Crop Model Research Articles

Ideotype map research based on a crop model in the context of a climatic gradient

Due to increasing climate uncertainties, optimizing plant traits is essential for sustainable agriculture. This article presents an approach that combines advanced modelling techniques to identify optimal plant traits under various agro-environmental conditions. By integrating a crop model, a climate generator, and our PEQI algorithm (Profile Expected Quantile Improvement), our method aims to create ideotype maps tailored to specific regions.We use the SAMARA model (Simulator of crop trait Assembly, MAnagement Response, and Adaptation), calibrated with trials carried in Sahel on a set of local varieties, to simulate crop growth in diverse environments. The PEQI algorithm adjusts varietal parameters to maximize expected yield, defining precise selection objectives known as ideotypes, which are particularly important in regions with irregular rainfall patterns like the Sahel.With the PEQI algorithm based on a kriging metamodel, we ensure effective adaptation to spatially variable environments. By leveraging a climate generator to simulate meteorological variability, our integrated approach optimizes crop yields in regions such as Senegal, southern Mali, Burkina Faso, and Guinea-Bissau. The outcome is an ideotype map for sorghum, providing breeders with a robust decision-support tool to enhance crop performance amidst climate uncertainty.

A Review on the Optimization of Irrigation Schedules for Farmlands Based on a Simulation–Optimization Model

Agriculture is the most important sector that is consuming water resources. In the context of global water scarcity, how to use limited water resources to improve water use efficiency in agriculture or achieve maximum crop yield and fruit quality is of great significance for ensuring food and water security. Optimizing irrigation schedules is an effective measure to improve water use efficiency, where crop models also play an important role. However, there is little research summarizing the optimization of irrigation schedules based on crop models. This study provides a systematic review on how to optimize irrigation schedules based on crop models and simulation–optimization models. When optimizing irrigation schedules based on crop models, the selected models are usually mechanistic agro-hydrological models. Irrigation scenarios and optimization objectives are mainly focused on both crop and water aspects, such as maximizing crop yield, fruit quality, water productivity, and irrigation water productivity. Minimizing crop water consumption and total irrigation amounts serve as optimization objectives, and irrigation quantity, irrigation frequency, and irrigation interval serve as decision variables. In saline areas or low fertilizer utilization areas, the optimization objectives and decision variables also involve some indicators related to salt and nitrogen, such as the maximum desalination rate, minimum salt content, fertilizer utilization efficiency, nitrogen fertilizer productivity, nitrogen fertilizer utilization efficiency, nitrogen leaching rate, which serve as the optimization objectives, and the irrigation water salinity, or fertilization schedules serve as the decision variables. When optimizing irrigation schedules based on simulation–optimization models, the models have mainly been upgraded from water-production function to crop mechanism models. In addition, optimization algorithms have been upgraded from traditional optimization techniques to intelligent optimization algorithms. Decision-making techniques are used to make decisions on optimization results. In addition, the spatial scale for the optimization problem of irrigation schedules was developed from fields to regions, and the time scale was developed from the growth stage, beginning with months, and shortening to ten days, then to a day, and then to an hour. This study also provides a detailed introduction to widely used optimization algorithms, such as genetic algorithms, as well as decision techniques. At the same time, it is proposed that the future should focus on improving crop models and analyzing uncertainty in research on irrigation schedule optimization, which is of great significance for the precise regulation of irrigation schedules.

Development and Characterization of a New TILLING Population for Forward and Reverse Genetics in Barley (Hordeum vulgare L.).

Mutagenesis is an important tool in crop improvement and free of the regulatory restrictions imposed on genetically modified organisms. Barley (Hordeum vulgare L.) is a diploid species with a genome smaller than those of other members of the Triticeae crops, making it an attractive model for genetic studies in Triticeae crops. In this study, we report an ethyl methane sulfonate (EMS)-mutagenized population in the Chinese barley landrace TX9425, which is tolerant to both abiotic and biotic stress. A TILLING (Targeting Induced Locus Lesion in Genomes) population consisting of 2000 M2 lines was also constructed based on the CEL I enzyme with subsequent polyacrylamide electrophoresis, which decreased the cost and labor investment. The mutant phenotypes of the M2 and M3 generations were scored and revealed the presence of a wide spectrum of morphological diversity. The population was evaluated by screening for induced mutations in five genes of interest. A detailed analysis was performed for the HvGLR3.5 gene and three mutations were identified by screening in 2000 M2 lines. Two of three mutations displayed tuft and yellow striped leaves compared to the wild type. Altogether, our study shows the efficiency of screening and the great potential of the new TILLING population for genetic studies in the barley crop model system.

Machine learning models for daily net radiation prediction across different climatic zones of China

Net radiation (Rn), a critical component in land surface energy cycling, is calculated as the difference between net shortwave radiation and longwave radiation at the Earth’s surface and holds significant importance in crop models for precision agriculture management. In this study, we examined the performance of four machine learning models, including extreme learning machine (ELM), hybrid artificial neural networks with genetic algorithm models (GANN), generalized regression neural networks (GRNN), and random forests (RF), in estimating daily Rn at four representative sites across different climatic zones of China. The input variables included common meteorological factors such as minimum and maximum temperature, relative humidity, sunshine duration, and shortwave solar radiation. Model performance was assessed and compared using statistical parameters such as the correlation coefficient (R2), root mean square errors (RMSE), mean absolute errors (MAE), and Nash–Sutcliffe coefficient (NS). The results indicated that all models slightly underestimated actual Rn, with linear regression slopes ranging from 0.810 to 0.870 across different zones. The estimated Rn was found to be comparable to observed values in terms of data distribution characteristics. Among the models, the ELM and GANN demonstrated higher consistency with observed values, exhibiting R2 values ranging from 0.838 to 0.963 and 0.836 to 0.963, respectively, across varying climatic zones. These values surpassed those of the RF (0.809–0.959) and GRNN (0.812–0.949) models. Additionally, the ELM and GANN models showed smaller simulation errors in terms of RMSE, MAE, and NS across the four climatic zones compared to the RF and GRNN models. Overall, the ELM and GANN models outperformed the RF and GRNN models. Notably, the ELM model's faster computational speed makes it a strong recommendation for Rn estimates across different climatic zones of China.

Mechanical Inter- and Intra-Row Weed Control for Small-Scale Vegetable Producers

Small-scale vegetable producers often do not have modern mechanical equipment; as a result, a significant amount of inter-row and all intra-row weeding is performed manually. The development of small, affordable machines increases the competitiveness of organic vegetable production, improves sustainable land use, and reduces dependence on unwanted herbicides. In this study, a simple modular lightweight e-hoe with the capability for both inter-row (1st degree of freedom) and intra-row (2nd degree of freedom) weeding was proposed. The e-hoe uses battery-powered in-wheel drives to move the platform (3rd degree of freedom) and additional drives to operate the tools. The e-hoe was evaluated in a small greenhouse using three different tools: a traditional hoe, an adjusted rounded hoe, and an adjusted spring tine narrow hoe. The experiments were conducted at four different tool rotation speeds, using specially designed 3D-printed models for crops and weeds for evaluation. The results indicate that the efficiency of the e-hoe rates up to 95% when the right tool design and rotation speed are combined. Based on the battery capacity, the machine can be operated for approximately 3.7 h, enabling the weeding of about 3050 plants.

Determination of yield response character function for Broccoli: A case study for Ayodhya region using FAO aqua crop model

Determination of yield response character function for Broccoli: A case study for Ayodhya region using FAO aqua crop model

A study on parameter calibration of a general crop growth model considering non-foliar green organs

Accurate crop yield estimation is essential for ensuring national food security and sustainable development. The crop growth model is one of the primary methods for estimating production and has been effectively applied in simulating crop growth and development, environmental factor effects and yield formation processes. However, many related studies have focused on Gramineae crops, such as wheat, rice and maize, while there has been little research on rape and soybean yield estimation. Non-foliar green organs such as rape siliques and soybean pods make vital contributions to plant photosynthesis and influence crop output. The parameter calibration method based on the leaf area index (LAI) cannot satisfy the existing demand for high-precision yield estimation for crops with active non-foliar green organs. Therefore, a new method for crop growth model calibration was proposed, which considers non-foliar green organs and plant photosynthetic succession processes and combines them with those of leaves to construct a photosynthetic area index (PAI). With wheat, rape and soybean as the research objects, their yields were simulated via the proposed crop growth model calibration method. The results showed that using the proposed calibration method to calibrate World Food Study (WOFOST) model parameters on the basis of the PAI improved the yield estimation accuracy over that of the crop model calibration method based on the LAI. The determination coefficient (R2) of the total dry weight of storage organs (TWSO) simulation value increased from 0.73 (using the LAI) to more than 0.90 (using the PAI). The R2 values of TWSO at the rape calibration and verification points were 0.910 and 0.922, respectively, and the R2 values of TWSO at the soybean calibration and verification points were 0.741 and 0.926, respectively. The above results verified the feasibility and effectiveness of the proposed calibration method. Consequently, the application of the crop model calibration method proposed in this paper is important for accurately estimating the crop yield of plants with active non-foliar green organs, promoting the expansion of general crop models for different types of crops and achieving high-precision crop yield estimations.

Capacity development training workshop on crop simulation modelling and effects of climate risks on agricultural production systems in Southeast Asia

Southeast Asia (SEA) has experienced frequent floods and droughts, posing severe challenges for farmers, agricultural scientists and extension officers. Consequently, crop modelling has become imperative in developing agricultural production systems and making informed decisions at the field level. The Decision Support System for Agrotechnology Transfer (DSSAT) can be effectively utilised at both farm and regional levels to assess the influence of climate change on production across different spatial scales. Moreover, it supports planning adaptation strategies tailored to the needs of farmers. The one-week hands-on training workshop aims to enhance technical and scientific proficiency in crop simulation modelling and evaluate the effects of climate risks on agricultural production systems in Southeast Asian countries such as Cambodia, Lao PDR, Thailand and Vietnam. A total of 62 participants from the selected countries, including Singapore and Ethiopia, were joined in the training. Among them were 48 male and 24 female individuals, comprising researchers, students, scientists, academicians and extension officers. The participants predominantly acquired knowledge of crop simulation modelling techniques by utilising existing case examples and lecture materials from the DSSAT foundation. Furthermore, the training workshops establish research networks and collaborations among the participant countries, facilitating the exchange of scientific knowledge related to innovative farm management practices and fostering interactions between local agricultural communities and scientists.

Improved Winter Wheat Yield Estimation by Combining Remote Sensing Data, Machine Learning, and Phenological Metrics

Accurate yield prediction is essential for global food security and effective agricultural management. Traditional empirical statistical models and crop models face significant limitations, including high computational demands and dependency on high-resolution soil and daily weather data, that restrict their scalability across different temporal and spatial scales. Moreover, the lack of sufficient observational data further hinders the broad application of these methods. In this study, building on the SCYM method, we propose an integrated framework that combines crop models and machine learning techniques to optimize crop yield modeling methods and the selection of vegetation indices. We evaluated three commonly used vegetation indices and three widely applied ML techniques. Additionally, we assessed the impact of combining meteorological and phenological variables on yield estimation accuracy. The results indicated that the green chlorophyll vegetation index (GCVI) outperformed the normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI) in linear models, achieving an R2 of 0.31 and an RMSE of 396 kg/ha. Non-linear ML methods, particularly LightGBM, demonstrated superior performance, with an R2 of 0.42 and RMSE of 365 kg/ha for GCVI. The combination of GCVI with meteorological and phenological data provided the best results, with an R2 of 0.60 and an RMSE of 295 kg/ha. Our proposed framework significantly enhances the accuracy and efficiency of winter wheat yield estimation, supporting more effective agricultural management and policymaking.

A novel mathematical method to estimate rice phenological parameters across spatial scales for the ORYZA model

While crop models are increasingly applied to large-scale areas, inadequate observation data make it difficult to calibrate the model’s phenological parameters at the regional scale. The present study proposed a simple mathematical method for estimating rice phenological parameters across spatial scales for the ORYZA model. The method establishes the cumulative function of the phenological parameters (CFPP). By fitting the CFPP with the so-called growth curve, the 4 phenological parameters of the ORYZA model were transformed into 3 fitted parameters in the equation of CFPP. Functions between the fitted parameters and several meteorological and field management factors were established. These established functions were substituted back into CFPP to construct a modified CFPP. Due to the inter-translational relationship between CFPP and the original phenological parameters, the values of phenological parameters could be estimated by the modified CFPP based on meteorological and field management factors. The newly proposed mathematical method was applied in the Yangtze River Basin (YRB), China. The results indicated that the multi-station average of the absolute value of relative errors for the rice panicle initiation, flowering, and maturity dates within the YRB were 12.3 %, 10.5 %, and 8.7 %, respectively, which were at most 4.8 % larger than that simulated using parameters calibrated using each station’s phenological data. The phenological parameters estimated by the novel mathematical method had close performance to those calibrated directly based on observed data at most stations in terms of rice phenology simulation. The present study provided a new solution for phenological parameter calibration for crop models when applied in a large-scale area.

Game analysis of future rice yield changes in China based on explainable machine-learning and planting date optimization

ContextGlobal warming's escalating severity necessitates sophisticated approaches for predicting rice yield. Research QuestionCombining crop models with data-driven techniques, such as machine learning, can more effectively grasp the complex interplay of variables influencing crop growth. It remains a significant challenge to balance accuracy and interpretability in such hybrid models. MethodsThe research integrated the Decision Support System for Agrotechnology Transfer (DSSAT) with statistical and machine learning models respectively, to assess rice yield changes in China under four future Shared Socio-economic Pathway (SSP). SSPs are scenarios that integrate socioeconomic trends with greenhouse gas emissions and radiative forcing pathways, which affect the phenology and yield of rice. The Shapley Additive Explanation (SHAP) method was employed to interpret the model, effectively determining the interplay among variables influenced rice yields. Mitigated the negative impacts of climate change on rice yield through the planting date optimization. ResultsProjections indicate significant rice yield losses in China without CO2, worsening with increased radiative forcing (p < 0.001). Considering rising CO2, single-season rice yields are projected to increase by 0.1–3.6 %, early rice by 4.6–9.5 %, while late rice yields are still decrease by 2.3–8.8 %. The rising CO2 can offset yield losses for single and early rice but not for late rice. The hybrid approach which combined the Random Forest (RF) with the DSSAT performed best in predicting rice yield. Studies showed that rising temperatures caused rice yield losses in China, yet we found that Growing Degree Days (GDD) exerted a more negative impact (p < 0.001). In high-precipitation regions, deep soil moisture is more influential than shallow soil moisture, whereas the reverse was true in drier areas (p < 0.001). Advancing planting dates for early and single rice and delaying for late rice can increase yields (p < 0.001). Adjusting to optimal planting dates, single-season rice yields increased by 3.3–6.3 %, early rice increased by 9.7–18.3 %, while late rice still decreased by 1.0–4.7 %. ConclusionsWithout considering the impact of CO2, significant rice yield losses in China are projected. Even with the fertilization effect of CO2, rice yields remain negatively impacted by climate change. However, implementing appropriate measures, such as optimizing planting dates, can help Chinese rice production benefit under changing climate. ImplicationsThis study offers insights into balancing accuracy and interpretability in hybrid models and provides guidance for local policymakers to address future climate change.

Modeling of cotton yield responses to different irrigation strategies in Southern Xinjiang Region,China

Southern Xinjiang region represents a prototypical oasis agricultural area, and cotton is the primary cash crop. The current cotton water productivity for this area remains room for improvement. In order to refine cotton drip irrigation strategy in Southern Xinjiang, this study combined field experiment with crop model simulation and prediction. In this experiment, two irrigation methods, furrow irrigation and drip irrigation, and different degrees of deficit irrigation were set up. The impacts of soil moisture on cotton yield and quality were assessed, with the additional utilization of the Root Zone Water Quality Model (RZWQM2) for simulating irrigation scenarios and predicting cotton yield field data showed that during the 2021–2022 growing season, drip irrigation significantly increased boll weight, yield, and Water Productivity compared to furrow irrigation. The deficit irrigation with 90 % of full irrigation during the flowering period resulted in the highest yield among all drip irrigation treatments, while simultaneously improving the fiber quality. The RZWQM2 model was calibrated and validated using 2021 and 2022 cotton field experimental data, respectively. The calibrated and verified RZWQM2 model shows good performance. In terms of soil water storage, cotton yield and evapotranspiration. The simulated value output by the model is close to the measured value in the field. Model prediction data suggest that there is still room for improvement in cotton yield, accompanied by an increase in evapotranspiration. But deficit irrigation below 80 % of the full irrigation quotas poses a notable risk of yield reduction. Targeting higher yields, the predicted scenario suggests an increase in water application (irrigation quota:314.8 mm). Additionally, this study proposes irrigation regimes with water savings of 10 % and 20 %, with irrigation quotas of 290.5 mm and 266.5 mm during the growth period, respectively. Based on this study, we provide a more detailed and reliable water-saving scheme for cotton drip irrigation in southern Xinjiang region.

Simulation of source sink partitioning in wheat under varying nitrogen regimes using DSSAT-CERES-wheat model

Grain yields in wheat can be limited by the assimilate supply (source) or by the carbohydrate demand of the grains (sink). Recently, there have been questions regarding the capability of crop models to simulate the physiology of source-sink interactions in crops; however, crop models scarcely tested with source-sink partitioning. DSSAT_CERES_Wheat model was used with details of field experimental data having treatments of manipulated source (i.e., assimilate supply), sink (i.e., kernel number). The aim of the present study was to assess the impact of different levels of nitrogen and source-sink manipulation on wheat crop and to model source-sink partitioning in wheat under varying N-Regimes and climatic conditions. The experiment was conducted during wheat growing seasons of 2015–16 and 2016–17, at two locations (Islamabad and URF Koont Chakwal), under five different levels of nitrogen and three source sink treatments (Control (100 % RUE), 50 % shading pre-anthesis (50 % RUE), 50 % spike removal i.e. spike halving) using randomized complete block design. Recommended rates of fertilizer were applied with the exception of nitrogen which was 0, 50, 100, 150 and 200 kg ha−1, while each treatment was replicated thrice. CERES-Wheat model was calibrated using 2015–16 observed data while model was evaluated using two-year field collected data of two sites i.e. Islamabad and Chakwal. The model was able to simulate treatments impacts on phenology (R2, RMSE and d-index values of 0.89, 2.80 days and 0.97 respectively at Islamabad while at Chakwal R2 = 0.89, RMSE = 2.65 days and d-index = 0.94), leaf area index (R2 = 0.94, 0.94, RMSE = 0.51, 0.38 and d-index = 0.98 and 0.92 at Islamabad and Chakwal respectively), biomass (R2 = 0.98, 0.96, RMSE = 370, 450 kg ha−1 and d-index = 0.96 and 0.95 at Islamabad and Chakwal respectively), grain yield (R2 = 0.97,0.96, RMSE = 0.17, 0.2 t ha−1, and d-index = 0.95 and 0.93 at Islamabad and Chakwal respectively), harvest index, soil nitrogen, crop nitrogen and grain nitrogen with good accuracy. The observed range for biomass water use efficiency (BM_WUE) was 34.1–14.5 kg ha−1 mm−1 while grain WUE remained in the range of 10.3–3.7 kg ha−1 mm−1. The results depicted that model could reproduce observed effects of shading and halving the spikes. Crop response to modified radiation use efficiency (RUE) was variable among sites which could be critical for studying crop environment interactions, improving WUE, estimating genetically and atmospheric CO2-related increased RUE, analyzing impact of solar dimming and source manipulations under biotic stress.

Vulnerability of barley yields in Québec.

Climate projections across Québec indicate increased water stress and recurrent vulnerability of cropping systems. In recent years, reports of droughts and water stress have been recorded across the province. Many parts of Québec have experienced droughts in the past few years, which have had uninvestigated impacts on crops. These droughts have been described as some of the most significant in the last 80 years. On the positive side, climate change is likely to trigger shorter winters and therefore longer growing seasons for several crops. However, for crops like maple syrup, the regions suitable for their cultivation will shift northwards. Despite these projections, studies monitoring the susceptibility of barley to environmental changes, climate variability, and adaptive capacity across Québec are still limited. This study aims to provide a provincial-scale portrait of vulnerability of barley in Québec based on historical growing season precipitation, barley yield, and socioeconomic data (literacy and poverty rates) using a composite statistical model. Growing season precipitation data were downloaded from Ouranos. Barley yield data were collected from the Institut de la Statistique du Québec, and the socio-demographic data were collected from the Advisory Council of Poverty and the Institut de la Statistique du Québec. A vulnerability index with sub-indices (sensitivity, exposure, and adaptive capacity) is deployed. It is hypothesised that (1) vulnerability is inversely associated with adaptive capacity, and (2) the peripheral regions of Québec are more vulnerable and less adaptive to climate stressors. Initial results show that when the vulnerability index for barley is more prominent, the associated index of adaptive capacity is relatively lower. A significant gradient between the peripheral and southern regions of Québec is observed, with vulnerability lowest around Montreal/Laval and gradually increasing towards the peripheral regions. A better understanding of vulnerability warrants a change in approach from focusing solely on climate-related variables to integrating socioeconomic proxies. The challenge, however, has been how to introduce these socioeconomic proxies into empirical and process-based crop models.

Time Series Modeling and Forecasting of Finger Millet Cultivation Area, Production and Productivity in Chhattisgarh, India: The Box Jenkins Methodology

Chhattisgarh has taken important steps towards promoting millets cultivation and improving the livelihood of farmers. To increase millets production in Chhattisgarh, the state government launched the Millet Mission in September 2021. This mission has been started with a view to make Chhattisgarh the ‘millet hub of India’. The present study was conducted on time series modelling and forecasting of finger millet crops in Chhattisgarh India using Box Jenkins methodology and used historical data on currently cultivated area, production and yield of finger millet crops. The time series data was collected from 2001 to 2023, and analysis of the study was carried out using path analysis and Box Jenkins ARIMA model; and among various 20 models the best and suitable ARIMA (0, 1, 2), (3, 1, 1) and (2, 0, 4) model was selected based on AIC, BIC, MAPE, RMSE, MAE. With the help of the selected appropriate model, the cultivation area, production and yield of finger millet cultivation in Chhattisgarh was forecasted for the year 2024 to 2030. But marvellous, diminishing and fluctuating trend was observed in finger millet cultivation area and production over the forecast period. Whereas increasing and stochastic trend was observed in finger millet yield over the forecast period.

Observational evidence for groundwater influence on crop yields in the United States

As climate change shifts crop exposure to dry and wet extremes, a better understanding of factors governing crop response is needed. Recent studies identified shallow groundwater-groundwater within or near the crop rooting zone-as influential, yet existing evidence is largely based on theoretical crop model simulations, indirect or static groundwater data, or small-scale field studies. Here, we use observational satellite yield data and dynamic water table simulations from 1999 to 2018 to provide field-scale evidence for shallow groundwater effects on maize yields across the United States Corn Belt. We identify three lines of evidence supporting groundwater influence: 1) crop model simulations better match observed yields after improvements in groundwater representation; 2) machine learning analysis of observed yields and modeled groundwater levels reveals a subsidy zone between 1.1 and 2.5 m depths, with yield penalties at shallower depths and no effect at deeper depths; and 3) locations with groundwater typically in the subsidy zone display higher yield stability across time. We estimate an average 3.4% yield increase when groundwater levels are at optimum depth, and this effect roughly doubles in dry conditions. Groundwater yield subsidies occur ~35% of years on average across locations, with 75% of the region benefitting in at least 10% of years. Overall, we estimate that groundwater-yield interactions had a net monetary contribution of approximately $10 billion from 1999 to 2018. This study provides empirical evidence for region-wide groundwater yield impacts and further underlines the need for better quantification of groundwater levels and their dynamic responses to short- and long-term weather conditions.

User-tailored sub-selection of climate model ensemble members for impact studies

Climate model ensembles serve as an input to all impact studies that use sector-specific models (e.g., hydrological, ecological, crop models, fire hazard) at regional or local scales. These models require regionally scaled climate information to simulate the potential environmental and socio-economic sectoral impacts of climate change. Such simulations are based on comprehensive multi-member climate ensembles derived from the bias-adjusted and downscaled global and regional climate models. Due to limited computational resources, users of climate scenarios often can only include a small number of the ensemble members in their calculations, and therefore they often select them at random. A pre-selection of meaningful, consistent and case-specific members is therefore desired by the climate data users. In this work, we aim to fill this gap and present a novel user-tailored procedure for sub-selecting ensemble members for a variety of applications. Our method is based on the ranking of the climate change signal (CCS) calculated for a set of climate indices (e.g., mean temperature or number of hot days). Based on the CCS strength, three ensemble members representing the strongest, weakest, and median CCS are selected for each application. We also demonstrate the robustness of our approach in a specific hydrological impact model framework. Providing a systematic procedure not only assists impact modelers in selecting appropriate members, but also improves the consistency and comparability of different impact studies.

Simulating water-limited potato yields across the Netherlands with (SWAP-)WOFOST: Experimentation, model improvement and evaluation

Water availability explains a large part of the spatial and temporal yield variability of ware potato in the Netherlands. Climate change is projected to lead to greater variability in water availability. Therefore, an accurate simulation of crop yield as a function of water availability is of high importance. Many crop models can simulate water-limited potato yields, but as detailed experimental data on drought and oxygen stress are scarce, models are often not well calibrated and evaluated. We set up a large experiment and an on-farm observational trial to calibrate and evaluate the simulation of water-limited yield of different potato cultivars across the Netherlands. In addition, we investigated the required model complexity to accurately simulate water-limited potato yields. The crop growth model WOFOST was used for simulations either as a single model or coupled with SWAP. Yields were simulated applying either only drought stress or by including both drought and oxygen stress. We employed two methods for simulating oxygen stress in SWAP-WOFOST. The calibration improved accuracy of simulated water-limited yields as demonstrated in our model evaluation. Both on sandy and clayey soils, water-limited yields were best simulated using SWAP-WOFOST while simulating oxygen stress using a process-based method in combination with a rooting density that decreases linearly with rooting depth. On sandy soils, the bottom boundary condition of free drainage was applicable. On clayey soils measured groundwater levels (or soil water pressure heads when measurements are not available) should be used as bottom boundary condition. On sandy soils both WOFOST and SWAP-WOFOST may be used, as results were similar. On clayey soils, WOFOST is less suitable, as it cannot simulate capillary rise. SWAP-WOFOST however has many options and parameters, which can result in large differences in results. Hence, a careful model set up and evaluation is required for each application.

Estimating soil mineral nitrogen from data-sparse field experiments using crop model-guided deep learning approach

Sandy soils are susceptible to excessive nitrogen (N) leaching under intensive crop production which is linked with the soil’s low nutrient holding capacity and high-water infiltration rate. Estimating soil mineral nitrogen (SMN) at the daily time-step is crucial in providing fertilizer recommendations balancing plant nitrogen use efficiency (NUE) and N losses to the environment. Crop models [e.g., Decision Support System for Agrotechnology Transfer (DSSAT)] can simulate SMN trend under varied fertilizer application rates and timings but struggle with accuracy in high-water table conditioned sub-irrigated soil. Alternatively, time-series deep learning (DL) models based on a long short-term memory (LSTM) are promising in understanding nonlinearity among complex variables. Yet, purely data-driven DL models for crops are challenging to develop due to limited field data availability and the excessive costs to collect more field data. Hence, a hybrid model (hybrid-LSTM) was developed by leveraging both the DSSAT and LSTM models to estimate daily SMN primarily using daily weather, applied fertilizer rates-timings, and the SMN sparse observations. This study used the observations from field trials conducted between 2011–2014 in Hastings, FL. The first step was to calibrate the DSSAT-SUBSTOR-Potato model to produce reliable SMN of the topsoil for treatments with varied N applied fertilizer rates split among the pre-planting, emergence, and tuber-initiation stages of the potato. Thereafter, the hybrid-LSTM model was trained on the calibrated DSSAT simulated SMN time-series and fine-tuned its predictions using the observed SMN to improve DSSAT simulated SMN. The hybrid-LSTM model was then tested on both calibrated and uncalibrated DSSAT SMN simulations where it outperformed the DSSAT model (improvement ranged ∼18–30 % on comparing the normalized root mean squared error) in providing reliable estimates of SMN across most of the farms and years. This novel hybrid modeling approach could guide stakeholders and farmers to build sustainable N management with improved crop NUE and yield and help in minimizing environmental losses.

Predicting rice phenology across China by integrating crop phenology model and machine learning

This study explores the integration of crop phenology models and machine learning approaches for predicting rice phenology across China, to gain a deeper understanding of rice phenology prediction. Multiple approaches were used to predict heading and maturity dates at 337 locations across the main rice growing regions of China from 1981 to 2020, including crop phenology model, machine learning and hybrid model that integrate both approaches. Furthermore, an interpretable machine learning (IML) using SHapley Additive exPlanation (SHAP) was employed to elucidate influence of climatic and varietal factors on uncertainty in crop phenology model predictions. Overall, the hybrid model demonstrated a high accuracy in predicting rice phenology, followed by machine learning and crop phenology models. The best hybrid model, based on a serial structure and the eXtreme Gradient Boosting (XGBoost) algorithm, achieved a root mean square error (RMSE) of 4.65 and 5.72 days and coefficient of determination (R2) values of 0.93 and 0.9 for heading and maturity predictions, respectively. SHAP analysis revealed temperature to be the most influential climate variable affecting phenology predictions, particularly under extreme temperature conditions, while rainfall and solar radiation were found to be less influential. The analysis also highlighted the variable importance of climate across different phenological stages, rice cultivation patterns, and geographic regions, underscoring the notable regionality. The study proposed that a hybrid model using an IML approach would not only improve the accuracy of prediction but also offer a robust framework for leveraging data-driven in crop modeling, providing a valuable tool for refining and advancing the modeling process in rice.