Model Agnostic Meta-Learning(MAML)-Based Ensemble Model for Accurate Detection of Wheat Diseases Using Vision Transformer and Graph Neural Networks

Wheat is a critical crop,extensively consumed worldwide,and its production enhancement is essential to meet escalating demand.The presence of diseases like stem rust,leaf rust,yellow rust,and tan spot significantly diminishes wheat yield,making the early and precise identification of these diseases vital for effective disease management.With advancements in deep learning algorithms,researchers have proposed many methods for the automated detection of disease pathogens;however,accurately detectingmultiple disease pathogens simultaneously remains a challenge.This challenge arises due to the scarcity of RGB images for multiple diseases,class imbalance in existing public datasets,and the difficulty in extracting features that discriminate between multiple classes of disease pathogens.In this research,a novel method is proposed based on Transfer Generative Adversarial Networks for augmenting existing data,thereby overcoming the problems of class imbalance and data scarcity.This study proposes a customized architecture of Vision Transformers(ViT),where the feature vector is obtained by concatenating features extracted from the custom ViT and Graph Neural Networks.This paper also proposes a Model AgnosticMeta Learning(MAML)based ensemble classifier for accurate classification.The proposedmodel,validated on public datasets for wheat disease pathogen classification,achieved a test accuracy of 99.20%and an F1-score of 97.95%.Compared with existing state-of-the-art methods,this proposed model outperforms in terms of accuracy,F1-score,and the number of disease pathogens detection.In future,more diseases can be included for detection along with some other modalities like pests and weed.

A Simplified Model with Soil Water Limitation on Spring Wheat Growth

The authors constructed a simplified model of spring wheat (Triticum aestivum L.) carbon assimilation and dry matter accumulation (DMA) process which consisted of two independent variables, day length (L) and total daily radiation (TDR). Leaf water potential (Ψ) was incorporated into the simplified growth model based on the assumption that both light use efficiency (α) and CO 2 conductance of assimilation (g c) were depressed by water limitation. Finally,Ψ was estimated from a regression equation in which the independent variables were relative soil water content in the upper 80 cm (θ R,80 ), ambient temperature (T a), vapor pressure deficit (VPD), the cumulative leaf water potential below thresholds of -1.5 MPa (Ψ c,1.5 ). Some applications in research program of field experiment of atmosphere_land surface processes in Heihe River region were tested. The simulated data agreed well with the data observed at Linze oasis in 1989 for various levels of water supply and at Zhangye oasis in 1992 in the field. The analysis and simulation using the model demonstrated that the simplified growth model could describe very well the DMA process of spring wheat with and without water limitation in the region of HEIFE (Heihe field experiment).

Comparison on Winter Wheat Yield Estimating Models Based on Radarsat-2 and HJ Satellite in Huaihe River Region

The establishment of crop yield estimating model based on microwave and optical satellite images can conduct the mutual verification of the accuracy of the reported crop yield and the precision of the estimating model. With Shou County and Huaiyuan County of Anhui Province as the experimental fields of winter wheat producing areas, the linear winter wheat yield estimating models were established by adopting backscattering coefficient and Normalized Difference Vegetation Index（NDVI） based on images from the synthetic aperture radar（SAR）—RDARSAT-2 and HJ satellite photographed in mid-April and early May, 2014, and then comparisons were conducted on the accuracy of the yield estimating models. The accuracies of the yield estimating models established using co-polarized（HH） and cross-polarized（HV） modes of SAR in Jiangou Town, Shou County were 68.37% and 74.01%, respectively, while the accuracies in Longkang Town, Huaiyuan County were 63.10%and 69.10%, respectively. Accuracies of yield estimating models established by HJ satellite data were 69.52% and 66.43% in Shou County and Huaiyuan County, respectively. Accuracies of winter yield estimating model based on HJ satellite data and that based on SAR were closed, and the yield difference of winter wheat in the lodging region was analyzed in detail. The model results laid the foundation and accumulated experience for the verification, parameters correction and promotion of the winter wheat yield estimating model.

Establishment and Application of Fertilization Model for Wheat Interplanted with Cotton in Jianghan Plain

3414 field experiment （including three nutrient elements at four gradient levels, a total of 14 unrepeated incomplete treatments） was designed to study the fertilization measures for wheat interplanted with cotton in Qianjiang City, Hubei Province. Fertilizer model for wheat interplanted with cotton in Jianghan Plain was finally established, based on which the soil nutrient indices in wheat-cotton inter- planting field were screened; and optimal nitrogen, phosphorus and potassium appli- cation for wheat was put forward as 130-210 kg/hm2 N, 40-70 kg/hm2 P2O5 and 40-60 kg/hm2 K2O.

The impacts of climate change on wheat yield in the Huang-Huai-Hai Plain of China using DSSAT-CERES-Wheat model under different climate scenarios

Climate change has been documented as a major threat to current agricultural strategies.Progress in understanding the impact of climate change on crop yield is essential for agricultural climate adaptation,especially for the Huang-Huai-Hai Plain(3H Plain)of China which is an area known to be vulnerable to global warming.In this study,the impacts of climate change on winter wheat(Triticum aestivum L.)yield between the baseline period(1981–2010)and two Representative Concentration Pathways(RCP8.5 and RCP4.5)were simulated for the short-term(2010–2039),the medium-term(2040–2069)and the long-term(2070–2099)in the 3H Plain,by considering the relative contributions of changes in temperature,solar radiation and precipitation using the DSSAT-CERES-Wheat model.Results indicated that the maximum and minimum temperatures(TMAX and TMIN),solar radiation(SRAD),and precipitation(PREP)during the winter wheat season increased under these two RCPs.Yield analysis found that wheat yield increased with the increase in SRAD,PREP and CO2 concentration,but decreased with an increase in temperature.Increasing precipitation contributes the most to the total impact,increasing wheat yield by 9.53,6.62 and 23.73%for the three terms of future climate under RCP4.5 scenario,and 11.74,16.38 and 27.78%for the three terms of future climate under RCP8.5 scenario.However,as increases in temperature bring higher evapotranspiration,which further aggravated water deficits,the supposed negative effect of increasing thermal resources decreased wheat yield by 1.92,4.08 and 5.24%for the three terms of future climate under RCP4.5 scenario,and 3.64,5.87 and 5.81%for the three terms of future climate under RCP8.5 scenario with clearly larger decreases in RCP8.5.Counterintuitively,the impacts in southern sub-regions were positive,but they were all negative in the remaining sub-regions.Our analysis demonstrated that in the 3H Plain,which is a part of the mid-high latitude region,the effects of increasing thermal resources were counteracted by the aggravated water deficits caused by the increase in temperature.

Global sensitivity analysis of the AquaCrop model for winter wheat under different water treatments based on the extended Fourier amplitude sensitivity test

Sensitivity analysis （SA） is an effective tool for studying crop models; it is an important link in model localization and plays an important role in crop model calibration and application. The objectives were to （i） determine influential and non-influential parameters with respect to above ground biomass （AGB）, canopy cover （CC）, and grain yield of winter wheat in the Beijing area based on the AquaCrop model under different water treatments （rainfall, normal irrigation, and over-irrigation）; and （ii） generate an AquaCrop model that can be used in the Beijing area by setting non-influential parameters to fixed values and adjusting influential parameters according to the SA results. In this study, field experiments were conducted during the 2012-2013,2013-2014, and 2014-2015 winter wheat growing seasons at the National Precision Agriculture Demonstration Research Base in Beijing, China. The extended Fourier amplitude sensitivity test （EFAST） method was used to perform SA of the AquaCrop model using 42 crop parameters, in order to verify the SA results, data from the 2013-2014 growing season were used to calibrate the AquaCrop model, and data from 2012-2013 and 2014-2015 growing seasons were val- idated. For AGB and yield of winter wheat, the total order sensitivity analysis had more sensitive parameters than the first order sensitivity analysis. For the AGB time-series, parameter sensitivity was changed under different water treatments; in comparison with the non-stressful conditions （normal irrigation and over-irrigation）, there were more sensitive parameters under water stress （rainfall）, while root development parameters were more sensitive. For CC with time-series and yield, there were more sensitive parameters under water stress than under no water stress. Two parameters sets were selected to calibrate the AquaCrop model, one group of parameters were under water stress, and the others were under no water stress, there were two more sensitive parameters （growing degree-days （GDD） from sowing to the maximum rooting depth （root） and the maximum effective rooting depth （rtx）） under water stress than under no water stress. The results showed that there was higher accuracy under water stress than under no water stress. This study provides guidelines for AquaCrop model calibration and application in Beijing, China, as well providing guidance to simplify the AquaCrop model and improve its precision, especially when many parameters are used.

Assimilation of temporal-spatial leaf area index into the CERES-Wheat model with ensemble Kalman filter and uncertainty assessment for improving winter wheat yield estimation

To accurately estimate winter wheat yields and analyze the uncertainty in crop model data assimilations, winter wheat yield estimates were obtained by assimilating measured or remotely sensed leaf area index （LAI） values. The performances of the calibrated crop environment resource synthesis for wheat （CERES-Wheat） model for two different assimilation scenarios were compared by employing ensemble Kalman filter （EnKF）-based strategies. The uncertainty factors of the crop model data assimilation was analyzed by considering the observation errors, assimilation stages and temporal-spatial scales. Overalll the results indicated a better yield estimate performance when the EnKF-based strategy was used to comprehen- sively consider several factors in the initial conditions and observations. When using this strategy, an adjusted coefficients of determination （R2） of 0.84, a root mean square error （RMSE） of 323 kg ha-1, and a relative errors （RE） of 4.15% were obtained at the field plot scale and an R2 of 0.81, an RMSE of 362 kg ha-1, and an RE of 4.52% were obtained at the pixel scale of 30 mx30 m. With increasing observation errors, the accuracy of the yield estimates obviously decreased, but an acceptable estimate was observed when the observation errors were within 20%. Winter wheat yield estimates could be improved significantly by assimilating observations from the middle to the end of the crop growing seasons. With decreasing assimilation frequency and pixel resolution, the accuracy of the crop yield estimates decreased; however, the computation time decreased. It is important to consider reasonable temporal-spatial scales and assimilation stages to obtain tradeoffs between accuracy and computation time, especially in operational systems used for regional crop yield estimates.

Rank correlation among different statistical models in ranking of winter wheat genotypes'

Several statistical methods have been developed for analyzing genotype×environment(GE)interactions in crop breeding programs to identify genotypes with high yield and stability performances.Four statistical methods,including joint regression analysis(JRA),additive mean effects and multiplicative interaction(AMMI)analysis,genotype plus GE interaction(GGE)biplot analysis,and yield–stability(YSi)statistic were used to evaluate GE interaction in20 winter wheat genotypes grown in 24 environments in Iran.The main objective was to evaluate the rank correlations among the four statistical methods in genotype rankings for yield,stability and yield–stability.Three kinds of genotypic ranks(yield ranks,stability ranks,and yield–stability ranks)were determined with each method.The results indicated the presence of GE interaction,suggesting the need for stability analysis.With respect to yield,the genotype rankings by the GGE biplot and AMMI analysis were significantly correlated(P<0.01).For stability ranking,the rank correlations ranged from 0.53(GGE–YSi;P<0.05)to0.97(JRA–YSi;P<0.01).AMMI distance(AMMID)was highly correlated(P<0.01)with variance of regression deviation(S2di)in JRA(r=0.83)and Shukla stability variance(σ2)in YSi(r=0.86),indicating that these stability indices can be used interchangeably.No correlation was found between yield ranks and stability ranks(AMMID,S2di,σ2,and GGE stability index),indicating that they measure static stability and accordingly could be used if selection is based primarily on stability.For yield–stability,rank correlation coefficients among the statistical methods varied from 0.64(JRA–YSi;P<0.01)to 0.89(AMMI–YSi;P<0.01),indicating that AMMI and YSi were closely associated in the genotype ranking for integrating yield with stability performance.Based on the results,it can be concluded that YSi was closely correlated with(i)JRA in ranking genotypes for stability and(ii)AMMI for integrating yield and stability.

A Knowledge Model System for Wheat Production Management

A knowledge model with temporal and spatial characteristics for the quantitative design of a cultural pattern in wheat production, using systems analysis and dynamic modeling techniques, was developed for wheat management, as a decision-making tool in digital farming. The fundamental relationships and algorithms of wheat growth indices and management criteria to cultivars, ecological environments, and production levels were derived from the existing literature and research data to establish a knowledge model system for quantitative wheat management using Visual C^＋＋. The system designed a cultural management plan for general management guidelines and crop regulation indices for timecourse control criteria during the wheat-growing period. The cultural management plan module included submodels to determine target grain yield and quality, cultivar choice, sowing date, population density, sowing rate, fertilization strategy, and water management, whereas the crop regulation indices module included submodels for suitable development stages, dynamic growth indices, source-sink indices, and nutrient indices. Ewluation of the knowledge model by design studies on the basis of data sets of different eco-sites, cultiwrs, and soil types indicated a favorable performance of the model system in recommending growth indices and management criteria under diverse conditions. Practical application of the knowledge model system in comparative field experiments produced yield gains of 2.4% to 16.5%. Thus, the presented knowledge model system overcame some of the difficulties of the traditional wheat management patterns and expert systems, and laid a foundation for facilitating the digitization of wheat management.

A Dynamic Knowledge Model for Design of Suitable Sowing Date and Sowing Rate in Winter Wheat

Based on research concerning dynamic relationships of winter wheat growth to environments and production conditions, a winter wheat model for selecting suitable sowing date, population density and sowing rate under different varieties, spatial and temporal environments was developed. Case studies on sowing date with the data sets of five different eco-sites, three climatic years and soil fertility levels, and on population density and sowing rate with the data sets of two different variety types, three different soil types, soil fertility levels, sowing dates and grain yield levels indicate a good model performance for decision-making.

A Dynamic Knowledge Model for Nitrogen Fertilizationin Wheat Management

By analyzing and extracting the research progress on nitrogen fertilization in wheat, a dynamic knowledge model for management decision-making on total nitrogen rate, ratios of organic to inorganic and of basal to dressing nitrogen under different environments and cultivars in wheat was developed with principle of nutrient balance and by integrating the quantitative effects of grain yield and quality targets, soil characters, variety traits and water management levels. Case studies on the nitrogen fertilization model with the data sets of different eco-sites, cultivars, soil fertility levels, grain yield and quality targets and water management levels indicate a good performance of the model system in decision-making and wide applicability.

A Knowledge Model-and Growth Model-Based Decision Support System for Wheat Management

By applying the system analysis principle and mathematical modeling technique to knowledge expression system for crop cultural management, the fundamental relationships and quantitative algorithms of wheat growth and management indices to variety types, ecological environments and production levels were analysed and extracted, and a dynamic knowledge model with temporal and spatial characters for wheat management(WheatKnow)was developed. By adopting the soft component characteristics as non language relevance , re-utilization and portable system maintenance. and by further integrating the wheat growth simulation model(WheatGrow)and intelligent system for wheat management, a comprehensive and digital knowledge model, growth model and component-based decision support system for wheat management(MBDSSWM)was established on the platforms of Visual C++ and Visual Basic. The MBDSSWM realized the effective integration and coupling of the prediction and decision-making functions for digital crop management.

西北春麦区APSIM-Wheat模型作物参数全局敏感性分析与比较

为提高西北春麦区APSIM-Wheat模型的适应性及校准的效率和精度,需要对作物参数进行重新率定,并对参数的取值范围进行标定,设定±10%、±20%、±30%、±40%和±50%五个参数取值范围,采用EFAST法和Morris法比较不同参数取值范围下模型的18个作物参数对春小麦产量和生物量的敏感性,并分析EFAST方法下最适宜的参数取值范围,同时对比两种敏感性分析方法的一致性。结果表明,对产量敏感的参数分别为每克茎籽粒数量(GPGS)、初花期积温(TFI)、灌浆期积温(TSGF)、出苗期到拔节期积温(TEOJ)、开花期积温(TF)、潜在灌浆速率(PGFR);对生物量敏感的参数分别为出苗期到拔节期积温(TEOJ)、灌浆期积温(TSGF)、初花期积温(TFI)、开花期积温(TF)。模型敏感参数和不敏感参数的适宜取值范围分别为±30%和±5%。不同参数取值范围下EFAST法和Morris法得到的敏感性结果一致性较好,在进行参数的敏感性分析时可以互相替代。

Modelling the meteorological influence on the yellowing of spring wheat leaves

The yellowed-leaf rate is one of the important variables in simulation models for thegrowth of spring wheat. Based on the field experiments (1985-1988), the evolution of yellowed-leafrate of spring wheat is analyzed. The functional relationship between the yellowing process of greenleaves and the development stages of spring wheat is established. Based on modelling and correctingfor the yellowing proass of green leaves affected by temperature and moisture, the synthetic modelfor simulating the dynaniical evolution of yellowed-leaf rate is constructed. The numerical experi-inents show that the result of the modelling is satisfactory.

Calibration and validation of SiBcrop Model for simulating LAI and surface heat fluxes of winter wheat in the North China Plain

The accurate representation of surface characteristic is an important process to simulate surface energy and water flux in land-atmosphere boundary layer.Coupling crop growth model in land surface model is an important method to accurately express the surface characteristics and biophysical processes in farmland.However,the previous work mainly focused on crops in single cropping system,less work was done in multiple cropping systems.This article described how to modify the sub-model in the SiBcrop to realize the accuracy simulation of leaf area index(LAI),latent heat flux(LHF)and sensible heat flux(SHF)of winter wheat growing in double cropping system in the North China Plain(NCP).The seeding date of winter wheat was firstly reset according to the actual growing environment in the NCP.The phenophases,LAI and heat fluxes in 2004–2006 at Yucheng Station,Shandong Province,China were used to calibrate the model.The validations of LHF and SHF were based on the measurements at Yucheng Station in 2007–2010 and at Guantao Station,Hebei Province,China in 2009–2010.The results showed the significant accuracy of the calibrated model in simulating these variables,with which the R2,root mean square error(RMSE)and index of agreement(IOA)between simulated and observed variables were obviously improved than the original code.The sensitivities of the above variables to seeding date were also displayed to further explain the simulation error of the SiBcrop Model.Overall,the research results indicated the modified SiBcrop Model can be applied to simulate the growth and flux process of winter wheat growing in double cropping system in the NCP.

Validation and Evaluation of a Mechanistic Model of Phasic and Phenological Development in Wheat

Three sets of data from the field experiments with different wheat( Triticum L. ) varieties and sowing dates in China and USA were used to test the performance of the mechanistic model of wheat development. The results showed that the absolute prediction errors for most phasic and phenological stages ranged within 0 - 5 days, and the root mean square errors were generally less than 5 days. The model was of high accuracy and low error especially for emergence, tillering, stamen and pistil initiation, and heading stages, reflecting an enhanced level of mechanism and prediction.

A simulation of winter wheat crop responses to irrigation management using CERES-Wheat model in the North China Plain

To improve efficiency in the use of water resources in water-limited environments such as the North China Plain（NCP）, where winter wheat is a major and groundwater-consuming crop, the application of water-saving irrigation strategies must be considered as a method for the sustainable development of water resources. The initial objective of this study was to evaluate and validate the ability of the CERES-Wheat model simulation to predict the winter wheat grain yield, biomass yield and water use efficiency（WUE） responses to different irrigation management methods in the NCP. The results from evaluation and validation analyses were compared to observed data from 8 field experiments, and the results indicated that the model can accurately predict these parameters. The modified CERES-Wheat model was then used to simulate the development and growth of winter wheat under different irrigation treatments ranging from rainfed to four irrigation applications（full irrigation） using historical weather data from crop seasons over 33 years（1981–2014）. The data were classified into three types according to seasonal precipitation： 〈100 mm, 100–140 mm, and 〉140 mm. Our results showed that the grain and biomass yield, harvest index（HI） and WUE responses to irrigation management were influenced by precipitation among years, whereby yield increased with higher precipitation. Scenario simulation analysis also showed that two irrigation applications of 75 mm each at the jointing stage and anthesis stage（T3） resulted in the highest grain yield and WUE among the irrigation treatments. Meanwhile, productivity in this treatment remained stable through different precipitation levels among years. One irrigation at the jointing stage（T1） improved grain yield compared to the rainfed treatment and resulted in yield values near those of T3, especially when precipitation was higher. These results indicate that T3 is the most suitable irrigation strategy under variable precipitation regimes for stable yield of winter wheat with maximum water savings in the NCP. The application of one irrigation at the jointing stage may also serve as an alternative irrigation strategy for further reducing irrigation for sustainable water resources management in this area.

Integrating a novel irrigation approximation method with a process-based remote sensing model to estimate multi-years'winter wheat yield over the North China Plain

Accurate estimation of regional winter wheat yields is essential for understanding the food production status and ensuring national food security.However,using the existing remote sensing-based crop yield models to accurately reproduce the inter-annual and spatial variations in winter wheat yields remains challenging due to the limited ability to acquire irrigation information in water-limited regions.Thus,we proposed a new approach to approximating irrigations of winter wheat over the North China Plain(NCP),where irrigation occurs extensively during the winter wheat growing season.This approach used irrigation pattern parameters(IPPs)to define the irrigation frequency and timing.Then,they were incorporated into a newly-developed process-based and remote sensing-driven crop yield model for winter wheat(PRYM–Wheat),to improve the regional estimates of winter wheat over the NCP.The IPPs were determined using statistical yield data of reference years(2010–2015)over the NCP.Our findings showed that PRYM–Wheat with the optimal IPPs could improve the regional estimate of winter wheat yield,with an increase and decrease in the correlation coefficient(R)and root mean square error(RMSE)of 0.15(about 37%)and 0.90 t ha–1(about 41%),respectively.The data in validation years(2001–2009 and 2016–2019)were used to validate PRYM–Wheat.In addition,our findings also showed R(RMSE)of 0.80(0.62 t ha–1)on a site level,0.61(0.91 t ha–1)for Hebei Province on a county level,0.73(0.97 t ha–1)for Henan Province on a county level,and 0.55(0.75 t ha–1)for Shandong Province on a city level.Overall,PRYM–Wheat can offer a stable and robust approach to estimating regional winter wheat yield across multiple years,providing a scientific basis for ensuring regional food security.

Study on Rational Irrigation for Wheat Production in Fengqiu Region:Ⅱ. A Model of Rational Irrigation forWheat and Its Evaluation

A 5-year experiment on water balance was conducted in a flat rainfed wheat field with an area of 66 m×100 m in Fengqiu, Henan Province, China. Based on the results of the 5-consecutive-year experiments,a reasonable irrigation model for wheat cultivation is suggested according to the principle of maintaining balance of the water resources. The irrigation program was designed by simulating the ideal soil moisture regimes during a wheat season. As far as the actual soil moisture was concerned, its deyiation from the ideal soil moisture was kept within 150 mm. If this model was put into practice, a grain yield of 5 250 kg ha-1 could be expected under optimal fertilization. Compared with the traditional irrigstion scheme, the suggested model saved irrigation water by 18%.

Wheat breeding model for steady and high yield and early maturation in middle and south Hebei province —Analysis of steady and high yield ability of wheat variety Ji 84-5418

A wheat breeding model for high yield in the middle and south of Hebei Province was developed. Wheat variety Ji 84-5418 has been bred on this model. The analysis results of high-yield and stability indicated that Ji 84-5418 was not only an aggregate of varied excellent characters,but a recombined biotype which could early differentiate spike and develop coordi-nately,and had better self-regulation ability and potential high productivity. Its yield is stable at 6000-8250 kg/ha.