Skip to main content

Sensores remotos para la obtención de información predictiva de la producción de cultivos cereales (PREDIC-PRO)

CPP2021-008733

•
Funding agency name Agencia Estatal de Investigación
Funding agency acronym AEI
Program Programa Estatal para Impulsar la Investigación Científico-Técnica y su Transferencia
Subprogram Subprograma Estatal de Transferencia de Conocimiento
Call Proyectos en Colaboración Público-Privada
Call year 2021
State Plan 2021-2023 Plan Estatal de Investigación Científica y Técnica y de Innovación
Beneficiary centre FERTINAGRO BIOTECH, SL

Resultados relacionados

Resultados totales (Incluyendo duplicados): 9
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

A Standardized Framework for Cleaning Non-Normal Yield Data from Wheat and Barley Crops, and Validation Using Machine Learning Models for Satellite Imagery

  • Arizo-García, Patricia|||0000-0002-6725-9098
  • Castiñeira-Ibáñez, Sergio|||0000-0001-6414-3906
  • Cruzado-Campos, Enric
  • Ricarte Benedito, Beatriz|||0000-0001-8094-1908
  • Rubio Michavila, Constanza|||0000-0002-4395-7473
  • San Bautista Primo, Alberto|||0000-0003-4846-6611
[EN] Modern combine harvesters can collect real-time geolocated yield data, but it is subject to errors. Various protocols have been proposed to clean this data, each with varying levels of complexity. This data is valuable for precision agriculture to implement site-specific management and to train models to predict yield using remote sensing data. Machine learning and deep learning techniques have shown their potential for precision agriculture, and their performance shows no significant differences between models trained with data cleaned using a computationally demanding protocol or a simpler one, such as parametric filtering. However, parametric filtering approaches primarily rely on statistics that are highly sensitive to data distribution and do not effectively filter inliers. The objective of this study is to develop a data-cleansing method that leverages robust statistical measures, specifically the median and interquartile range, to effectively identify and filter outliers and inliers while retaining valid observations in datasets collected from combine harvesters, thereby minimizing the influence of non-normal data distributions. Different levels of data cleaning were applied to a total of 7399 ha of wheat and barley crops, and the quality of each cleaning level was compared. The selected protocol improved the spatial structure of the data, deleting up to 42% and 33% of the data at the polygon level, for wheat and barley, respectively. It increased the mean and median, and decreased the standard deviation and coefficient of variation of the data. Between 78.7% and 82.9% of the fields showed a normal distribution after applying the selected method, and machine learning performance improved compared with the raw data. Compared with previous data cleaning studies, the present work proposes an automatic, low-computational, parametric filtering method that uses robust statistics for non-normal distributions. In addition, its scalability has been demonstrated by applying the method to a large dataset, improving data quality and the performance of yield-prediction ML models in all cases., This research was funded by the PREDIC-PRO, project SCPP2100C008733XV0, of the State Research Agency of the Ministry of Science, Innovation and Universities, and the ACIF Generalitat Valenciana, European Union (European Social Fund: Investing in Your Future), grant number CIACIF/2022/255.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Improving harvester yield maps postprocessing leveraging remote sensing data in rice crop

  • Fita-Silvestre, David
  • Franch-Gras, Belen
  • Rubio Michavila, Constanza|||0000-0002-4395-7473
  • Castiñeira-Ibáñez, Sergio|||0000-0001-6414-3906
  • Tarrazó-Serrano, Daniel|||0000-0001-9986-0884
  • San Bautista Primo, Alberto|||0000-0003-4846-6611
[EN] Precision Agriculture relies significantly on yield data obtained from combine harvesters, which constitutes a pivotal tool for optimizing crop productivity. Despite its potential, challenges in data accuracy persist, necessitating the development of novel automated postprocessing protocols for yield data refinement. In this paper, different automatic postprocessing protocols were evaluated using remote sensing data, specifically Sentinel-2 satellite imagery. Various automatic postprocessing protocols were applied to a dataset spanning 946 hectares over a four-year period. Commercial sensors on combine harvesters acquired the yield data. The analysis included global (field-level) adjustments and local adjustments at a finer scale (40 × 40 m² level), employing interval mean ± n·(standard deviation) calculations. Three n values (1, 1.5, and 2.5) were tested, resulting in thirteen distinct postprocessing variations. Finally, a mean filter was also applied. The results demonstrated that the yield correlation with satellite data increased with the reduction of yield variability at the pixel level (10 m). The best results were obtained using n = 1 with a 3 × 3 mean filter, where Sentinel-2 pixels remained unaffected, and the average Root Mean Square Error (RMSE) during validation was 0.572 t·ha¿¹. In addition, the geostatistical parameters (coefficient of variation, semivariance, and range within a 10 m pixel) reached optimal values. Finally, the postprocessing uncertainty was determined to be 0.200 t·ha¿1. These results validate the efficacy of a novel postprocessing protocol for refining yield data in rice crops. The integration of pixel-level data from combine harvesters with Sentinel-2 imagery emerges as a promising approach for optimizing crop management, offering valuable insights for the advancement of Precision Agriculture., This research was funded by PREDIC-PRO Project SCPP2100C008733XVD funded by the State Research Agency of theMinistry of Science, Innovation and Universities, and Conselleria de Cultura, Educación y Ciencia, Generalitat Valenciana (CIACIF/2021/143), European Union European Social Fund Investing in Your Future.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Remote Sensor Images and Vegetation Indices to Optimize Rice Yield Analysis for Specific Growth Stages Within Extensive Data

  • Fita-Silvestre, David
  • Franch-Gras, Belen
  • Rubio Michavila, Constanza|||0000-0002-4395-7473
  • Uris Martínez, Antonio|||0000-0002-2005-2305
  • Castiñeira-Ibáñez, Sergio|||0000-0001-6414-3906
  • Tarrazó-Serrano, Daniel|||0000-0001-9986-0884
  • San Bautista Primo, Alberto|||0000-0003-4846-6611
[EN] The crop yield in commercial fields is a very important parameter for farmers. The use of Precision Agriculture tools has been shown to improve rice crop yields. One of these tools is remote sensing on satellite platforms. Sentinel-2 provides free data on reflectance at different wavelengths. Focusing on commercial farms, correlations between the yield and satellite reflectance were studied over several years and locations for `JSendra¿ rice crops. Four years of yield maps for 706 ha composed the database. Mid tillering-MT, panicle initiation-PI and grain filling-GF reflectance values and Vegetation Indices (VIs) were used. At MT, correlations with the yield were variable (0.23¿0.70). At PI, correlations with the yield increased in NIR (0.39¿0.85), but the other regions and VIs experienced a decrease. Visible bands and B05 Red Edge were significantly correlated with each other; similarly, B08 NIR was highly correlated with B06, B07, and B8A; SWIR bands were correlated with each other but not with the yield. At GF, the previous pattern was similar. Substantial limitations in estimating yield variability directly from reflectance or VIs were discussed. Two periods were established. The first is designing strategies to increase NIR and decrease red reflectance from MT to PI. The second is avoiding the relationship between crop greenness and NIR from PI to harvest. NIR was a better variable than VIs, but the single use of this band is challenging. Future recommendations focus on the visible¿NIR collinearities to interpret differences between years or locations., This research has been funded by the PREDIC-PRO project SCPP2100C008733XV0 (CPP2021-008733) of the State Research Agency of the Ministry of Science, Innovation and Universities, and ACIF Generalitat Valenciana, European Union (European Social Fund. Investing in Your Future) (CIACIF/2021/143).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Spectral Phenological Typologies for Improving Cross-Dataset in Mediterranean Winter Cereals

  • Arizo-García, Patricia|||0000-0002-6725-9098
  • Castiñeira-Ibáñez, Sergio|||0000-0001-6414-3906
  • Ricarte Benedito, Beatriz|||0000-0001-8094-1908
  • San Bautista Primo, Alberto|||0000-0003-4846-6611
  • Rubio Michavila, Constanza|||0000-0002-4395-7473
[EN] Accurate monitoring of crop phenology is essential for precision agriculture and yield forecasting. However, satellite-derived time series often suffer from inherent noise, such as residual atmospheric effects and mixed pixels, as well as a frequent lack of ground-truth data in agriculture. In response, this study proposes an algorithm to define the type of spectral signatures for the principal phenological stages of crops, using them as the foundation for training supervised machine learning classification models. The algorithm was developed using Fuzzy C-Means (FCM) clustering to identify the spectral signature reference groups in winter wheat across the Burgos region (Spain) during the 2020 and 2021 growing seasons. To enhance cluster independence and biological coherence, a multi-step filtering process was implemented, including spectral purity (membership degree, SAM, and SAMder) and temporal coherence filters. The filtered and labeled dataset (80% original Burgos dataset) was used to train supervised classification models (KNN and XGBoost). The models¿ reliability was verified through three wheat tests (remaining 20%), labeled using other clustering techniques, and an independent barley dataset from diverse geographic locations (Valladolid and Soria). The filtering process significantly improved cluster stability by removing outliers and transition spectral signatures. The supervised models demonstrated exceptional performance; the KNN model slightly outperformed XGB, achieving a mean Accuracy of 0.977, a Kappa of 0.967, and an F1-score of 0.977 in the wheat external test. Furthermore, the model showed, when applied to barley, that its phenological spectral signatures are equivalent in shape to those of wheat, with an Accuracy of 0.965 and an F1-score of 0.974. In addition, it was verified that the type spectral signatures remain the same regardless of the location. This study presents a robust classification tool capable of labeling four key phenological stages (tillering, stem elongation, ripening, and senescence) without ground truth. By effectively removing inherent satellite noise, the proposed methodology produces organized, cleaned datasets. This structured foundation is critical for future research integrating spectral signatures with harvester data to develop high-precision yield prediction models., This research was funded by the PREDIC-PRO project SCPP2100C008733XV0 of the State Research Agency of the Ministry of Science, Innovation and Universities, and the ACIF Generalitat Valenciana, European Union (European Social Fund. Investing in Your Future) grant number CIACIF/2022/255.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Use of Sentinel-2 Images to Elaborate a VRT Sensor-Based and Map-Based Nitrogen Fertilization in Wheat and Barley Crops

  • Arizo-García, Patricia|||0000-0002-6725-9098
  • Castiñeira-Ibáñez, Sergio|||0000-0001-6414-3906
  • Tarrazó-Serrano, Daniel|||0000-0001-9986-0884
  • Rubio Michavila, Constanza|||0000-0002-4395-7473
  • San Bautista Primo, Alberto|||0000-0003-4846-6611
  • Franch-Gras, Belen
[EN] Precision agriculture can determine the amount of nitrogen (N) required in each area to optimize yield and nitrogen use efficiency (NUE). The use of variable rate technology (VRT) for planning N fertilization has often relied on techniques that are unfeasible for farmers with limited resources. This study aims to present a variable fertilization plan for wheat and barley, along with a protocol to determine the optimal timing for the second nitrogen (N) application, thereby minimizing the need for in situ crop monitoring. Two approaches are studied: a more straightforward sensor-based method and a map-based method. The sensor-based approach involved modeling the maximum NDVI based on the observed value at the time of application and the required N level, achieving an R2 of 0.55 ± 0.06 and 0.72 ± 0.04, an MAE of 0.025 ± 0.002 and 0.039 ± 0.002, and an RMSE of 0.049 ± 0.007 and 0.055 ± 0.004 for wheat and barley, respectively. The map-based approach relied on training models to estimate the nitrogen dose to be applied based on the target yield and reflectance data from Sentinel-2 at the time of application. Using random forest algorithms, an R2 of 0.97 ± 0.01 and 0.96 ± 0.02, an MAE of 3.33 ± 0.20 kg N ha¿1 and 2.01 ± 0.13 kg N ha¿1, and an RMSE of 4.79 ± 0.31 kg N ha¿1 and 3.27 ± 0.58 kg N ha¿1 for wheat and barley, respectively., This research was funded by the PREDIC-PRO project SCPP2100C008733XV0 of the State Research Agency of the Ministry of Science, Innovation and Universities, and the ACIF Generalitat Valenciana, European Union (European Social Fund. Investing in Your Future) grant number CIACIF/2022/255.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

High-Resolution Wheat and Barley Yield Forecasting Using Multi-Temporal Satellite Time Series and Machine Learning

  • Arizo-García, Patricia|||0000-0002-6725-9098
  • Castiñeira-Ibáñez, Sergio|||0000-0001-6414-3906
  • Cruzado-Campos, Enric
  • San Bautista Primo, Alberto|||0000-0003-4846-6611
  • Rubio Michavila, Constanza|||0000-0002-4395-7473
[EN] High-resolution yield forecasting is essential for advancing precision agriculture and improving the sustainability of wheat and barley production. While most previous studies focus on field-scale predictions, pixel-level approaches are needed to capture intra-field variability and support site-specific management. This paper evaluates the performance of machine learning models for 10 m resolution yield prediction using multi-temporal Sentinel-2 surface reflectance data across seven major cereal-producing regions in Spain. Yield monitor data from winter wheat and barley fields collected over five growing seasons (2020¿2024) were combined with spectral bands and vegetation indices. Random Forest (RF) and XGBoost (XGB) models were trained at five phenological stages expressed as days before harvest (DBH) and validated using both internal (2020¿2023) and independent external (2024) datasets. Model accuracy increased as harvest approached. In external validation, RF achieved the best performance for wheat (R2 = 0.77; RMSE ¿ 697 kg · ha¿1), while XGB performed best for barley (R2 = 0.86; RMSE ¿ 744 kg · ha¿1). Visible, red-edge, and SWIR bands were the most informative predictors, especially during grain filling and senescence. Results demonstrate the potential of multi-temporal Sentinel-2 data and machine learning for accurate, transferable, pixel-level yield forecasting in Mediterranean cereal systems., This research was funded by the PREDIC-PRO project SCPP2100C008733XV0 of the State Research Agency of the Ministry of Science, Innovation and Universities, and the ACIF Generalitat Valenciana, European Union (European Social Fund. Investing in Your Future) grant number CIACIF/2022/255.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Assessment of Water Depth Variability and Rice Farming Using Remote Sensing

  • Simeon-Brocal, Ruben
  • Rubio Michavila, Constanza|||0000-0002-4395-7473
  • Uris Martínez, Antonio|||0000-0002-2005-2305
  • Coronado-Campos, Javier
  • Agenjos-Moreno, Alba
  • San Bautista Primo, Alberto|||0000-0003-4846-6611
[EN] Remote sensing is a widely used tool for crop monitoring to improve water management. Rice, a crop traditionally grown under flooded conditions, requires farmers to understand the relationship between crop reflectance, water depth and final yield. This study focused on seven commercial rice fields in 2022 and six in 2023, analyzing the correlations between water depth and Sentinel-2 reflectance over two growing seasons in Valencia, Spain. During the tillering stage across both seasons, water depth showed positive correlations with visible bands and negative correlations with NIR and SWIR bands. There were no correlations with the indices NDVI, GNDVI, NDRE and NDWI. The NIR band showed significant correlations across both seasons, with R2 values of 0.69 and 0.71, respectively. In addition, the calculation of NIR anomalies for each field proved to be a good indicator of final yield anomalies. In 2022, anomalies above 10% corresponded to yield deviations above 500 kg·ha−1, while in 2023, anomalies above 15% were associated with yield deviations above 1000 kg·ha−1. The response of final yield to water level was positive up to average values of 9 cm. The use of the NIR band during the rice crop tillering stage can support farmers in improving irrigation management., This research has been funded by the PREDIC-PRO project SCPP2100C008733XV0 (CPP2021-008733) of the State Research Agency of the Ministry of Science, Innovation and Universities.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Remote Sensing Dynamics for Analyzing Nitrogen Impact on Rice Yield in Limited Environments

  • Fita-Silvestre, David
  • Franch, Belén
  • Domingo Carrasco, Concha
  • San Bautista Primo, Alberto|||0000-0003-4846-6611
  • Castiñeira-Ibáñez, Sergio|||0000-0001-6414-3906
  • Rubio Michavila, Constanza|||0000-0002-4395-7473
[EN] Rice production remains highly dependent on nitrogen (N). There is no positive linear correlation between N concentration and yield in rice cultivation because an excess of N can unbalance the distribution of photo-assimilates in the plant and consequently produce a lower yield. We intended to study these imbalances. Remote sensing is a useful tool for monitoring rice crops. The purpose of this study was to evaluate the effectiveness of using remote sensing to assess the impact of N applications on rice crop behavior. An experiment with three different doses (120, 170 and 220 kg N·ha¿1) was carried out over two years (2021 and 2022) in Valencia, Spain. Biomass, Leaf Area Index (LAI), plants per m2, yield, N concentration and N uptake were determined. Moreover, reflectance values in the green, red, and NIR bands of the Sentinel-2 satellite were acquired. The two data matrices were merged in a correlation study and the resulting interpretation ended in a protocol for the evaluation of the N effect during the main phenological stages. The positive effect of N on the measured parameters was observed in both years; however, in the second year, the correlations with the yield were low, being attributed to a complex interaction with climatic conditions. Yield dependence on N was optimally evaluated and monitored with Sentinel-2 data. Two separate relationships between NIR¿red and NDVI¿NIR were identified, suggesting that using remote sensing data can help enhance rice crop management by adjusting nitrogen input based on plant nitrogen concentration and yield estimates. This method has the potential to decrease nitrogen use and environmental pollution, promoting more sustainable rice cultivation practices., This research has been funded by the PREDIC-PRO project SCPP2100C008733XVD, of the State Research Agency of the Ministry of Science, Innovation and Universities, and ACIF Generalitat Valenciana, European Union (European Social Fund. Investing in Your Future) (CIACIF/2021/143) and DETECTORYZA project INNEST/2022/227, INNEST/2022/319 and INNEST/2022/361 Regional Operational Programme, FEDER Comunitat Valenciana de la Innovació, Generalitat Valenciana.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

The yield strikes back: Enhancing the transferability of field scale wheat and barley yield models by leveraging Sentinel-1/2

  • Franch, Belen
  • Moletto-Lobos, Italo
  • Tarin-Mestre, Javier
  • Mascolo, Lucio
  • Vermote, Eric
  • Kalecinski, Natacha
  • Becker-Reshef, Inbal
  • San Francisco, Sara
  • Naranjo, Miguel Angel
  • Paredes, Vanessa
  • Nafria, David
  • Cantero-Martinez, Carlos
  • San Bautista Primo, Alberto|||0000-0003-4846-6611
  • Rubio Michavila, Constanza|||0000-0002-4395-7473
[EN] Accurate and transferable crop monitoring from remote sensing remains challenging because vegetation signals are strongly affected by phenological asynchrony, climatic variability, and sensor-specific responses. Existing approaches rely on local calibrated relationships , limiting their effectiveness in data-sparse regions. This study investigates whether models calibrated on high-quality localized reference data can generalize to other regions by stabilizing sensor-biophysical relationships. The proposed methodology integrates two components (i) thermal time normalization based on growing degree days (GDD) to reduce phenology-driven variability, and (ii) physically motivated optical and optical-SAR fusion indices designed within this normalized framework to enhance the consistency of learned relationships across contrasting environments. The approach was evaluated through within-region cross-season, and cross-region experiments. Results show that GDD normalization improves performance relative to calendar-based approaches by up to 35%. In cross-season validation, fusion-based linear models achieved R-2 > 0.60 for wheat and 0.65 in barley. Cross-region validation shows reduced but meaningful transferability, with both crops reaching R-2 approximate to 0.45, and fusion reducing RMSE by similar to 200 kg ha(-1) compared to optical-only models. Machine-learning models did not improve generalization over simple parametric fits. These findings confirm that stabilizing phenological and multi-sensor relationships is critical for transferring models from data-rich to data-limited areas, providing a foundation for scalable, global agricultural monitoring., This research has been funded by the PREDIC-PRO project SCPP2100C008733XV0 (CPP2021-008733) of the State Research Agency of the Ministry of Science, Innovation and Universities.