Search

The price gap between the prices of beef and lamb in Qinghai and the average national price of beef and lamb persists and even tends to expand due to the contradiction between supply and demand, which is not only detrimental to the enhancement of income of Qinghai herders but also the healthy development of the beef and lamb industry in Qinghai. To better manage the prices of beef and lamb in Qinghai, to increase the income of Qinghai herders, and to build "a green and organic agricultural and livestock products exporting place" in Qinghai, an effective price forecasting model needs to be constructed and applied in Qinghai. Accordingly, we propose a data-driven ensemble forecasting model based on the idea of spectral clustering to forecast the future prices trend of beef and lamb in Qinghai and hope to provide a theoretical basis for the scientific and healthy development of the Qinghai livestock industry. Specifically, this paper analyzed the fluctuation pattern and characteristics of beef and lamb prices in Qinghai through the spectral clustering method, and extracted their data fluctuation features, respectively. And these features are correspondingly embedded into the proposed ensemble forecasting model to predict the future prices trend of beef and lamb in Qinghai, respectively. The competitiveness of the proposed model was verified on about 14 years' weekly Qinghai beef and lamb prices (RMB yuan/Kg) data which was obtained from the website of National Animal Husbandry Services using web crawler technology. The experimental results show that the proposed model can significantly improve the prediction accuracy of beef and lamb prices in Qinghai, and provide strong support for the management of beef and lamb prices in Qinghai, the realization of increased income of Qinghai herders, and the building of "a green and organic agricultural and livestock products exporting place" in Qinghai. [ABSTRACT FROM AUTHOR]

The development and utilization of solar energy, meteorological forecasting, and agricultural production all require accurate predictions of radiance. However, due to the influence of meteorological conditions and the failure of observation instruments and design defects, the data collection of all-sky radiance distribution is often missing or inaccurate. In this paper, a new modified method of all-sky radiance distribution (CASR) is proposed based on the Spatial distribution of daylight model recommended by Commission Internationale de l'Eclairage (CIE model). Firstly, a functional relationship between sky luminance and radiance was established based on measured data. Secondly, the abnormal values were divided into four categories according to the number of anomalous regions of radiance and the frequency of abnormal values in each sky region is determined, and the reasons were analyzed according to different altitude angles. Thirdly, the radiance and sky luminance anormal values were screened to determine their sky type by comparing them to a dataset of 15 standard sky types. The abnormal values were then corrected using the CIE luminance model with the relationship between radiance and sky luminance. Finally, the CASR model was validated against existing models. The results show that the CASR model accuracy was better than the existing model, which can accurately correct the sky luminance and radiance abnormal values and accurately characterize the distribution of radiance of the whole sky. It provides a theoretical basis for radiance prediction, natural lighting analysis and building energy consumption analysis. • Linear and polynomial functional relationships between radiance and sky luminance were established. • Classified and summarized the anomaly data according to their causes and the number of radiance anomaly regions. • A new modified method of all-sky radiance distribution based on the principle of photothermal integration was proposed. [ABSTRACT FROM AUTHOR]

Although Heilongjiang Province is the largest commodity grain producing area and modern agricultural reform pilot area, its agricultural modernization still has a long way to go. According to the actual situation of 34 regions in Heilongjiang Province, this paper evaluates the agricultural development of Heilongjiang Province from four main aspects: input level, output level, social development ability and sustainable ability. An index system of four first-level indicators and ten second-level indicators has been established. Analytic hierarchy process method and entropy weight method were used to calculate the weight of each index layer. Secondly, the least square method was used to fit the three-dimensional nonlinear value, and the extreme value standardization method was used to evaluate each index. Finally, based on the longitude and latitude data of 34 regions in Heilongjiang Province, the Moran index is calculated and the comprehensive evaluation results of the development level index of agricultural modernization in each region were given. [ABSTRACT FROM AUTHOR]

Highlights • A QR-RBF neural network model is proposed to forecast the soybean price in China. • We propose a hybrid algorithm (GDGA) to optimize the parameters of QR-RBF model. • The proposed GDGA has better forecasting performance and faster convergence. • The results show influential factors are diverse at different soybean price levels. Abstract As the price of soybean affects the soybean market development and food security in China, its forecasting is essential. A quantile regression-radial basis function (QR-RBF) neural network model is introduced in this paper. The model has two characteristics: (1) using quantile regression models to describe the distribution of the soybean price range; and (2) using RBF neural networks to approximate the nonlinear component of the soybean price. In order to optimize the QR-RBF neural network model parameters, a hybrid algorithm known as GDGA, based on a combination of the genetic algorithm (performing a global search) and a gradient descent method (performing a local search), is proposed in this paper. Data regarding the monthly domestic soybean price in China were analyzed and the results indicate that the proposed hybrid GDGA is effective. Furthermore, the results suggest that the influencing factors of soybean price vary at different price levels. Money supply and port distribution price of imported soybean were found to be important across a range of quantiles; output of domestic soybean and consumer confidence index were important only for low quantiles; and import volume of soybean and consumer price index were important only for high quantiles. [ABSTRACT FROM AUTHOR]

Time series characterized by periodic fluctuations are widespread in reality behavioural systems, and accurate forecasting of them has important practical application value. In this paper, based on the idea of discrete grey model, dual interaction effects and the nonlinear characteristics of the system are considered under the perspective of dual processing, and the fluctuating trend of the periodic sequence is portrayed through the introduction of the periodic dummy variable, and at the same time, combined with the extended periodic rolling time window method, the rolling discrete grey periodic power model with the interaction effect under dual processing is proposed, namely DIE-RDGM(1,1, T) α model. Differential evolution algorithm (DE) are employed to determine the optimum power index. Finally, the new model was compared with nine other forecasting models by three instances and applied to the problem of agricultural drought forecasting in Zhoukou City and Anyang City, Henan Province. The results show: (1) The new model has demonstrated strong competitiveness, confirming its rationality, validity, and superiority. (2) The DE is suitable for power index optimization search. (3) The new model can better describe the periodic change rule of seasonal-scale soil moisture in the two cities, and the forecast results are consistent with the actual drought conditions. [ABSTRACT FROM AUTHOR]

Accurate yield estimation and optimised nitrogen management is essential in agriculture. Remote sensing (RS) systems are being more widely used in building decision support tools for contemporary farming systems to improve yield production and nitrogen management while reducing operating costs and environmental impact. However, RS based approaches require processing of enormous amounts of remotely sensed data from different platforms and, therefore, greater attention is currently being devoted to machine learning (ML) methods. This is due to the capability of machine learning based systems to process a large number of inputs and handle non-linear tasks. This paper discusses research developments conducted within the last 15 years on machine learning based techniques for accurate crop yield prediction and nitrogen status estimation. The paper concludes that the rapid advances in sensing technologies and ML techniques will provide cost-effective and comprehensive solutions for better crop and environment state estimation and decision making. More targeted application of the sensor platforms and ML techniques, the fusion of different sensor modalities and expert knowledge, and the development of hybrid systems combining different ML and signal processing techniques are all likely to be part of precision agriculture (PA) in the near future. [ABSTRACT FROM AUTHOR]

This paper offers solutions to some of the challenges around maintaining productive agricultural land close to cities in countries facing a decline in urban populations. In such circumstances, some farmers have been observed to convert their land into real estate and leave farming before land prices decline, therefore decreasing the area of agricultural land close to large cities. In contrast, many suburban farmers in developed countries remain in farming even when land prices decline and suburbs shrink. We argue that such behaviour can be explained by a desire to remain in farming, even at the expense of profits. In such cases, agricultural income may be supplemented by rental income or by selling land. This paper demonstrates that, when land prices are high, a preferential taxation system may help farmers with real estate income to retain more of their land. This study is based on data from a survey of farmers in Tokyo, Japan where, in 1992, a programme combining preferential taxation and restrictions on the conversion of farmland was implemented. Our findings suggest that farmers in more populated areas with a strong dependence on real estate income tend to continue farming, as do those in less populated areas who are less dependent on this income source. Analysis further suggests that imposing heavy taxes on residential property simply increases living costs for farmers and results in the loss of agricultural land and that policies which promote diversification and reduce housing costs are important for keeping urban fringe land in agriculture. [ABSTRACT FROM AUTHOR]

Time-Series data has been of great importance to the research field of prediction models. Over the past two decades, multitudinous fuzzy time series blueprints have been put forth for agricultural yield production. However, most of these predictions were based on 7th interval partitioning. A surprising insight was that nobody gave a sound reason to justify the choice of that particular interval. So, this paper focuses on predicting data values on a large spectrum of fuzzy logic computations based on second and third-degree relationships. This paper showcases work on 4 different types of the fuzzy interval, where each interval is tested with 4 degrees of regression equations. Each of these 16 cases is performed for the fuzzy logic relationship (FLR) 2 and 3 separately. Apart from this, the robustness of algorithm is a testament to an incredible solution for the time series model. In addition to this, A Regression analysis model has been enforced to accomplish the efficient defuzzification operation. To elucidate the process of forecasting, the historical data of wheat yield of University of Agriculture and Technology has been used. [ABSTRACT FROM AUTHOR]

Predicting agricultural commodity prices accurately is of utmost importance due to various factors such as perishability, seasonality, production uncertainty etc. Moreover, the substantial volatility that may be exhibited in time series further adds to the complexity and constitutes a significant challenge. In this paper, a Hidden Markov (HM) guided Deep Learning (DL) models has been developed on nonlinear and nonstationary price data of agricultural commodities for forecasting by considering technical indicators viz., Moving Average (MA), Bollinger Bands (BB), Moving Average Convergence Divergence (MACD), Exponential MA (EMA) and Fast Fourier Transformation (FFT). HM Models (HMMs) can effectively handle the sequential dependencies and hidden states, while DL approach can learn complex patterns and relationships within the price series and thus the drawback of lack of generalization capability in the DL model has been overcome by HMM. In this study, the Potato price data of the Champadanga district of West Bengal, India has been utilized to assess the performance of the proposed technique. HMM has been combined with six baseline DL models viz., Recurrent Neural Networks (RNN), Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), Gated Recurrent Units (GRU), Bidirectional LSTM (BiLSTM) and Bidirectional GRU (BiGRU) for forecast modeling. Performance evaluation metrics viz., Root Mean Squared Error (RMSE), Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE) and the insightful Diebold–Mariano (DM) test revealed that Hidden Markov hybridized with DL models surpassed baseline DL models in forecasting accuracy for 1-week, 4-week, 8-week and 12-week ahead DL predictions. The proposed approach holds significant promise for enhancing the precision of agricultural commodity price forecasting with far-reaching implications for various stakeholders such as farmers and planners. • A novel Hidden Markov based Deep Learning (DL) models for accurate agricultural commodity price forecasting. • Addresses DL model generalization challenges by integrating sequential dependencies and hidden states obtained from HMMs. • Proposed models outperform baseline models - RNN, CNN, LSTM, GRU, BiLSTM, and BiGRU for price predictions of agriculture commodities. • Evaluated using RMSE, MAPE, MAE and Diebold Mariano test, for accuracy and reliability of the proposed approach. • Offers precision in commodity price forecasts, benefiting stakeholders such as farmers, planners and policymakers. [ABSTRACT FROM AUTHOR]

The optimized agricultural water schedule and management, especially in the large irrigation district along the river basins, were closely related to the spatial and temporal runoff variations. However, the impacts of climate change and human activities leads to non-linear and non-stationary monthly runoffs. Under this condition, effectively capturing the variation of the runoff time series and improving the accuracy of the prediction model are of vital importance. However, some existing monthly data-driven prediction studies mainly focus on the model structure and calculation load, ignoring the influence of each frequency component of runoff and the fluctuation of runoff series caused by meteorological factors, which may not effectively capture the potential change process. In this paper, a novel method (CEEMD-MPE-EMD-GRU) for the non-stationary monthly runoff prediction was proposed. It fully took advantages of the complementary ensemble empirical mode decomposition (CEEMD), multi-scale permutation entropy (MPE), empirical mode decomposition (EMD) and gated recurrent unit (GRU). The combined model in general is a data-driven model, and compared with the traditional mechanism model, its most notable advantage is that it successfully overcomes the redundant information of the prediction model. In addition, atmospheric input factor analysis is added on the basis of fully decomposing and identifying non-stationary pseudo-components. The hydrological runoff data (1956–2014) obtained from the Manas River locating at Xinjiang, China were used for prediction. The results indicated that the new CEEMD-MPE-EMD-GRU model reached higher accuracy, as its Nash-Sutcliffe efficiency coefficient (0.960) was significantly larger than those of the GRU model (0.813) and the CEEMD-GRU model (0.889). Meanwhile, the root mean square error and the absolute relative error of the CEEMD-MPE-EMD-GRU model decreased to 0.279 and 0.195, respectively. The new runoff prediction model established in this paper would provide more precise evaluation of the monthly runoff prediction and better guidelines for high-efficiency agricultural water scheduling in the irrigation district. • Introduced CEEMD-MPE-EMD-GRU to predict the monthly series of streamflow. • Optimize the hydrometeorological elements as the input items of the model. • The CEEMD-MPE-EMD-GRU is found to be stable on tendency and extreme streamflow. • It overcomes a lot of steps brought by redundant information of the model. • Effectively improve the accuracy of nonlinear and non-stationary flow simulation. [ABSTRACT FROM AUTHOR]

A forecasting method, easily understood and applied by administrators/policy-makers, is developed in this paper. Based on a simple, but quite realistic assumption, that production in a given region ‘pulls’ or ‘shapes’ the volume of production nationally a co-integration forecasting method is proposed in the paper. Using data from administrative regions, the empirical analysis suggests that the property of co-integration can be used to forecast agricultural production in the short-run. [ABSTRACT FROM AUTHOR]

Accurate forecasting of the agricultural power load is important for rural electricity networks' safe and stable operation. But it is more uncertain and difficult to forecast than industrial, commercial, and residential loads. Multivariate time series prediction methods have problems such as poor refinement of weather services and difficulty in quantifying the influencing factors. Univariate time series prediction methods should be emphasized, and the characteristic information in the historical load data should be fully exploited. Therefore, this paper systematically investigates the decomposition strategy and attention-based recurrent neural network for univariate time-series multi-step prediction. The decomposition strategy is employed to generate derived variable sequences from the target sequence as inputs for the model, while the dual-stage attention mechanism is used for dynamically selecting features. A one-year real-world dataset of agricultural power loads in southern China is used for validation. The effectiveness of two different decomposition algorithms on dual-stage attention-based recurrent neural network (DARNN) is discussed, then the proposed ensemble empirical mode decomposition and dual-stage attention-based recurrent neural network (EEMD-DARNN) is compared with four univariate time series prediction models and three ablation models. The proposed model is optimal in four evaluation indicators, with R 2 , R M S E , M A P E , and M A E of 0.970, 12.298, 0.040, and 7.335 respectively in the shortest forecast horizon (15 min), and 0.941, 17.305, 0.065 and 11.347 respectively in the longest forecast horizon (1 h). It is a significant improvement over the existing models. In conclusion, EEMD-DARNN can achieve more accurate ultra-short-term agricultural power load forecasting. [ABSTRACT FROM AUTHOR]

This paper presents a study focused on the analysis of phytosanitary treatment sales in the Souss Massa region of Morocco. The objective of the study is to predict the sales of agricultural products, particularly crop protection solutions, aiming to optimize supply chain operations and meet customer demand effectively. Data for this study are collected from multiple sources, including the Enterprise Resource Planning (ERP) system called Microsoft Dynamics AXAPTA used by a leading agricultural company operating in the region. Information such as the date of sale, farming type, climate, and specific sales locations within the Sous Massa region is gathered. Machine learning techniques are applied for forecasting. Various regression models, including the Gradient Boosting Regressor algorithm, are employed to determine the most accurate predictor. Evaluation of the models reveals promising results, with a Mean Absolute Error (MAE) of 0.0035 and a Root Mean Square Error (RMSE) of 0.0066. The results obtained by applying various regression models, including the Gradient Boosting Regressor algorithm, demonstrate promising prediction scores. These findings contribute to the field of sales prediction in the agricultural industry while considering the impact of climate conditions, farming practices, and regional factors. [ABSTRACT FROM AUTHOR]

Scientific and accurate prediction of energy production is important for exploring energy alternative paths, adjusting energy structure and industrial layout in a targeted manner, and promoting strategic energy transformation. In this paper, a time-lagged logistic grey forecasting model with harvesting term is established by combining the Logistic model, which has the ability of historical recurrence and short- and medium-term forecasting ability, and the grey forecasting model, which has the characteristics of strong adaptability and small computational effort. The least squares method is used to estimate the parameters of the new model, and the mathematical method is used to calculate the time response equation of the new model, and the modeling steps and flow chart of the new model are obtained. Finally, the oil production data of a province in China is used as the validity analysis of the model, and the new model is compared with two classical grey prediction models, three optimized grey prediction models, and one other prediction model. Through the comparison of seven indicators, the results show that the new model is significantly better than other models. Based on the validity analysis results, the new model is applied to the energy production forecast of three typical provinces in China. The horizontal and vertical comparison shows that the model can effectively predict the three kinds of energy production, and can provide technical support for the strategy and measures to adjust the energy security structure and promote the healthy and sustainable development of the energy industry. • A novel model (HDLGM(1,1,τ)) is established by considering the time-lagged Logistic. • Logistic harvest items are introduced into the novel model. • Modeling short- and medium-term forecasting ability was Enhanced. • The proposed model has been effectively applied to energy production. [ABSTRACT FROM AUTHOR]

Freight transportation has a key role in the economic competitiveness of any nation. India is one of the fastest-growing nations in the world; its agricultural sector plays a vital role in contributing to the country's economy. In this paper, the freight production in the agricultural sector of India is modelled using multiple linear regression (MLR) and generalized additive modelling (GAM) approaches. Using district-level data, factors influencing agricultural freight production are identified and the relationship between them is modelled. The study considered 210 districts of eight states covering South India and parts of Central and Western India. Population, employment in the agricultural sector, gross cropped area, and gross irrigated area are identified as influential factors. GAM approach is a flexible method which can predict non-linear responses for the given predictor variables by allowing non-linear functions for each of them while maintaining additivity. The applicability of GAM in modelling freight production while tackling the non-linear effects of predictor variables on freight production is investigated and encouraging results are achieved. The results revealed that both MLR and GAM models have good modelling efficiency, however, the GAM model outperformed the MLR model in both fitting and predicting. The study can be used in evaluating the commodity movements and transportation demand which in turn will help in decision-making for the provision of freight transportation facilities and policymaking. • Regional level freight generation in India. • Production of agricultural commodities at district level modelled. • Multiple linear regression (MLR) and Generalized additive modeling (GAM) adopted. • GAM performed better than MLR. • Nonlinear relationship between freight generation and predictor variables. [ABSTRACT FROM AUTHOR]

Agricultural production prediction is a challenging task in the deep neural network field. This paper presents a novel approach to fruit production prediction using deep neural networks to build a fast and reliable prediction system for agricultural production. In this article, we have considered different types of fruit production data (apples, bananas, citrus, pears, grapes, and total fruits), analyzed this data, and predicted the future production of these fruits using deep neural networks. The data are taken from the National Bureau of Statistics of Pakistan and the production output of major fruits. We implemented 3 different methods to predict the data for future fruit production. The first method is Levenberg-Marquardt optimization (LM), which was 65.6% accurate; the second method is called scale conjugate gradient back propagation (SCG), which had an accuracy of 70.2%, and the third method, is Bayesian regularization back propagation (BR), which was 76.3% accurate. These methods to predict fruit production are applicable in developing countries because they can compare production with increasing populations and assist in making new policies to increase production. The estimated results reveal that the government of Pakistan needs to further increase fruit production and create better policies for farmers to improve their production. [ABSTRACT FROM AUTHOR]

Summary: The overall goal of the present study is to show through a simulation experiment the value of adaptive management under climatic variability and change. The Northern California water resources management system is used as an example of a multi-agency, multi-site, multi-objective water resources management system that supports a trillion dollar economy to demonstrate the utility of the methods and to exemplify the assessment studies. Simulations are performed using an adaptation of the Integrated Forecast and Reservoir Management (INFORM) system of Northern California with 6-hourly 1.4° input from the CCSM3.0 climate model for historical/control (1970–2019) and future periods (2050–2099). The present paper examines the assessment of the hydroclimatological sensitivities and a companion paper () examines the assessment of reservoir management sensitivities. The future period was based on the moderate CCSM3.0 A1B scenario. The INFORM system includes an intermediate complexity regional model for dynamic downscaling of the CCSM3.0 scenario output to produce surface precipitation and temperature fields with a 10×10km2 resolution, and snow–soil-channel modeling of all the watersheds upstream of the major reservoirs in Northern California. The flow results indicate a shift in the monthly average flow volume toward earlier times in the year and higher flow variability for the future period. Higher average temperatures in the future period and, consequently, earlier snow pack melt are mainly responsible for these flow changes. Both daytime and nighttime average monthly temperatures are higher in the simulation of the future period for the entire domain with the northern catchments experiencing higher temperature increases. Simulated monthly average precipitation for the future period is higher in the southern high Sierras and lower in the northern drainage basins than the historical period precipitation. The uncertainty in the assessments for Northern California mentioned may be estimated with additional simulations similar to those shown in this work using high spatial and temporal resolution output from different climate models. [ABSTRACT FROM AUTHOR]

• New improved drought forecast model incorporating eight synoptic climate drivers. • Multivariate decomposition splits variables into intrinsic mode functions (IMF). • Simulated Annealing (SA) approach selects pertinent IMFs to inform drought pattern. • Kernel Ridge Regression (KRR) algorithm used to design new hybrid drought model. • Hybrid model provides significant hydrological and water management implications. New and improved drought models based on the World Meteorological Organization approved Standardized Precipitation Index, principally at multiple timescale horizons, are providing significant benefits to the hydrological community, by its widespread acceptance in the sub-field of water resources management, sustainable water use and precision agriculture. In this research paper, the existing challenges faced by a drought forecasting model trained at multiple time-scales are resolved where a new multi-phase, multivariate empirical mode decomposition model integrated with simulated annealing and Kernel ridge regression algorithms (i.e., MEMD-SA-KRR) is designed to attain significantly accurate drought forecasts for 3 agricultural sites (i.e., Faisalabad, Islamabad and Jhelum, located in Pakistan). Utilizing the multi-scalar Standardized Precipitation Index (SPI) time series as a target variable for characterization of drought, twelve multivariate datasets (derived from statistically significant lagged combinations of precipitation, temperature & humidity), that are enriched with eight synoptic-scale climate mode indices and periodicity, are utilized in designing a new drought model. The study constructs a hybrid MEMD-SA-KRR model, where firstly, the data are partitioned into their respective training and testing subsets after creating historically lagged SPI at timescale (t – 1). Secondly, the MEMD algorithm is conditioned to demarcate multivariate climate indices from their training and testing sets, separately, into their decomposed intrinsic mode functions (IMFs) and residues. Thirdly, the SA method is employed to decide the most suitable IMFs. Finally, the KRR algorithm is applied to the selected IMFs to forecast multi-scaler SPI, at 1-, 3-, 6- and 12-monthly forecast horizon. The results are benchmarked with Random Forest, integrated with MEMD and SA to develop the MEMD-SA-RF equivalent model. The multi-phase MEMD-SA-KRR model is tested geographically in Pakistan, revealing that the MEMD-SA-KRR hybrid model generates reliable performance in forecasting multi-scaler SPI series, relative to the comparative models based on error analysis metrics. The hybrid drought model incorporating the most pertinent synoptic-scale climate drivers, as the model inputs has significant implications for hydrological applications and water resources management including its potential use in drought policy and drought recovery plans. [ABSTRACT FROM AUTHOR]

• Machine learning algorithms have been applied to estimate wheat yield for Australia. • Combining climate and satellite data achieves high performance for yield prediction. • The unique and overlapping information of climate and satellite data is quantified. • Optimal prediction performance can be achieved two-month lead time before maturity. Wheat is the most important staple crop grown in Australia, and Australia is one of the top wheat exporting countries globally. Timely and reliable wheat yield prediction in Australia is important for regional and global food security. Prior studies use either climate data, or satellite data, or a combination of these two to build empirical models to predict crop yield. However, though the performance of yield prediction using empirical methods is improved by combining the use of climate and satellite data, the contributions from different data sources are still not clear. In addition, how the regression-based methods compare with various machine-learning based methods in their performance in yield prediction is also not well understood and needs in-depth investigation. This work integrated various sources of data to predict wheat yield across Australia from 2000 to 2014 at the statistical division (SD) level. We adopted a well-known regression method (LASSO, as a benchmark) and three mainstream machine learning methods (support vector machine, random forest, and neural network) to build various empirical models for yield prediction. For satellite data, we used the enhanced vegetation index (EVI) from MODIS and solar-induced chlorophyll fluorescence (SIF) from GOME-2 and SCIAMACHY as metrics to approximate crop productivity. The machine-learning based methods outperform the regression method in modeling crop yield. Our results confirm that combining climate and satellite data can achieve high performance of yield prediction at the SD level (R2 ˜ 0.75). The satellite data track crop growth condition and gradually capture the variability of yield evolving with the growing season, and their contributions to yield prediction usually saturate at the peak of the growing season. Climate data provide extra and unique information beyond what the satellite data have offered for yield prediction, and our empirical modeling work shows the added values of climate variables exist across the whole season, not only at some certain stages. We also find that using EVI as an input can achieve better performance in yield prediction than SIF, primarily due to the large noise in the satellite-based SIF data (i.e. coarse resolution in both space and time). In addition, we also explored the potential for timely wheat yield prediction in Australia, and we can achieve the optimal prediction performance with approximately two-month lead time before wheat maturity. The proposed methodology in this paper can be extended to different crops and different regions for crop yield prediction. [ABSTRACT FROM AUTHOR]

Annual forecasts for the Australian macadamia crop have been issued since 2001, with varying (and not always improving) degrees of accuracy. Regression models using climate variables have formed the basis for these forecasts, with general linear model (GLM) ensembles being adopted more recently. This has proven to be a challenging task, as there are only a small number of observations (18) combined with a large number (90+) of independent variables – these being different climate measures for different times of the year (representing 'key physiological periods for macadamia trees'). Also, these 'assumedly-independent' variables contain various degrees of correlation. This study uses cross-validation, with the most recent data for the two dominant production regions of Australia (Lismore and Bundaberg), to investigate the relative performance of alternate modelling methods. These modelling methods were GLMs, partial least squares (PLS) regression and LASSO (least absolute shrinkage and selection operator) penalised regression. Model ensembles, which have been shown to be beneficial in many alternate disciplines, are used to advantage. Both GLMs and PLS produced quite-disappointing results, failing to meet the project's benchmarked accuracy of ±10% error. The optimal LASSO models performed notably better, with a further improvement when ensembles were incorporated. The lowest mean absolute error (MAE) rates here were 9.0% for Lismore and 5.9% for Bundaberg. Hence LASSO ensembles will be adopted for future forecasts of the Australian macadamia crop. • Since 2001 crop forecasts have been provided for the Australian industry to facilitate macadamia operations. • This paper reviews this process, and conducts a cross-validation comparison of some of the more modern regression methods. • These are also incorporated in with the new ensemble modelling methods, also known as 'wisdom of the crowds'. • Results from the ensemble LASSO penalised regressions show a marked improvement over the previous methods. • This method is now being adopted for the 2019 macadamia crop forecast. [ABSTRACT FROM AUTHOR]

Abstract Regional crop production forecasting is growing in importance in both, the public and private sectors to ensure food security, optimize agricultural management practices and use of resources, and anticipate market fluctuations. Thus, a model and data driven, easy-to-use forecasting and a risk assessment system can be an essential tool for end-users at different levels. This paper provides an overview of the approaches, algorithms, design, and capabilities of the CCAFS Regional Agricultural Forecasting Toolbox (CRAFT) for gridded crop modeling and yield forecasting along with risk analysis and climate impact studies. CRAFT is a flexible and adaptable software platform designed with a user-friendly interface to produce multiple simulation scenarios, maps, and interactive visualizations using a crop engine that can run the pre-installed crop models DSSAT, APSIM, and SARRA-H, in concert with the Climate Predictability Tool (CPT) for seasonal climate forecasts. Its integrated and modular design allows for easy adaptation of the system to different regional and scientific domains. CRAFT requires gridded input data to run the crop simulations on spatial scales of 5 and 30 arc-minutes. Case studies for South Asia for two crops, including wheat and rice, shows its potential application for risk assessment and in-season yield forecasting. Highlights • An integrated modeling framework, CRAFT, has been developed for gridded simulations at 5 and 30 arc-minutes resolutions. • It allows within-season yield forecasting, risk analysis, and climate change impact studies. • It incorporates seasonal climate forecasts using the Climate Predictability Tool. • The current version can run the pre-installed DSSAT, APSIM, and SARRA-H, but support for any crop model can be added. • CRAFT is a desktop application that can produce multiple simulation scenarios, maps, and interactive visualizations. [ABSTRACT FROM AUTHOR]

Abstract Global and regional scale agricultural monitoring systems aim to provide up-to-date information regarding food production to different actors and decision makers in support of global and national food security. To help reduce price volatility of the kind experienced between 2007 and 2011, a global system of agricultural monitoring systems is needed to ensure the coordinated flow of information in a timely manner for early warning purposes. A number of systems now exist that fill this role. This paper provides an overview of the eight main global and regional scale agricultural monitoring systems currently in operation and compares them based on the input data and models used, the outputs produced and other characteristics such as the role of the analyst, their interaction with other systems and the geographical scale at which they operate. Despite improvements in access to high resolution satellite imagery over the last decade and the use of numerous remote-sensing based products by the different systems, there are still fundamental gaps. Based on a questionnaire, discussions with the system experts and the literature, we present the main gaps in the data and in the methods. Finally, we propose some recommendations for addressing these gaps through ongoing improvements in remote sensing, harnessing new and innovative data streams and the continued sharing of more and more data. Highlights • Eight global and regional agricultural monitoring systems are compared • Gaps in data are described, where cropland maps, crop calendars and meteorological data are viewed as the most critical • Gaps in methods include the need for better predictions of yield and crop production and how to operationalize research methods • New sources of remote sensing data, greater data sharing and new sources of information, e.g. from mobile phones, will all help to improve agricultural monitoring [ABSTRACT FROM AUTHOR]

Abstract Each month the JRC issues the MARS Bulletin detailing the agro-meteorological and expert analysis underpinning the assessment of European crops' status and yield forecasts. In this context a resume is provided to give an overview on the geographical distribution of eventual crop damages. The MARS Bulletin provides such information in a set of synthetic maps (Areas of Concern), produced in each Bulletin, depicting extreme weather events and their impact on crops that have occurred in Europe during the analysis period. The present article describes the mix of quantitative and qualitative datasets and methodologies that drive the delineation of the Areas of Concern (AOC) maps and evaluates their capability to resemble crop production losses. The quantitative analysis is based on the Mars Crop Yield Forecasting System (MCYFS) indicators coming from meteorological models, crop growth models and remote sensing data. Indicators are considered in absolute and relative terms and in their relation with standard statistical metrics. The outcome of the quantitative analysis is a set of potential Areas of Concern. Experts' judgment is thus necessary to discriminate potential results through a qualitative analysis focused on: past occurred events and climatologic conditions; agro-management practices; regional agricultural systems peculiarity and their historical resilience and resistance to adverse conditions. In this article the experts' judgment of the team of current MARS analysts, as used in the AOC analysis, is translated into a warning index. Such index condenses the specific contribution to the final production of each development stage and the adverse agrometeorological events occurred, as depicted into the AOC maps. The warning index is computed at country scale for the past five agricultural seasons, from season 2011-2012 to season 2015-2016. Two crops are considered, winter wheat and grain maize as proxy for winter and summer crop groups. The warning indexes calculated are then compared to the national production in a qualitative way. To support the analysis few study cases are presented. The findings of this article highlight that the events depicted in the AOC maps are informative about production losses and specific knowledge is needed to full understand the information carried. Highlights • The comparison of the issued warnings through AOC analysis shows some good agreements with production anomalies. • AOC analysis properly assessed most of the winter wheat production anomalies. • AOC analysis could provide warning relevant at regional scale only. • AOC analysis tends to underestimate unfavorable conditions at harvest. • This paper represents the first attempt to formalize the tacit knowledge within the group of MARS analysts. [ABSTRACT FROM AUTHOR]

Abstract This paper presents an evaluation framework followed to identify cost-efficient alternatives to extend the MARS Crop Yield Forecasting System (MCYFS), run by the European Commission Joint Research Centre since 1992, to other main producing areas of the world: Eastern European Neighbourhood, Asia, Australia, South America and North America. These new systems would follow the principles and components of the MCYFS Europe: a meteorological data infrastructure, a remote sensing data infrastructure, a crop modelling platform, statistical tools, a team of analysts and a crop area estimation component. The framework designed evaluates the performance of the possible MCYFS-like system realizations against six defined objectives and their costs. Possible monitoring systems are based on a combination of different technical solutions for each of the MCYFS components, and are evaluated through an automatic algorithm that calculates the expected system performance –relying on a priori expert judgement–, the costs, and possible risks to construct some technical solutions, to finally identify the cost-efficient ones. A baseline system, achieving the minimum required performance, was identified as the most efficient starting point for the MCYFS extension in all the geographical areas. Such system would be built upon: (i) near real-time reanalysis meteorological products; (ii) remote sensing data from low-resolution (~1 km) platforms with a long-term product archive; (iii) crop models based on crop-specific model calibration from experimental data published in scientific literature; (iv) statistical methods based on trend and regression analysis applied to national level; (v) a team of analysts with specific technical profiles (on meteorology, remote sensing, and agronomy); and (vi) digital classification of very high resolution imagery supported by non-expensive ground surveys for area estimation. In countries where accessibility to local data and resources is high the baseline system can be upgraded enhancing some of the components: sub-national statistical analysis with additional statistical methods like multiple regression or scenario analysis; recruitment of experts on local agricultural conditions in the team of analysts; local calibration of crop models with experimental data; and exploiting high and low resolution biophysical products from remote sensing for crop monitoring. Highlights • We present a framework to identify cost-efficient solutions for a crop yield forecasting system in the main producing areas. • An automatic algorithm evaluates the expected performance, costs, and possible risks to construct some technical solutions. • The analysis identified a baseline system common to all the geographical areas reaching the minimum required performance. • A roadmap to upgrade the baseline system indicates those solutions improving the system performance at lower costs. [ABSTRACT FROM AUTHOR]

Abstract Since 1993, the JRC has put in operation a crop monitoring and yield forecasting system for Europe, the results of which are published in the JRC MARS Bulletin (currently every month). This paper outlines how the agro-meteorological analyses, country-specific overviews of crop conditions, and crop yield forecasts reported in the Bulletin are used and how these respond to the diverse needs of different types of stakeholders. Stakeholders from more than 32 countries download the JRC MARS Bulletin, in peak-season up to 1500 downloads occur in the first days after publication. The readership of the Bulletin is diverse, coming from governments (e.g. Ministries of Agriculture), private companies (e.g. commodity traders, banking), media, and research and academia. On the list of stakeholders that want to be notified of the release of the Bulletin, roughly 37% originate from business, 35% from research and development, 22% from government, and 6% from the media. The primary user is the Directorate General for Agriculture and Rural Development (DG-AGRI) of the European Commission , which uses the forecasts to quantify the production estimates for crop supply balance sheets and to identify regions with exceptional (mostly weather related) challenges that might require a policy response. The information for wheat, maize and rice is shared through the Agricultural Market Information System (AMIS) thus contributing to increased global market transparency and better governance of agriculture and food policies. The largest business use is in market information, financial, and consultancy services, followed by commodity trading. Examples of use in media reports as well as online feedback to those, e.g. by farmer's organizations, are also presented. Downloads of the Bulletin peak in the month before harvest at the time when the forecasts can be of most value for stakeholder decision-making. Highlights • The primary user of crop yield forecasts & the JRC MARS Bulletin is DG-AGRI. • Stakeholders from more than 32 countries download the JRC MARS Bulletin. • Stakeholders are from business (37%), R&D (35%), government (22%), and media (6%). • Bulletin & yield forecast downloads (up to ~1500) peak in the month before harvest. • Knowledge gaps on actual use, perceived usefulness, impacts of crop forecasts exist. [ABSTRACT FROM AUTHOR]

Abstract: Aiming to solve the problem of crop yield forecasting and accurate fertilization, this paper presents a method based on optimized design ideas about fertilization decision-making. The algorithm is based on a variety of fertilizer function fittings, and weighting optimal solutions to obtain the fertilizing amount. The algorithm in this paper which uses the interpolation algorithm improves the rationality and reliability of fertilization by solving the problem of which need large-scale test, and large computation so as to become difficult to implement. This algorithm is composed of three sub-programs. One or some of its sub-programs can be used flexibly. The goals to access better economic efficiency with less inputs can be achieved. [Copyright &y& Elsevier]

• A new multi-task prediction algoritom TPMv2 was developed. • It estimates 3D pose of tomato bunch based on 2D- and 3D- Bounding Box and Keypoints prediction. • TPMv2 can predict Keypoints on partially visible plant organ from coarse data. The automatic harvesting of tomatoes has been achieved for many years in the laboratory. The new research topic is harvesting the tomato more flexibly and nondestructively at any tomato bunch pose according to the agronomic demands. Although the tomato pose can be predicted by keypoints detection, the poor data quality of commercial RGBD cameras, occlusion between plant organs, various tomato poses, and unstructured working environments pose some challenges to the tomato bunch pose detection. Therefore, our research proposed an improved version of the Tomato Pose Method (TPM), namely TPMv2, which is a two-stage end-to-end multi-task network. This network provides comprehensive information on the tomato bunch, including the positions and poses of the stem, peduncle, and fruits, by predicting the two-dimensional bounding box (2D BBox), three-dimensional bounding box (3D BBox), two-dimensional key point (2D Kpt), and three-dimensional key point (3D Kpt). Aiming at the problems of occlusion and poor-quality point cloud, this paper specially designs a key point network (KPN) for tomatoes, where a keypoints processing pipeline was innovatively proposed, improving the accuracy of key point positioning and reducing abnormal prediction effectively. TPMv2 makes it possible to detect tomato bunch pose precisely with an economical camera, avoiding dangerous situations caused by abnormal prediction. The precision of 2D BBox and 3D BBox reached 0.9372 and 0.8700, and the Percentage of correct Keypoints (PCK) of 2D Kpt and 3D Kpt reached 0.8882 and 0.7836. About 78.36 % of 3D Kpts' positioning errors are less than 20 mm, sufficient to describe a correct pose trend based on the 3D Kpt, benefiting the manipulator to plan a more reasonable trajectory for non-destructive harvesting. [ABSTRACT FROM AUTHOR]

This paper attempts to analyze the variations in technical efficiency of the individual farmers of distinct (small, medium and large) size of land holdings and different types (family and hired) of labor with given set of input variables for the agricultural production across different Agro-Climatic Zones. A field survey was conducted to collect the information from a randomly stratified sample of 300 farmers operating in six districts falls under Trans-Gangetic Plains and Semi Sandy Desert zones of the states of Punjab and Haryana in India. A model of stochastic frontier analysis for cross sectional data with half normal truncated features has been used to measure the technical efficiency under the assumptions of Cobb-Douglas production function. The model comprises eight dummy variables for the farmers owning distinct size of land and using distinct type of labor in their production. The results for the estimates of technical inefficiency with family and hired labor depict; small farmers are 50.30 and 68.11 per cent, medium farmers are 61.09 and 67.75 per cent, large farmers are 71.03 and 56.67 per cent and average farmers are 71.94 and 70.07 per cent inefficient respectively. This paper helps the farmers and the policy makers in identifying the second best-alternative crop for obtaining the sustainable agricultural production. [Display omitted] • Ratio of land to labor applied by farmers lead to loss of income and productivity. • The demand for food is increasing, motivating the importance of farmers for sustainable agricultural production and development. • Sustainable agricultural production can deal with help is required to meet the growing demand for food. • Kernel density function measures variations in technical inefficiency of individual farmers. • Second-best alternative crops must be suggested to farmers for cultivation. [ABSTRACT FROM AUTHOR]

Spatial awareness and memory are key factors for a robot to evolve in semi-structured and dynamic environments as those found in agriculture, and particularly in fruit crops where the trees are regularly distributed. This paper proposes a probabilistic method for mapping out-of-structure objects (weeds, workers, machines, fallen branches, etc.) using a Kernel density estimator. The methodology has theoretical and practical advantages over the well-known occupancy grid map estimator such as optimization of storage resources, online update, high resolution, and straightforward adaptability to dynamic environments. An example application would be a control scheme through which a robot is able to perform cautious navigation in areas with high probability of finding obstacles. Simulations and experiments show that large extensions can be online mapped with few data and high spatial resolution. [ABSTRACT FROM AUTHOR]

In this paper, we evaluated the performance of the model AquaCrop for crop yield forecasting in the Brazilian semiarid (BSA) using meteorological observation and Eta model seasonal climate forecasts as input data. The study area is characterized by low rainfall that is poorly distributed throughout the rainy season; thus, the region's agricultural productivity is vulnerable to climate conditions. AquaCrop was first calibrated using field experiments and subsequently applied to simulate an operational crop yield forecast system for maize under rainfed conditions. Simulations were performed with daily data for 37 growing seasons for the period 2001–2010. The seasonal climate forecast was used in combination with observed meteorological data to anticipate the crop forecast. Soil characteristics were derived from pedotransfer functions (PTFs). We were able to demonstrate the ability of the seasonal crop yield forecast system to provide timely and accurate information about maize yield at least 30 days in advance of the harvest. The development of improved crop yield forecasting system is crucial for implementing drought-preparedness measures in the BSA region. [ABSTRACT FROM AUTHOR]

Precise evapotranspiration (ET) estimation is critical for agricultural water management, particularly in water-stressed developing countries. Vapor Pressure Deficit is one of the ET parameters that has a significant impact on its calculation (VPD). This paper forecasts VPD using ensemble learning-based modeling in eight different regions (Dakahliyah, Gharbiyah, Kafr Elsheikh, Dumyat, Port Said, Ismailia, Sharqiyah, and Qalubiyah) in Egypt. In this study, six machine learning algorithms were used: Linear Regression (LR), Additive regression trees (ART), Random SubSpace (RSS), Random Forest (RF), Reduced Error Pruning Tree (REPTree), and Quinlan's M5 algorithm (M5P). Monthly vapor pressure data were obtained from the Japanese 55-year Reanalysis JRA-55 from 1958 to 2021. The dateset has been divided into two segments: the training stage (1958–2005) and the testing stage (2006–2021). Five statistical measures were used to evaluate the model performances: Correlation Coefficient (CC), Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Relative absolute error (RAE), and Root Relative Squared Error (RRSE), across both training and testing stages. RF model outperformed the rest of the models [CC = 0.9694; MAE = 0.0967; RMSE = 0.1252; RAE (%) = 21.7297 and RRSE (%) = 24.0356], followed closely by REPTree and RSS models. On the other hand, M5P model performance remained moderate and both LR and AR model were the worst. During the testing stage, RF outperformed the rest of the models in terms of (which statistic), followed closely by REPTree and RSS models. On the other hand, M5P performance remained moderate and both LR and AR models were the worst. This study recommended using the RF model for future hydro-climatological studies in general, and vapor pressure deficit modeling and prediction in particular. This study enables future magnitudes to be predicted, alerting the authorities and administrators involved to focus their policy-making on more specific pathways toward climate adaptation. • Six machine learning models (LR, BART, RSS, RF, REPTree, and M5P) were developed for modeling VPD. • Long time series of eight Egyptian regions was used for estimation of VPD. • The RF model outperformed the rest models. • The performance of M5P and LR models was the second rank, while RSS model performance remained moderate. [ABSTRACT FROM AUTHOR]

This paper established hybrid prediction models based on variational mode decomposition (VMD), empirical mode decomposition (EMD), and ensemble empirical mode decomposition (EEMD) combined with the backpropagation neural network model (BPNN) to improve the prediction accuracy of reference crop evapotranspiration (ET 0) time series with the characteristics of nonlinearity and instability. The daily ET 0 data of 11 representative stations in Xinjiang from 1993 to 2016 were selected for model training and testing and compared with the results of support vector regression (SVR) and gradient boosting regression tree (GBRT) as the two empirical machine learning models. The results indicated the superiority of the VMD-BPNN hybrid prediction model to EMD-BPNN and EEMD-BPNN in terms of accuracy and stability, with root mean square error (RMSE) = 0.405 mm/d, mean absolute error (MAE) = 0.268 mm/d, and coefficient of determination (R2) = 0.979. When employing the VMD-BPNN model to forecast ET 0 for seven days, the RMSE and Nash-Sutcliffe efficiency coefficient (NSE) of the VMD-BPNN model were 0.588 mm/d and 0.952, respectively, and the prediction results indicated high precision and reliability. The prediction accuracy of the VMD-BPNN model was significantly higher than that of single machine learning models, such as BPNN, SVR, and GBRT. The RMSE and MAE values of the VMD-BPNN model were more than 60% smaller than BPNN, SVR, and GBRT models, and R2 and NSE were approximately 18% higher than BPNN, SVR, and GBRT models, respectively. This demonstrates the effectiveness of the VMD method in reducing the non-stationarity of the original daily ET 0 data. The BPNN model predicted the decomposed data series, and the prediction accuracy and stability were significantly enhanced. This indicates the high reliability of the VMD-BPNN model and its capability for ET 0 prediction in Xinjiang. • The VMD-BPNN model is the best among three hybrid models in predicting daily ET 0. • The VMD-BPNN model performs good on sites with different spatial distributions. • The VMD-BPNN model had higher accuracy in predicting 7-day ET 0. • Hybrid prediction model has higher accuracy than the BPNN, SVR, GBRT model. [ABSTRACT FROM AUTHOR]

• This paper is focused on predicting price signals for energy arbitrage of storage systems. • A methodology is presented based on feature selection tools and classification techniques. • A Li-Ion battery in the Ontario competitive electricity market is considered as a case study. • Results indicate that the proposed methodology outperforms alternative point forecasting models. When participating in a competitive electricity market, energy storage systems could deliver various services and stack multiple revenue streams. One potential venue for gaining profit is energy arbitrage, i.e., to harvest the price differential that might exist in some markets between the peak and off-peak hours. To take advantage of price arbitrage, however, it is necessary to have an insight into the price fluctuations of upcoming hours. In this paper, we propose a method for generating predictive electricity price signals to help storage operators make arbitrage decisions. The proposed method delivers signals that are integrated into an optimization platform to schedule the arbitrage operation of a storage system. We examine a Lithium-ion battery as the choice of storage system and the Ontario's competitive electricity market in Canada is considered as the case study. We compare the performance of the proposed approach against the cases that alternative point forecasts of the electricity price are used for self-scheduling of the storage system. The results show that the obtained arbitrage profit of the storage system can be increased significantly using the proposed approach. [ABSTRACT FROM AUTHOR]

Rice production encountered several challenges, especially among rural farmers in Northern Ghana. The current debate encompasses the belief of cultural norms and religious values as the right theory for the sustainability of agriculture. The belief in abstracts is prioritised to the adoption of improved rice. This paper examined the value attached to traditional rice varieties as a result of societal beliefs, and theoretically underpinned by the Dual-system hypothesis. The finding from three ethnic and religious groups through a survey indicates heterogeneity in the results but also demonstrates the factual importance of culture and religion. This paper, therefore, proposes a new way of categorising farming; as a business and a belief and that policy stream aim at small scale, rural farmers with belief peculiarities needs to adopt cultural and religious strategies to promote adoption. • Little is know about the value farmers attached to culture and religion. • Cultural and religious values affect the adoption of agricultural technology. • Chiefs and religious leaders are important in the diffusion of innovation. • Accessibility and affordability is less priority relative to social norms to farmers. • Exploring Psychology Dual-system theory in Agricultural Economics. [ABSTRACT FROM AUTHOR]

Water footprint forecasting is essential to measuring the embodied water resource consumption and achieving the sustainable water governance. Agricultural sector is conventionally a water intensive sector and accounts for large amount of water consumption in the river basins. In this paper, a system dynamics model is combined with Markov Chain, considering economic development, agriculture water consumption, population and agricultural ecosystem, to forecast the total agricultural water footprint (AWF) as well as its pressure on the freshwater ecosystem. Wheat, coin, potato, alfalfa, vegetables and flax are chosen as representative crops for AWF accounting in the integrated model. A case study of the Heihe River Basin in China during 2010–2030 shows that, the AWFs are 9.67 × 10 8 m 3 , 1.02 × 10 9 m 3 , 1.05 × 10 9 m 3 and 9.27 × 10 8 m 3 under Baseline Scenario, Moderate Risk Scenario, High Risk Scenario and Sustainable Scenario, respectively. It is concluded that the improvement on agricultural water efficiency may decrease the AWF, which can be achieved by agricultural water conservation, irrigation canal construction, maintenance funding and investments, agricultural planting adjustment, and virtual water strategies. [ABSTRACT FROM AUTHOR]

• Deep learning models learn plausible relationships between features and crop yield. • Neural networks learn features that perform similar to expert-designed features. • Feature influence on yield matches expert knowledge when model accuracy is high. Machine learning models for crop yield forecasting often rely on expert-designed features or predictors. The effectiveness and interpretability of these handcrafted features depends on the expertise of the people designing them. Neural networks have the ability to learn features directly from input data and train the feature learning and prediction steps simultaneously. In this paper, we evaluate the performance and interpretability of neural network models for crop yield forecasting using data from the MARS Crop Yield Forecasting System of the European Commission's Joint Research Centre. The selected neural networks can handle sequential or time series data and include long short-term memory (LSTM) recurrent neural network and 1-dimensional convolutional neural network (1DCNN). Performance was compared with a linear trend model and a Gradient-Boosted Decision Trees (GBDT) model, trained using hand-designed features. Feature importance scores of input variables were computed using feature attribution methods and were analyzed by crop yield modeling and agronomy experts. Results showed that LSTM models perform statistically better than GBDT models for soft wheat in Germany and similar to GBDT models for all other case studies. In addition, LSTM models captured the effect of yield trend, static features (e.g. elevation, soil water holding capacity) and biomass features on crop yield well, but struggled to capture the impact of extreme temperature and moisture conditions. Our work shows the potential of deep learning to automatically learn features and produce reliable crop yield forecasts, and highlights the importance and challenges of involving human stakeholders in assessing model interpretability. [ABSTRACT FROM AUTHOR]

• This paper presents a novel logarithmic mean Divisia index decomposition framework. • It contributes four new models to the existing LMDI model family. • First LMDI models that separate quantity and price effects. • Four case studies apply the LMDI framework to the total RUEC indicator. • This paper provides a new tool for monitoring the energy transition. This paper presents a novel logarithmic mean Divisia index (LMDI) decomposition framework that is tailor-made for unit cost indicators. It adds four new models to the existing LMDI model family. The main novelty of the new framework lies in the separation of quantity and price effects captured in unit cost indicators, while retaining the same desirable properties of traditional models. Four case studies apply the novel LMDI framework to the total real unit energy costs (total RUEC) indicator. Total RUEC represents the sum of direct energy costs (for energy products) and indirect energy costs (energy costs embedded in intermediate inputs and passed on along global value chains) as a fraction of value added. This yardstick allows for monitoring shifts in the burden of energy costs on industries. The first three case studies, based on the World Input-Output Database, cover the period between 1995 and 2009. For an up-to-date analysis, a fourth case study collects additional data for 2009–2016 from energy and economic statistics' institutions. Globally, up until 2009, rising costs for crude petroleum/natural gas and the rise of China in the global economy were the largest drivers of total RUEC. In general, increases of indirect energy costs were more substantial than were those of direct energy costs. The total RUEC of Chinese car manufacturers increased much more strongly than did that of American car manufacturers. After 2009 (until 2016), prices for crude petroleum/natural gas and value added generation were major decelerating factors of global direct RUEC, while increases in energy consumption had offsetting effects. This paper provides a suitable tool to scientists who want to build on unit cost indicators in their research in general and to all policy-oriented institutions concerned with monitoring and analysing the energy transition in particular. [ABSTRACT FROM AUTHOR]

• Forecast crop yield accurately by using dataset expanded and corrected by yield data compensation method. • Calculate the effects of different environmental factors on yield of different crop varieties and identify the key growth periods affecting crop yield. • Promote crop varieties according to comprehensive analysis of the promoted variety yield, preferences for different meteorological factors and the climatic characteristics of the target ecological regions. Forecasting crop variety yield, selecting suitable ecological regions for designated crop varieties, and finding high-yield crop varieties for designated ecological regions are all critical aspects of crop variety popularization. A crop variety yield prediction system based on variety yield data compensation (CVYPS-VYDC) is described in this paper. This novel system is used to predict the yield of designated crop varieties using the large amount of breeding data stored in the Golden Seed Breeding Cloud Platform and meteorological data from the National Meteorological Science Data Center. The prediction system was devised to (a) forecast the yield of designated crop varieties in the target ecological regions, (b) calculate the effects of different environmental factors on the yield of different crop varieties, (c) identify the key growth periods affecting crop yield, and (d) promote crop varieties according to a comprehensive analysis of the promoted variety yield, preferences for different meteorological factors, and climatic characteristics of the target ecological regions. In addition to meteorological factors, soil factors affect crop yield. However, large-scale measurements of soil composition require a large number of sensors and considerable human resources. By using the yield data compensation method, the yield data of predicted varieties across different regions were expanded and corrected, which reduces the yield prediction error caused by soil differences among regions to some extent. In addition, the coefficient of determination (R 2) between actual and predicted yield was improved to 0.82. In addition to predicting the yield of 90 maize varieties in the Huang-Huai-Hai area of China, the preferences of different maize varieties for different environmental factors were analyzed at the same time. CVYPS-VYDC identified the key environmental factors greatly impacting the maize yield, and the proportion by which different factors affected the yield prediction of different maize varieties was determined. Moreover, the growth periods that have great impact on maize yield were identified by analyzing 39 cultivation sites of various maize varieties in the Huang-Huai-Hai area of China. [ABSTRACT FROM AUTHOR]

Agriculture is the mainstay of many developing economies, and successful production is intricately linked to food security, economic development, and regional stability. Estimates of crop yield for strategic grain crops, such as maize (Zea mays L.) have been used in national food security planning to develop response strategies in years of shortfalls and secure markets in years of surplus. Past studies have shown that despite the potential of models in maize crop yield assessment, they have not been effectively used in understanding seasonal and annual production dynamics. Thus, stakeholders require the availability of accurate and timely data on maize production potential and hence the development and application of crop yield models for maize yield estimation. However, current methods of assessing maize crop yields are based on field assessments, which are expensive, laborious and inaccurate. This mixed methods paper, therefore, aimed to; (i) review information sources for maize crop yield assessments, looking at their strengths, limitations, and potential for application in sub-Saharan Africa, (ii) perform trend and distribution analyses of publications in maize crop yield simulation, and (iii) discuss the challenges in the application of models in agriculture planning in the African agriculture systems. The general aim was to understand these crop yield assessments and the current approaches in maize yield estimation methods, and their potential use in the early warning system. The study narrowed the review to crop growth simulation models and explored the different growth simulation models and their potential integration into real-time monitoring frameworks for grain crop assessments. It was observed in this review, using graphical presentation of trend and distribution analysis of one thousand three hundred and thirty-thre scientific publications, that there is an increase in research interest in crop simulation modelling, with current research being done mostly in developed countries. However, the application of models in maize crop yield assessment is dependent on the availability of data, modelled crop characteristics, model calibration requirements, technical capacity, and model implementation costs. Therefore, it was concluded that using crop yield estimation models integrating remote sensing is an important step in local, national and regional agricultural planning in the sub-region and beyond. • Agriculture is intricately linked to food security, economic development, and regional stability. • Estimates of crop yield for strategic grain crops have been used in national food security planning. • Current methods of assessing maize crop yields are based on field assessments. • We reviewed information sources for maize crop yield assessments. [ABSTRACT FROM AUTHOR]

The root zone soil moisture (RZSM) is essential for monitoring and forecasting agricultural, hydrological, and meteorological systems. Accordingly, researchers are determined to improve robust machine learning (ML) models to increase the accuracy of the RZSM predictions. This paper designed new complementary forecasting paradigms hybridizing Empirical Wavelet Transform (EWT) and two modern ensemble-based ML models, namely, extreme gradient boosting (XGBoost) and categorical boosting (CatBoost), to forecast long-term multi-step ahead daily RZSM in very cold and very warm semi-arid regions of Iran. For this purpose, the required datasets consisting of soil properties and meteorological information were extracted from the satellite datasets during 2005–2020 for Ardabil and Minab sites. Afterward, the significant lags of RZSM time series and optimal influence candidate inputs were sought based on the partial autocorrelation function (PACF) and mutual information techniques, respectively. Selected lagged components of RZSM time series were decomposed using EWT into different sub-sequences and consequently concatenated with candidate inputs to feed the ensemble ML models to forecast one-, three-, and seven-day-ahead RZSM at each case study. The performance of EWT-CatBoost and EWT-XGBoost and their counterpart standalone approaches was firstly evaluated in forecasting one-, three-, and seven-day-ahead RZSM using satellite data in this study and their accuracy were compared with a standalone kernel ridge regression (KRR) and complementary EWT-KRR models based on several statistical metrics (e.g., correlation coefficient (R), root mean square error (RMSE), Nash–Sutcliffe model efficiency coefficient (NSE)) and diagnostic analysis. The outcomes of testing phase in Ardabil site ascertained that the EWT-CatBoost (for RZSM(t + 1), R= 0.9979, RMSE= 0.0019, and NSE= 0.9985; for RZSM(t + 3), R= 0.9934, RMSE= 0.0035, and NSE= 0.9885; for RZSM(t + 7), R= 0.9489, RMSE= 0.0109, and NSE= 0.8634) outperformed the other models. On the other hand, the EWT-XGBoost model according to its best results (for RZSM(t + 1), R= 0.9911, RMSE= 0.0064, and NSE= 0.9805; for RZSM(t + 3), R= 0.9807, RMSE= 0.0092, and NSE= 0.9589; for RZSM(t + 7), R= 0.9680, RMSE= 0.0120, and NSE= 0.9309) yielded the most promising accuracy in forecasting multi-step ahead daily RZSM followed by the EWT-CatBoost, and EWT-KRR, respectively. [Display omitted] • Multi-step ahead forecasting the daily root zone soil moisture. • Developing two new complementary ensemble models; EWT-CatBoost and EWT-XGBoost. • Application of microwave satellite (SMAP) datasets for two regions of Iran. • Validating the models of RZSM in very cold and very warm semi-arid climates. [ABSTRACT FROM AUTHOR]

• An adaptive feature fusion pyramid network is developed to extract the optimal features. • A feature augmentation module is introduced for reducing the information loss of pest features. • A multi-stage detector is introduced for recognizing and classifying multi-classes pests. The accurate and robust crop pest detection system is an important step to enable the reliable forecasting of agricultural pest in the community of precision agriculture, attracting great attention in many countries. For achieving the automatic recognition and detection of agricultural pest, previous methods adopt image processing-based methods, leading to lower efficiency. Then, machine vision-based methods are introduced into crop pest detection by using hand-crafted feature descriptors, improving the detection precision and speed. However, the manual feature is powerless for precise recognition. Considering powerful ability of feature extraction of convolutional neural network(CNN), we have developed a CNN-based method for multi-classes pest detection under complex scenes. In this paper, an adaptive feature fusion is introduced into feature pyramid network for extracting richer pest features. Then, an adaptive augmentation module has been developed for reducing the information loss of the highest-level feature maps. Finally, a two-stage region-based convolutional neural network (R-CNN) was built for refining predicted bounding boxes, which can obtain the categories and locations of pests of each image. We have conducted large quantities of comparison experiments on AgriPest21 dataset. Our method could achieve an accuracy of 77.0%, which significantly outperforms other state-of-the-art methods, including SSD, RetinaNet, FPN, Dynamic R-CNN, and Cascade R-CNN. [ABSTRACT FROM AUTHOR]

• Rules to attribute APSIM's yield forecast skill to seasonal climate forecasts (SCF). • Use ECPP and Schaake shuffle to downscale four climate variables suitable for APSIM. • Simulate yield forecasts for 50 stations in 23 years with 1 to 6 months lead times. • ECPP outperforms QM, raw SCFs and climatology especially for early-season forecasts. • As an early-season yield forecast alternative to bring SCFs closer to agriculture. Seasonal climate forecasts (SCF) are evolving rapidly alongside improvements in climate modelling and downscaling research, and have great potential for weather-sensitive sectors, especially agriculture, by reducing weather-related risks and increasing productivity. Skilful yield forecasts at the beginning of, or before, a cropping season can provide farmers and other stakeholders in agribusiness with the necessary information for early planning and actions. Only a few yield forecast studies have a forecast lead time of four months or longer due to the problem complexity. To enable SCFs from Global Climate Models (GCMs) to be used for early-season yield forecasts, this paper uses a statistical downscaling technique, Extended Copula Post-Processing (ECPP) and the Schaake shuffle, to downscale four climate variables to generate weather-like daily data that are suitable for agricultural applications. Climate forecasts drive a process-based crop model APSIM (Agricultural Production Systems sIMulator) to simulate crop forecasts on 50 stations, well-distributed across the Australian grain zone. To focus on yield forecast skills attributable to SCF, we propose best practice management rules to predict water-limited winter wheat yield. Yield forecasts from ECPP have a significant improvement over quantile mapping downscaling and raw SCF from the Australian recent seasonal forecast model ACCESS-S1 in terms of bias, accuracy, reliability, and overall forecast skill. In addition, even at the beginning of a cropping season with a forecast lead time of four or more months, yield forecasts driven by ECPP illustrate higher skill than climatology, a benchmark for yield forecast. Early-season yield forecasts driven by SCFs provides a promising alternative to regression/machine-learning-based forecasts. Performance sensitivity and issues, and gaps on using skilful SCFs to help growers with their farming decision-making are discussed. [ABSTRACT FROM AUTHOR]

• Empirical density functions are linked to other datasets for crop-yield modeling. • Artificial covariates can represent complex phenomena in predictive models. • The proposed approach provides measures of uncertainty along with yield predictions. Machine learning methods are increasingly used in analyzing remotely sensed data and studying different aspects of agricultural production. In particular, several of these flexible models are widely adopted to predict regional crop yield during or after the growing season. However, most existing models cannot be applied when dealing with functional covariates. In this paper, an approach based on multidimensional scaling is proposed to generate a set of artificial covariates from empirical density functions of different phenomena captured within specific administrative boundaries through satellites. In contrast to traditional aggregation methods, this approach is designed to reduce the loss of information associated with the use of summary statistics as covariates. The proposed methodology is applied to NASA remote sensing data, combined with information from surveys and USDA's end-of-season county estimates, to study the prediction accuracy of different crop-yield models for three major crops in North Dakota. [ABSTRACT FROM AUTHOR]

Abstract: Distributed crop simulation models are typically confronted with considerable uncertainty in weather variables. In this paper the use of MeteoSat-derived meteorological products to replace weather variables interpolated from weather stations (temperature, reference evapotranspiration and radiation) is explored. Simulations for winter-wheat were carried for Spain, Poland and Belgium using both interpolated and MeteoSat-derived weather variables. The results were spatially aggregated to national and regional level and were evaluated by comparing the simulation results of both approaches and by assessing the relationships with crop yield statistics over the periods 1995–2003 from EUROSTAT. The results indicate that potential crop yield can be simulated well using MeteoSat-derived meteorological variables, but that water-stress hardly occurs in the water-limited simulations. This is caused by a difference in reference evapotranspiration which was 20–30% smaller in case of MeteoSat. As a result, the simulations using MeteoSat-derived meteorological variables performed considerably poorer in a regression analyses with crop yield statistics on national and regional level. Our results indicate that a recalibration of the model parameters is necessary before the MeteoSat-derived meteorological variables can be used operationally in the system. [Copyright &y& Elsevier]

Phenology is the study of recurring plant and animal life-cycle stages which can be observed across spatial and temporal scales that span orders of magnitude (e.g., organisms to landscapes). The variety of scales at which phenological processes operate is reflected in the range of methods for collecting phenologically relevant data, and the programs focused on these collections. Consideration of the scale at which phenological observations are made, and the platform used for observation, is critical for the interpretation of phenological data and the application of these data to both research questions and land management objectives. However, there is currently little capacity to facilitate access, integration and analysis of cross-scale, multi-platform phenological data. This paper reports on a new suite of software and analysis tools – the "Pheno-Synthesis Software Suite," or PS3 – to facilitate integration and analysis of phenological and ancillary data, enabling investigation and interpretation of phenological processes at scales ranging from organisms to landscapes and from days to decades. We use PS3 to investigate phenological processes in a semi-aride, mixed shrub-grass ecosystem, and find that the apparent importance of seasonal precipitation to vegetation activity (i.e., "greenness") is affected by the scale and platform of observation. We end by describing potential applications of PS3 to phenological modeling and forecasting, understanding patterns and drivers of phenological activity in real-world ecosystems, and supporting agricultural and natural resource management and decision-making. • Proper phenological analysis must consider the scale at which phenological observations are made. • However, there is currently little capacity to facilitate analysis of cross-scale, multi-platform phenological data. • Therefore, software suite was developed to enable phenological analysis at scales ranging from organisms to landscapes and from days to decades. [ABSTRACT FROM AUTHOR]

Advances in our ability to predict climate fluctuations months in advance suggest opportunity to improve management of climatic risk in agriculture, but only if particular conditions are in place. This paper outlines prerequisites to beneficial forecast use; highlights key issues, approaches and challenges related to each; and suggests an evolutionary strategy. The first prerequisite is that forecast information must address a need that is both real and perceived. Second, benefit arises only through viable decision options that are sensitive to forecast information. Third, benefit depends on prediction of the components of climate variability that are relevant to viable decisions. Fourth, appropriate forecast use requires effective communication of relevant information. Finally, sustained use requires institutional commitment and favorable policies. It is useful to consider three phases of effort: an exploratory phase to gain understanding and assess potential, a pilot phase characterized by co-learning between researchers and target decision makers, and an operational phase focusing on engaging and equipping relevant institutions. Although examples of use and potential use, and advances in institutional support, are cause for optimism, use of climate prediction by agriculture s still too new to support strong generalizations about its value. [Copyright &y& Elsevier]

Information and communication technology provides e-agriculture with data storage and data analysis capabilities, among which data collection is mainly achieved through a large number of sensors with limited computing and storage capabilities. Since sensors with low security protection are vulnerable to various attacks in agriculture environment, it is challenging to ensure the correctness of the collected data and secure data processing. Most of the existing works mainly consider how to perform data analysis and secure storage, while ignoring the security of the collection task. In this paper, we propose a green supply chain framework based on non-cooperative game, in which we designed a more scientific game theory model based on the Bayesian formula that combines past experience with the current situation. This model ensures the reliability of the data provided by various sensors. In order to manage the equipment in the system more intelligently and flexibly, we propose a green data supply chain and develop trust management to record the trust values of sensors in the blockchain, which significantly improved the transparency and trust of the system. Finally, we conduct simulations to evaluate the performances. [Display omitted] • Consider a blockchain-based green supply chain architecture. • Develop trust management technology to remove and isolate malicious sensors. • Propose a non-cooperative game model to filter malicious data from the collected data. • Conduct simulations to evaluate the agricultural device management scheme. [ABSTRACT FROM AUTHOR]

This paper presents a methodological framework to conduct a water stress assessment of water resources through water stress index. A modified version of the water stress index was developed and applied for cereal grown in the Kebir Rhumel Basin, Algeria. The water stress index defined as Crop Stress Index (CSI) and derived by accumulating differences in water supply and demand over time to assess the maximum cumulative deficit. The forecasting crop water stress index for cereals was developed here as a way of understanding and hedging risk. The CSI was forecasted using the appropriate climate predictors as predictors in a non-parametric k -nearest neighbors (k -NN) approach that issues probabilistic forecasts. The accuracy of the CSI forecasts in the years 2002–2013 was judged based directional concordance and contingency metrics such as hit , miss and false alarm rate. The results showed that our forecasts were correct 8 out of 12 times, with three misses and just one false alarm. These results confirmed that this is a useful approach in investigating irrigation needs. Finally, it was found that this water stress index is a useful tool to estimate water storage and irrigation requirements in the interest of agricultural water resources. • Water stress assessment of water resources through water stress index. • The useful tool of the Crop Stress Index (CSI) to estimate water storage and irrigation requirements in the interest of agricultural water resources. • The significant impact of the climate variability on the Crop Stress Index (CSI). • Forecasting the CSI using a non-parametric k-nearest neighbors (k-NN) approach. [ABSTRACT FROM AUTHOR]

Coffee is one of the trademarks of Colombia. Currently, up to a half million Colombian families depend directly on coffee production for their livelihoods. As such, there has been increasing concerns about how coffee productivity will react to changing climate conditions and how coffee growers could adapt their production practices. This paper is one of the first to estimate the production function of Colombian coffee at the municipal level and to make projections about its future productivity. Using a panel dataset measured across municipalities over 2007–2013, we find that productivity depends on altitude as well as on March temperature and precipitation. We estimate projections based on the 2.6, 4.5, and 6.0 Representative Concentration Pathways derived from Global Circulation Models. We find that productivity over 2041–2060 is expected to increase by 7.6% on average. However, we find that this forecast varies greatly according to altitude. Indeed, municipalities above median elevation will increase their productivity by 16%, while those below the median will experience a 8.1% decrease in productivity. Our result implies that place-tailored strategies for coffee production in Colombia are required to adapt to changing climate conditions in the future. [Display omitted] • Colombia is a major coffee producer and there is uncertainty about the effect of climate variability at the municipal level. • We estimate a dynamic panel model and use it to forecast productivity at the municipal level for 2041–2060. • Productivity will increase by 7.6% nation-wide and 16.7% in high-altitude municipalities. It will decrease by 8.16% in low-altitude municipalities. • Although productivity on average will rise, attention must be paid to areas that will be less productive. • The government and coffee institutions must facilitate crop shifting efforts in areas expected to decrease productivity. [ABSTRACT FROM AUTHOR]

• We model the water content prediction problem as a learning problem solved by supervised learning. • We predict the water potential of a potato crop for up to a week. • We use feature selection methods, in order to automatically choose the most relevant features in the model design. • We compare performance and robustness of different algorithms under different climatic conditions. Agricultural decision-making is crucial for future yields. In the context of smart farming, grower combine information from sensors located close to their crops with agronomic models to help them to better understand their crops. Irrigation management is therefore based on extrapolation of data and/or agronomic model responses. This problem can be seen as a learning task for which machine learning techniques have proven their relevance in many and diverse applications. In this paper we place ourselves in the context of potato farming, a crop for which irrigation plays a crucial role. We model the problem of soil water potential prediction as a learning problem solved by supervised leaning algorithms. The problem appears to be difficult since there are several potential inputs, and several outputs to predict. Experiments are conducted on several scenarios with data acquired during 3 years. We demonstrate the possibility of applying feature selection method to automatically design models with features relevant to the problem at hand while having good performances. We have also demonstrated the relevance of the machine learning for this kind of problem, since the methods are able to correctly predict the next water potential values. [ABSTRACT FROM AUTHOR]