Search

Yield prediction models can be divided between data-driven and process-based models (crop growth models). The first category contains many different types of models with parameters learned from the data themselves and where domain knowledge is only used to select the predictors and engineer features. In the second category, models are based upon biophysical principles, whose structure and parameters are derived primarily from domain knowledge. Here we investigate if the integration of the two approaches can be beneficial as it allows to overcome the limitations of the two approaches taken individually - lack of sufficiently large, reliable and orthogonal datasets for data-driven approaches and the need of many inputs for process-based models. The applications of the two categories of models have been reviewed, paying special attention to the cases where the two approaches have been mixed. By analysing the literature we identified three major cases of integration between the two approaches: (1) using crop growth models to engineer features and expand the predictors space, (2) use data-driven approaches to estimate missing inputs for process-based models (3) using data-driven approaches to produce meta-models to reduce computation burden. Finally we propose a methodology based on metamodels and transfer learning to integrate data-driven and process-based approaches.

Funding Information: This work was in part supported by the Collaborative Research Center 1253 CAMPOS (Project 7: Stochastic Modeling Framework), funded by the German Research Foundation ( DFG , Grant Agreement SFB 1253/1 2017 ), the Academy of Finland through projects AICropPro ( 316172 ) and DivCSA ( 316215 ) and Natural Resources Institute Finland (Luke) through a strategic project BoostIA, the BonaRes project ''Soil3'' ( BOMA 03037514 ) of the Federal Ministry of Education and Research ( BMBF ), Germany, the Deutsche Forschungsgemeinschaft ( DFG , German Research Foundation ) under Germany's Excellence Strategy - EXC 2070–390732324 EXC (PhenoRob), the project BiomassWeb of the GlobeE programme (Grant number: FKZ031A258B ) funded by the Federal Ministry of Education and Research ( BMBF , Germany), the INRA ACCAF meta-programme, the German Federal Ministry of Education and Research ( BMBF ) in the framework of the funding measure “Soil as a Sustainable Resource for the Bioeconomy – BonaRes”, project “BonaRes (Module B): BonaRes Centre for Soil Research , subproject B” (grant 031B0511B ), the National Key Research and Development Program of China ( 2017YFD0300205 ), the National Science Foundation for Distinguished Young Scholars ( 31725020 ), the Priority Academic Program Development of Jiangsu Higher Education Institutions ( PAPD ), the 111 Project ( B16026 ), and China Scholarship Council , the Agriculture and Agri-Food Canada's Project 1387 under he Canadian Agricultural Partnership, the DFG Research Unit FOR 1695 ‘Agricultural Landscapes under Global Climate Change – Processes and Feedbacks on a Regional Scale, the U.S. Department of Agriculture National Institute of Food and Agriculture (award no. 2015-68007-23133 ) and USDA / NIFA HATCH grant N. MCL02368 , the National Key Research and Development Program of China ( 2016YFD0300105 ), The Broadacre Agriculture Initiative, a research partnership between University of Southern Queensland and the Queensland Department of Agriculture and Fisheries , the Academy of Finland through project AI-CropPro ( 315896 ), the JPI FACCE MACSUR2 project, funded by the Italian Ministry for Agricultural, Food, and Forestry Policies ( D.M. 24064/7303/15 of 6/Nov/2015), the Ministry of Education, Youth and Sports of Czech Republic through SustES - Adaption strategies for sustainable ecosystem services and food security under adverse environmental conditions (project no. CZ.02.1.01/0.0/0.0/16_019/000797 ). The order in which the donors are listed is arbitrary. Funding Information: This work was in part supported by the Collaborative Research Center 1253 CAMPOS (Project 7: Stochastic Modeling Framework), funded by the German Research Foundation (DFG, Grant Agreement SFB 1253/1 2017), the Academy of Finland through projects AICropPro (316172) and DivCSA (316215) and Natural Resources Institute Finland (Luke) through a strategic project BoostIA, the BonaRes project ''Soil3'' (BOMA 03037514) of the Federal Ministry of Education and Research (BMBF), Germany, the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany's Excellence Strategy - EXC 2070?390732324 EXC (PhenoRob), the project BiomassWeb of the GlobeE programme (Grant number: FKZ031A258B) funded by the Federal Ministry of Education and Research (BMBF, Germany), the INRA ACCAF meta-programme, the German Federal Ministry of Education and Research (BMBF) in the framework of the funding measure ?Soil as a Sustainable Resource for the Bioeconomy ? BonaRes?, project ?BonaRes (Module B): BonaRes Centre for Soil Research, subproject B? (grant 031B0511B), the National Key Research and Development Program of China (2017YFD0300205), the National Science Foundation for Distinguished Young Scholars (31725020), the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD), the 111 Project (B16026), and China Scholarship Council, the Agriculture and Agri-Food Canada's Project 1387 under he Canadian Agricultural Partnership, the DFG Research Unit FOR 1695 ?Agricultural Landscapes under Global Climate Change ? Processes and Feedbacks on a Regional Scale, the U.S. Department of Agriculture National Institute of Food and Agriculture (award no. 2015-68007-23133) and USDA/NIFA HATCH grant N. MCL02368, the National Key Research and Development Program of China (2016YFD0300105), The Broadacre Agriculture Initiative, a research partnership between University of Southern Queensland and the Queensland Department of Agriculture and Fisheries, the Academy of Finland through project AI-CropPro (315896), the JPI FACCE MACSUR2 project, funded by the Italian Ministry for Agricultural, Food, and Forestry Policies (D.M. 24064/7303/15 of 6/Nov/2015), the Ministry of Education, Youth and Sports of Czech Republic through SustES - Adaption strategies for sustainable ecosystem services and food security under adverse environmental conditions (project no. CZ.02.1.01/0.0/0.0/16_019/000797). The order in which the donors are listed is arbitrary. Publisher Copyright: © 2021 Elsevier Ltd Calibration, the estimation of model parameters based on fitting the model to experimental data, is among the first steps in many applications of process-based models and has an important impact on simulated values. We propose a novel method of developing guidelines for calibration of process-based models, based on development of recommendations for calibration of the phenology component of crop models. The approach was based on a multi-model study, where all teams were provided with the same data and asked to return simulations for the same conditions. All teams were asked to document in detail their calibration approach, including choices with respect to criteria for best parameters, choice of parameters to estimate and software. Based on an analysis of the advantages and disadvantages of the various choices, we propose calibration recommendations that cover a comprehensive list of decisions and that are based on actual practices.

Plant phenology, which describes the timing of plant development, is a major aspect of plant response to environment and for crops, a major determinant of yield. Since climate change is projected to alter crop phenology worldwide, there is a large effort to predict phenology as a function of environment. Many studies have focused on comparing model equations for describing how phenology responds to weather but the effect of crop model calibration, also expected to be important, has received much less attention. The objective here was to obtain a rigorous evaluation of prediction capability of wheat crop phenology models, and to analyze the role of calibration. The 27 participants in this multi-model study were provided experimental data for calibration and asked to submit predictions for sites and years not represented in those data. Participants were instructed to use and document their 99usual99 calibration approach. Overall, the models provided quite good predictions of phenology (median of mean absolute error of 6.1 days) and did much better than simply using the average of observed values as predictor. Calibration was found to compensate to some extent for differences between models, specifically for differences in simulated time to emergence and differences in the choice of input variables. Conversely, different calibration approaches led to major differences in prediction error between models with the same structure. Given the large diversity of calibration approaches and the importance of calibration, there is a clear need for guidelines and tools to aid with calibration. Arguably the most important and difficult choice for calibration is the choice of parameters to estimate. Several recommendations for calibration practices are proposed. Model applications, including model studies of climate change impact, should focus more on the data used for calibration and on the calibration methods employed.

Understanding subfield crop yields and temporal stability is critical to better manage crops. Several algorithms have proposed to study within-field temporal variability but they were mostly limited to few fields. In this study, a large dataset composed of 5520 yield maps from 768 fields provided by farmers was used to investigate the influence of subfield yield distribution skewness on temporal variability. The data are used to test two intuitive algorithms for mapping stability: one based on standard deviation and the second based on pixel ranking and percentiles. The analysis of yield monitor data indicates that yield distribution is asymmetric, and it tends to be negatively skewed (p −1 for maize, 0.39 Mg ha−1 for cotton, 0.34 Mg ha−1 for soybean, and 0.59 Mg ha−1 for wheat. The yield of the unstable zones was similar to the pixels classified as low-and-stable by the standard deviation algorithm, whereas the two-way outlier algorithm did not exhibit this bias. Furthermore, the increase in the number years of yield maps available induced a modest but significant increase in the certainty of stability classifications, and the proportion of unstable pixels increased with the precipitation heterogeneity between the years comprising the yield maps.

Predicting wheat phenology is important for cultivar selection, for effective crop management and provides a baseline for evaluating the effects of global change. Evaluating how well crop phenology can be predicted is therefore of major interest. Twenty-eight wheat modeling groups participated in this evaluation. Model predictions depend not only on model structure but also on the parameter values. This study is thus an evaluation of modeling groups, which choose the structure and fix or estimate the parameters, rather than an evaluation just of model structures. Our target population was wheat fields in the major wheat growing regions of Australia under current climatic conditions and with current local management practices. The environments used for calibration and for evaluation were both sampled from this same target population. The calibration and evaluation environments had neither sites nor years in common, so this is a rigorous evaluation of the ability of modeling groups to predict phenology for new sites and weather conditions. Mean absolute error (MAE) for the evaluation environments, averaged over predictions of three phenological stages and over modeling groups, was 9 days, with a range from 6 to 20 days. Predictions using the multi-modeling group mean and median had prediction errors nearly as small as the best modeling group. For a given modeling group, MAE for the evaluation environments was significantly correlated with MAE for the calibration environments, which suggests that it would be of interest to test ensemble predictors that weight individual modeling groups based on performance for the calibration data. About two thirds of the modeling groups performed better than a simple but relevant benchmark, which predicts phenology by assuming a constant temperature sum for each development stage. The added complexity of crop models beyond just the effect of temperature was thus justified in most cases. Finally, there was substantial variability between modeling groups using the same model structure, which implies that model improvement could be achieved not only by improving model structure, but also by improving parameter values, and in particular by improving calibration techniques.

International audience; Smallholder farmers in sub-Saharan Africa (SSA) currently grow rainfed maize with limited inputs including fertilizer. Climate change may exacerbate current production constraints. Crop models can help quantify the potential impact of climate change on maize yields, but a comprehensive multimodel assessment of simulation accuracy and uncertainty in these low-input systems is currently lacking. We evaluated the impact of varying [CO2], temperature and rainfall conditions on maize yield, for different nitrogen (N) inputs (0, 80, 160 kg N/ha) for five environments in SSA, including cool subhumid Ethiopia, cool semi-arid Rwanda, hot subhumid Ghana and hot semi-arid Mali and Benin using an ensemble of 25 maize models. Models were calibrated with measured grain yield, plant biomass, plant N, leaf area index, harvest index and in-season soil water content from 2-year experiments in each country to assess their ability to simulate observed yield. Simulated responses to climate change factors were explored and compared between models. Calibrated models reproduced measured grain yield variations well with average relative root mean square error of 26%, although uncertainty in model prediction was substantial (CV = 28%). Model ensembles gave greater accuracy than any model taken at random. Nitrogen fertilization controlled the response to variations in [CO2], temperature and rainfall. Without N fertilizer input, maize (a) benefited less from an increase in atmospheric [CO2]; (b) was less affected by higher temperature or decreasing rainfall; and (c) was more affected by increased rainfall because N leaching was more critical. The model intercomparison revealed that simulation of daily soil N supply and N leaching plays a crucial role in simulating climate change impacts for low-input systems. Climate change and N input interactions have strong implications for the design of robust adaptation approaches across SSA, because the impact of climate change in low input systems will be modified if farmers intensify maize production with balanced nutrient management.

Not all areas of a farmer’s field are equal; some always produce more relative to the rest of the field, others always less, while still other areas fluctuate in their production capacity from one year to the next, depending on the interaction between climate, soil, topography and management. Understanding why the yield in certain portions of a field has a high variability over time—we call these areas unstable—is of paramount importance both from an economic and an environmental point of view, as it is through the better management of these areas that we can improve yields or reduce input costs and environmental impact. In this research, we analyzed data from 338 fields cultivated with maize, soybean, wheat and cotton in the US Midwest to understand how topographic attributes and rain affect yield stability over time. In addition to this high resolution yield monitor dataset, we used publicly available data on topography, rain and soil information to test the hypothesis that within-field areas characterized by a low topographic wetness index (proxy for areas with probability of lower water content) always perform poorly (low and stable yield) compared to the rest of the field because they are drier, and that areas of a field characterized by a mid-high wetness index (high and stable yield) always perform well relative to rest of the field because they have greater water availability to plants. The relative performance of areas of a field with a very high wetness index (e.g. depressions) strongly depends on rain patterns because they may be waterlogged in wet years, yielding less than the rest of the field, or wetter during dry years, yielding more than the rest of the field. We present three different observations from this dataset to support our hypothesis. First, we show that the average topographic wetness index in the different stability zones is lower in low and stable yield areas, high in high and stable yield areas and even higher in unstable yield areas (p

Over the last two decades, there has been significant advancements in the application of geospatial technologies in agriculture. Improved resolutions (spectral, spatial and temporal) of remotely sensed images, coupled with more precise on-the-ground systems (yield monitors, geophysical sensors) have allowed farmers to become more sensitive about the spatial and temporal variations of crop yields occurring in their fields. Previous research has extensively looked at spatial variability of crop yields at field scale, but studies designed to predict within-field spatial patterns of yield over a large number of fields as yet been reported. In this paper, we analyzed 571 fields with multiple years of yield maps at high spatial resolution to understand and predict within-field spatial patterns across eight states in the Midwest US and over corn, soybean, wheat and cotton fields. We examined the correlation between yield and 4 covariates, three derived from remote sensing imagery (red band spectral reflectance, NDVI and plant surface temperature) and the fourth from yield maps from previous years. The results showed that for spatial patterns that are stable over time the best predictor of the spatial variability is the historical yield map (previous years’ yield maps), while for zones within the field that are more sensitive to weather and thus fluctuate from one year to the next the best predictor of the spatial patterns are the within-season images. The results of this research help quantify the role of historical yield maps and within-season remote sensing images to predict spatial patterns. The knowledge of spatial patterns within a field is critical not only to farmers for potential variable rate applications, but also to select homogenous zones within the field to run crop models with site-specific input to better understand and predict the impact of weather, soil and landscape characteristics on spatial and temporal patterns of crop yields to enhance resource use efficiency at field level.

Pyrogenic carbon (PyC) is hypothesized to play an important role in the carbon (C) cycle due to its resistance to decomposition; however, much uncertainty still exists regarding the stocks of PyC that persist on-site after the initial erosion in post-fire forests. Therefore, understanding how fire characteristics influence PyC stocks is vital, particularly in the context of California forests for which an increase of high-severity fires is predicted over the next decades. We measured forest C and persistent PyC stocks in areas burned by low-to-moderate and high-severity fire, as well as in adjacent unburned areas in a California mixed-conifer forest, two to three years after wildfire. We measured C and PyC stocks in the following compartments: standing trees, downed wood, forest floor and mineral soil (0–5 cm), and we identified PyC using the weak nitric acid digestion method. We found that the total stock of PyC did not differ among fire severity classes (overall mean 248 ± 30 g C m−2); however, fire severity influenced the distribution of PyC in the individual compartments. Areas burned by high-severity fire had 2.5 times more PyC stocked in the coarse woody debris (p < 0.05), 3.3 times more PyC stocked in standing trees (p < 0.05) and a lower PyC stock in the forest floor (−22%, p < 0.05) compared to low-to-moderate fire severity areas. These results have important implications for the permanence time of PyC, which is putatively higher in standing trees and coarse woody debris compared to the forest floor, where it is susceptible to rapid losses through erosion.

The weak nitric acid digestion method (also known as KMD) is a method to quantify black carbon (BC) in mineral soil. Here we extended the use of this methodology from mineral samples to organic matrices. We tested this methodology on known mixtures of pyrolyzed pine wood and needles (at 300 °C and 550 °C) and non-pyrolyzed pine needles and measured: (1) the fraction of pyrolyzed material identified as BC by the methodology, (2) the fraction of non-pyrolyzed pine needles identified as BC, and (3) the precision of the method. We found that the fraction of pyrolyzed-C identified as BC ranges from 10% (needles at 300 °C) to 90% (wood at 550 °C), indicating that the weak nitric acid digestion method can be extended to measure BC in organic matrices and that the methodology can be used in studies aimed to quantify BC stocks in post-fire environments.

Soil organic matter (SOM) represents the biggest pool of carbon within the biosphere and influences the flux of greenhouse gases between land surface and atmosphere. In this regard, conservation tillage systems have been adopted to reduce negative impacts of conventional tillage practices on greenhouse gases (GHG) emissions. However, the role of these techniques to increase carbon sequestration also depends upon soil features and climatic conditions, which is studied and managed by precision agriculture (PA) principles and technologies. Simulation models have shown to be useful tools to understand the interaction between soil, climate, genotypes and management practices to simulate the long-term effects of management approaches of different soils on crop yield, soil organic carbon (SOC) storage, and GHG emissions. The research goals of this study are (1) to examine the mid-term (15 years) trajectory of SOC in the upper 0.4 m of the soil profile under different tillage systems using the SALUS model; (2) determine the impact of PA on the inputs to the crop and CO2 emissions; (3) identify the strategies, derived from the synergy between conservation agriculture and PA, so as to decrease the CO2 emissions of agricultural systems. The validated SALUS simulation showed a significant reduction in SOC losses in minimum tillage (MT) and no-tillage (NT), 17% and 63% respectively, compared to conventional tillage (CT). Furthermore, the adoption of conservation tillage techniques decreased carbon emissions related to farming operations, while PA technologies led to an optimization of the exhaustible sources such as fossil fuels and fertilizers. Finally, the synergy between conservation tillage systems, especially NT, and PA strategies represents a useful tool in terms of carbon emissions mitigation with a reduction of 56% of total CO2 as compared to CT.

Pyrogenic organic matter (PyOM) plays an important role as a stable carbon (C) sink in the soils of terrestrial ecosystems. However, uncertainties remain about in situ turnover rates of fire-derived PyOM in soil, the main processes leading to PyOM-C and nitrogen (N) losses from the soil, and the role of N availability on PyOM cycling in soils. We measured PyOM and native soil organic carbon losses from the soil as carbon dioxide and dissolved organic carbon (DOC) using additions of highly 13C-labelled PyOM (2.03 atom %) and its precursor pinewood during 1 year in a temperate forest soil. The field experiment was carried out under ambient and increased mineral N deposition (+60 kg N-NH4NO3 ha−1 year−1). The results showed that after 1 year: (1) 0.5% of PyOM-C and 22% of wood-C were mineralized as CO2, leading to an estimated turnover time of 191 and 4 years, respectively; (2) the quantity of PyOM and wood lost as dissolved organic carbon was negligible (0.0004 ± 0.0003% and 0.022 ± 0.007% of applied-C, respectively); and (3) N additions decreased cumulative PyOM mineralization by 43%, but did not affect cumulative wood mineralization and did not affect the loss of DOC from PyOM or wood. We conclude that mineralization to CO2 was the main process leading to PyOM losses during the first year of mineralization in a forest soil, and that N addition can decrease PyOM-C cycling, while added N showed no effect on wood C cycling.

Pyrogenic organic matter (PyOM) is considered an important soil carbon (C) sink. However, there are evidences that its addition to soil may induce a priming effect (PE) thus influencing its C abatement potential. The direction, the size and the mechanisms responsible for PyOM induced PE are far from being understood. We collected approximately 650 data points from 18 studies to analyse the characteristics of the PE induced by PyOM. The database was divided between the PE induced on the native soil organic matter and on fresh organic matter. Most of the studies were short-term incubation therefore the projections of findings on the long term may be critical. Our findings indicate that over 1 year PyOM induces an average positive PE of 0.3 mg C g−1 soil on native soil organic matter and a PE of approximately the same size but opposite direction on fresh organic matter. We studied the correlation of PE with several properties of soil, of the added PyOM, and time after PyOM addition. We found that PyOM primes positively the native soil organic matter in the first 20 days while negative PE appears in a later stage. Negative PE was correlated with the soil C content. PyOM characterized by a low C content induced a higher positive PE on native soil organic carbon. No correlation was found between the factors record in our database and the PE induced on the fresh organic matter. We reviewed the mechanisms proposed in literature to explain PE and discussed them based on findings from our meta-analysis. We believe that the presence of a labile fraction in PyOM may trigger the activity of soil microorganisms on the short term and therefore induce a positive PE, while on the long term PyOM may induce a negative PE by promoting physical protection mechanisms.

Pyrogenic organic matter (PyOM) decomposes on centennial timescale in soils, but the processes regulating its decay are poorly understood. We conducted one of the first studies of PyOM and wood decomposition in a temperate forest using isotopically labeled organic substrate, and quantified microbial incorporation and physico-chemical transformations of PyOM in situ. Stable-isotope (13C and 15N) enriched PyOM and its precursor wood were added to the soil at 2 cm depth at ambient (N0) and increased (N+) levels of nitrogen fertilization. The carbon (C) and nitrogen (N) of added PyOM or wood were tracked through soil to 15 cm depth, in physically separated soil density fractions and in benzene polycarboxylic acids (BPCA) molecular markers. After 10 months in situ, more PyOM-derived C (>99% of initial 13C-PyOM) and N (90% of initial 15N-PyOM) was recovered than wood derived C (48% of 13C-wood) and N (89% under N0 and 48% under N+). PyOM-C and wood-C migrated at the rate of 126 mm yr-1 with 3-4% of PyOM-C and 4-8% of wood-C recovered below the application depth. Most PyOM C was recovered in the free light fraction (fLF) (74%), with 20% in aggregate-occluded and 6% in mineral associated fractions —fractions that typically have much slower turnover times. In contrast, wood C was recovered mainly in occluded (33%) or dense fraction (27%). PyOM addition induced loss of native C from soil (priming effect), particularly in fLF (13%). The total BPCA-C content did not change but after ten months the degree of aromatic condensation of PyOM decreased, as determined by relative contribution of benzene hexa-carboxylic acid (B6CA) to the total BPCA C. Soil microbial biomass (SMB) assimilated 6-10% of C from the wood, while PyOM contributions was negligible (0.14–0.18%). The addition of N had no effect on the dynamics of PyOM while limited effect on wood.