (i) control, (ii) inorganic fertilizer alone (INOFRecom), combining organic resources [empty fruit bunch of palm (EFB) and compost with inorganic fertilizer (INOF) to make up equivalent nutrients in (ii)]; (iii) EFB + INOF and (iv) Comp + INOF.ResultsThough all the soil amendments boosted grain output, the EFB + INOF treatment outperformed the other treatments in the major season with gains of between 161 and 211% and the most stable (least inter-annual variability of 27%) yield. Across the planting windows, the INOFRecom and EFB + INOF treatments achieved comparable yield increments in the minor season. Though grain yield variability was high during the minor rainy season, combining organic and inorganic fertilizers reduced inter-annual yield variability, thus, lowering uncertainty in yield due to variable inter-annual rainfall. Combining local organic resources with a reduced amount of inorganic fertilizer produced higher yields and better yield stability compared to using only inorganic fertilizer.DiscussionsThus, such soil fertility management solutions might sustain resource use and boost maize grain yield in the study area, where strategies for sustainable crop nutrition remain a critical necessity. The improved nitrogen management regimes may result in fewer environmental hazards for vulnerable rainfed agricultural systems.

Predicting maize yield in Northeast China by a hybrid approach combining biophysical modelling and machine learning

ContextChina produces more than 20 % of maize grain in the world, and Northeast China (NEC) accounts for ∼30 % of the nation's total maize production. Previous studies have used either climate data, satellite data, or crop growth model (CGM) to predict or forecast maize yield. However, maize is highly susceptible to the effect of extreme climate events (such as drought, heat) in NEC, and there is a lack of studies to predict/forecast maize yield by integrating climate data, satellite data, extreme climate events, and CGM-simulated data. ObjectiveWe aim to develop a hybrid approach with machine learning to blend different sources of data (climate data, satellite data, extreme climate events) and process-based modelling results to improve predictive accuracy of maize yield in NEC. MethodsUsing maize data from 44 sites during the period of 2000–2013 in NEC, we firstly optimized Agricultural Production System sIMulator (APSIM) using Differential Evolution Adaptive Metropolis combined with Gaussian likelihood function and Bayesian multiplication method. Next, we divided the growing season into five phases, and selected variables of different phases using exploratory data analysis and Random Forest. Then, we developed a hybrid model using Random Forest by blending of multiple sources of data and APSIM simulations to predict maize yield from the start to the end of the growing season, and quantified the relative contribution of predictors. Results and conclusionsA hybrid model developed with random forest by combining climate data, NDVI, extreme climate events and APSIM simulations can achieve high performance for predicting yield toward the end of the growing season. The accuracy of in-season yield prediction showed a linear increase with MAPE/KGE changing from 19 %/0.05 to 13 %/0.53 from start to end of the growing season. Yield forecasts are acceptable with RMSE/MAE of 1.20/1.01 Mg ha−1 (16 %/13 % of the observed mean yield) approximately one-month prior to harvest. The most important predictor that affect yield forecast was APSIM-simulated biomass or yield, and the most important extreme climate event was drought during early grain-filling stage. SignificanceWith the increasing availability of crop-related data, we expect that the in-season forecasting capacity of the proposed methodology could be further improved, and the methodology can be extended to other crops and other regions for yield forecast.

Investigating the effects of APSIM model configuration on model outputs across different environments

IntroductionSoil type plays a major role in nutrient dynamics and soil water which impacts crop growth and yield. The influence of soil characteristics on crop growth is usually evaluated through field experimentation (in the short term) and through crop-soil modelling (in the long-term). However, there has been limited research which has looked at the effect of model structural uncertainty of model outputs in different soil types.MethodsTo analyze the impact of soil inputs on model structural uncertainty, we developed eight model structures (a combination of two crop models, two soil water models and two irrigation models) within the Agricultural Production Systems sIMulator (APSIM) across three soil types (Ferralsols, Alisols and Chernozems). By decomposing the mean proportion of variance and simulated values of the model outputs (yield, irrigation, drainage, nitrogen leaching and partial gross margin) we identified the influence of soil type on the magnitude of model structural uncertainty.ResultsFor all soil types, crop model was the most significant source of structural uncertainty, contributing >60% to variability for most modelled variables, except irrigation demand which was dominated by the choice of irrigation model applied. Relative to first order interactions, there were minimal (<12%) contributions to uncertainty from the second order interactions (i.e., inter-model components). We found that a higher mean proportion of variance does not necessarily imply a high magnitude of uncertainty in actual values. Despite the significant impact of the choice of crop model on yield and PGM variance (contributing over 90%), the small standard deviations in simulated yield (ranging from 0.2 to 1 t ha-1) and PGM (ranging from 50.6 to 374.4 USD ha-1) compared to the mean values (yield: 14.6 t ha-1, PGM: 4901 USD ha-1) indicate relatively low actual uncertainty in the values. Similarly, the choice of irrigation model had a contribution of over 45% to variance, but the relatively small standard deviations ranging from 11 to 33.3 mm compared to the overall mean irrigation of 500 mm suggest low actual uncertainty in the values. In contrast, for the environmental variables- drainage and nitrogen leaching, the choice of crop model had contributions of more than 60% and 70% respectively, yet the relatively large standard deviations ranging from 7.1 to 30.6 mm and 0.6 to 7.7 kg ha-1 respectively, compared to the overall mean values of drainage (44.4 mm) and nitrogen leaching (3.2 kg ha-1), indicate significant actual uncertainty.DiscussionWe identified the need to include not only fractional variance of model uncertainty, but also magnitude of the contribution in measured units (e.g. t ha-1, mm, kg ha-1, USD ha-1) for crop model uncertainty assessments to provide more useful agronomic or policy decision-making information. The findings of this study highlight the sensitivity of agricultural models to the impacts of moisture availability, suggesting that it is important to give more attention to structural uncertainty when modelling dry/wet conditions depending on the output analyzed.

Short-term extreme heat at flowering amplifies the impacts of climate change on maize production

Extreme weather poses a threat to global crop production, food security and farmer livelihoods. High temperatures have been identified as detrimental to crop yields; however, how heat stress during the critical flowering stage will influence future maize (Zea mays L.) yields remains unclear. Here, we combined statistical and process-based models to assess impacts of short-term extreme heat at flowering on Chinese maize yield under climate change. We showed that heat around flowering has a stronger impact on yields than heat at other times in the growing season, especially for temperatures >30 °C. Heat stress during flowering was responsible for 23% of total yield loss from extreme degree days (EDDs) in 1990–2012. An improved process-based model (Agricultural Production Systems sIMulator (APSIM)-maize) incorporating a grain-temperature function was then applied and indicated that extreme heat at flowering amplified the impacts of climate change on maize production compared to the original model. The improved APSIM-maize predicted an 8.7% yield reduction across the Chinese Maize Belt as EDDs increased more than quadrupled at the end of the century (2070–2099) under a high emissions pathway (SSP585) in comparison with the baseline period (1990–2019). Our study highlights the importance of extreme heat at flowering on maize yield and can inform farmers and policy makers on adaptive measures as well as providing a reference for other crop areas facing similar challenges.

Effect of soil water on flowering and pod-set in chickpea: implications for modelling and managing frost and heat stress

Phenological development is critical for crop adaptation. Phenology models are typically driven by temperature and photoperiod, but chickpea phenology is also modulated by soil water, which is not captured in these models. This study is aimed at evaluating the hypotheses that accounting for soil water improves (i) the prediction of flowering, pod-set, and flowering-to-pod-set interval in chickpea and (ii) the computation of yield-reducing frost and heat events after flowering. To test these hypotheses, we compared three variants of the Agricultural Production System Simulator (APSIM): (i) APSIMc, which models development with no temperature threshold for pod-set; (ii) APSIMx, which sets a threshold of 15 °C for pod-set; and (iii) APSIMw, derived from APSIMc with an algorithm to moderate the developmental rate as a function of soil water, in addition to temperature and photoperiod common to all three models. Comparison of modelled and actual flowering and pod-set of a common cheque cultivar PBA BoundaryA in 54 diverse environments showed that accuracy and precision were superior for APSIMw. Because of improved prediction of flowering and pod-set timing, APSIMw improved the computation of the frequency of post-flowering frosts compared to APSIMc and APSIMx. The number of heat events was similar for all three models. We conclude that accounting for water effects on plant development can allow better matching between phenology and environment.

Assessing the effect of using different APSIM model configurations on model outputs

The three major sources of uncertainty in crop models are model inputs, structure and parameters. Model structure is one of the major contributors to this uncertainty, however, its quantification is difficult due to limitations in controlling confounding effects from parameter and input uncertainty. The objective of this study was to quantify the contribution of structural uncertainty to the variance in model outputs produced by the Agricultural Production Systems sIMulator (APSIM). Outputs investigated were yield, irrigation requirements, partial gross margin, drainage and nitrogen (N) leaching. Eight model structures differing in choice of soil water model, crop model and irrigation model were developed within a single APSIM version (v.7.10) and tested under three contrasting environments (climate × soil) across 120 years. We quantified: (i) the model structure uncertainty (from soil water, crop and irrigation models) using analysis of variance (ANOVA) and deviation analysis; and (ii) the variability of outputs due to model structure and climate using the coefficient of variation. Confounding effects from inputs, parameters and model users were controlled. Most structural uncertainty resulted from first order effects of the choice of model components (crop model: 12.2–98.9%, irrigation model: 0–78.4%, soil water model:1–33.7%) rather than second order interactions between components (0.1–18.9%). Furthermore, uncertainty from choice of sub-model/model used was not necessarily related to the structural complexity of these components. The effects of structural uncertainty on predictions commonly used to inform agronomic, ecological or policy decision making were strongly impacted by site and climate conditions i.e., high rainfall site (∼1330 mm year−1) had less uncertainty and variability as compared to low rainfall site (∼610 mm year−1), highlighting the need for any uncertainty assessment to cover the entire range of conditions for model application. Here we show the value of a component-based modelling framework for quantifying uncertainty in crop modelling studies.

Simulating Gibberellic Acid Effect on Pasture Yield on Naturally Deposited and Fixed Area Urine

Nitrate (NO3−-N) leaching from urine patches is a serious environmental concern in dairy pastoral systems. In our previous research, we established that application of a plant growth hormone, gibberellic acid (GA), can potentially reduce NO3−-N leaching in urine patches; however, this was investigated in two locations in New Zealand. The performance of GA in influencing pasture nitrogen (N) uptake and NO3−-N leaching needs to be undertaken in multi-locations to draw conclusions. However, multi-location studies are a challenge due to a lack of funding and time constraints, so models such as the agricultural production systems simulator (APSIM) have been used. Therefore, field studies were conducted to determine whether APSIM can be used to quantify and simulate the effect of GA on NO3−-N leaching and pasture yield in three experimental sites known as Ashburton, Stratford, and Rotorua in New Zealand. Treatments examined were control (no urine applied), urine at 600 kg N ha−1, urine + GA at 8 g ha−1. The observed data was used to calibrate and validate the model. APSIM simulated that application of GA reduced NO3−-N leaching (relative to urine treatment) by 4.6, 5.1, and 8.8 kg NO3−-N ha−1 in Ashburton, Stratford, and Rotorua, respectively. APSIM reliably simulated pasture dry matter yield, and this was confirmed by the coefficient of determination ranging from R2 = 0.8562 to 0.995 in all treatments and experimental sites. This study demonstrated that APSIM can effectively be used to simulate the effect of GA application on NO3−-N leaching and pasture yield. Therefore, APSIM can be applied in other areas to simulate NO3−-N leaching and pasture yield.