European Journal of Agronomy最新文献

{"title":"CO-ResNetRS50-SSL: Enhanced convolution and semi-supervised learning for accurate rice growth stage recognition in complex field conditions","authors":"Changqing Yan ,&nbsp;Guangpeng Yang ,&nbsp;Zeyun Liang ,&nbsp;Han Cheng ,&nbsp;Genghong Wu ,&nbsp;Amit Kumar Srivastava ,&nbsp;Qiang Yu ,&nbsp;Gang Zhao","doi":"10.1016/j.eja.2025.127631","DOIUrl":"10.1016/j.eja.2025.127631","url":null,"abstract":"<div><div>Effective crop management decisions, such as fertilization, irrigation, and crop protection, are closely tied to the crop growth stages. Precise identification of development stages is essential to optimize management practices in line with crop needs. While deep learning has shown promise in identifying growth stages, existing models often face challenges due to limited data availability and reduced accuracy in complex field conditions. To overcome these limitations, this study proposes a semi-supervised image classification method built on an enhanced ResNetRS50 architecture, named CO-ResNetRS50-SSL. This model leverages ResNetRS50 as its backbone, integrating Coordinate Attention (CA) for improved positional feature extraction and Omni-Dimensional Dynamic Convolution (ODConv) to enhance the adaptability of convolutional kernels to varying targets. Additionally, a semi-supervised learning framework is employed to boost generalization while minimizing dependence on labeled data. Ablation experiments show that semi-supervised learning boosted ResNetRS50’s accuracy from 88.58 % to 89.36 %. Adding Coordinate Attention further increased accuracy to 89.89 %, while incorporating ODConv in the final CO-ResNetRS50-SSL model achieved 90.38 % accuracy, 90.59 % precision, and 90.19 % F1 score (with 65.38 M parameters). Comparisons reveal that CO-ResNetRS50-SSL outperforms state-of-the-art models (FasterNet-T1, ShuffleNetV2, Swin Transformer, Vision Transformer, ConvNeXt-base) with highly significant differences (p &lt; 0.001) and delivers robust performance across rice growth stages, with an optimal trade-off at 224 × 224 resolution. CO-ResNetRS50-SSL can accurately detect rice growth stages with limited labeled data, and its improvements in accuracy and generalization are expected to enhance decision-making in precision agriculture, optimizing resource allocation, reducing inputs, and advancing progress in the field of digital agriculture. Future work will focus on improving efficiency in utilizing unlabeled data, ensuring more balanced performance across different growth stages, and enhancing the model’s adaptability to other crops and more complex agricultural scenarios.</div></div>","PeriodicalId":51045,"journal":{"name":"European Journal of Agronomy","volume":"168 ","pages":"Article 127631"},"PeriodicalIF":4.5,"publicationDate":"2025-04-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143786029","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing winter wheat growth indicator prediction with multi-task learning and multi-source data","authors":"Haoxi Song ,&nbsp;Tingxuan Zhuang ,&nbsp;Xueye Li ,&nbsp;Guojie Ruan ,&nbsp;James Schepers ,&nbsp;Dashuai Wang ,&nbsp;Xiaojun Liu ,&nbsp;Yongchao Tian ,&nbsp;Yan Zhu ,&nbsp;Weixing Cao ,&nbsp;Qiang Cao","doi":"10.1016/j.eja.2025.127629","DOIUrl":"10.1016/j.eja.2025.127629","url":null,"abstract":"<div><div>Timely and accurate prediction of wheat growth indicators is crucial for yield enhancement and extreme weather impact mitigation. Research on efficient monitoring of growth indicators using multi-task learning combined with multi-source information remains limited. Furthermore, the growth stage-specific prediction should be emphasized to reveal the effect of growth stages on the indicators. This study aims to predict winter wheat growth indicators at different growth stages using machine learning, deep learning, and multi-task learning based on multi-source and multi-temporal features, such as from spectral, moisture, and meteorological data, to evaluate and improve the accuracy of prediction. Field-collected growth indicators including leaf area index (LAI), chlorophyll (CHL), plant nitrogen accumulation (PNA), plant dry matter (PDM), plant nitrogen content (PNC), nitrogen nutrition index (NNI), and the above features were analyzed alongside feature selection based on Pearson correlation coefficients (PCCfs). Models were developed using Random Forest (RF), Long Short-Term Memory (LSTM), and Multi-Task Learning (MTL), with consideration given to the contribution of features to indicators. The results demonstrated that RF model outperformed LSTM, with average R² values ranging from 0.54 to 0.92 versus 0.08–0.88, respectively. The MTL enhanced model speed and accuracy, particularly with large datasets or deep learning applications. Each indicator exhibited optimal performance at specific growth stages, such as LAI during the jointing and PDM during the flowering. Vegetation Index (VI) emerged as the most significant features for growth indicators, followed by the canopy equivalent water thickness (CEWT) and meteorological features. This study presents a novel approach to winter wheat growth indicator prediction, significantly enhancing prediction accuracy and contributing to the achievement of precise field management.</div></div>","PeriodicalId":51045,"journal":{"name":"European Journal of Agronomy","volume":"168 ","pages":"Article 127629"},"PeriodicalIF":4.5,"publicationDate":"2025-04-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143786030","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Modern maize hybrids have increased grain yield and efficiency tolerances to nitrogen-limited conditions","authors":"Raziel A. Ordóñez ,&nbsp;Lia Olmedo-Pico ,&nbsp;Antonella Ferela ,&nbsp;Slobodan Trifunovic ,&nbsp;Douglas Eudy ,&nbsp;Sotirios Archontoulis ,&nbsp;Tony J. Vyn","doi":"10.1016/j.eja.2025.127621","DOIUrl":"10.1016/j.eja.2025.127621","url":null,"abstract":"<div><h3>Context or problem</h3><div>Previous studies confirmed continued yield improvements in maize (<em>Zea mays</em> L.) since the introduction of hybrids. However, corresponding yield genetic gains (GGs) and associated physiological changes occurring under low- nitrogen (N) conditions are less understood.</div></div><div><h3>Objective or research question</h3><div>This study assessed the growth, yield and key trait GGs of 38 non-GMO legacy hybrids under low- (34–99 kg N ha⁻¹) versus high-N (202–280 kg N ha⁻¹) conditions to understand the role of N in trait changes over time.</div></div><div><h3>Methods</h3><div>Hybrids, commercialized in the USA by Bayer Crop Science from 1983 until 2020, were evaluated in eight site-years N trails at locations across Indiana, Illinois, and Iowa between 2020 and 2022. Seventeen traits were studied pre- and post-anthesis and at physiological maturity. The GGs for each trait are reported as Best Linear Unbiased Predictions (BLUPs), linked to the year of hybrid release.</div></div><div><h3>Results</h3><div>Grain yield increased by 88 kg ha⁻¹ year⁻¹ under low-N and 102 kg ha⁻¹ year⁻¹ under high-N. Over the past decade, modern hybrids grown under low-N outyielded older hybrids grown under high-N. Despite typical trade-offs between yield and grain N concentration, the grain C:N ratio increased in most recent hybrids subjected to low-N, indicating greater carbon (C) accumulation per unit of plant N uptake. Harvest index (HI) increased similarly under both N conditions, with no sign of a plateauing. Newly released hybrids displayed improved N use efficiency at maturity under low-N, and GGs in biomass accumulation during the post-silking period followed observations of enhanced ear dry weight in newer hybrids at the R2 growth stage.</div></div><div><h3>Conclusions</h3><div>Breeding for yield indirectly improved maize tolerance to soils with limited available N and enhanced N use efficiency via more grain C capture per unit plant N uptake.</div></div><div><h3>Implication or significance</h3><div>This study highlights the potential for maize breeders to target specific post-silking traits while simultaneously improving yield and inherent N efficiencies. Furthermore, this research contributes to a deeper understanding of whole-plant N dynamics, and the role of breeding in advancing sustainability. However, future research should also focus on assessing the N rate dependency of GGs in maize hybrids in the context of multiple plant densities or other potential management system interactions.</div></div>","PeriodicalId":51045,"journal":{"name":"European Journal of Agronomy","volume":"168 ","pages":"Article 127621"},"PeriodicalIF":4.5,"publicationDate":"2025-04-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143783046","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The response of crop yield, water and nitrogen use efficiency to organic fertilizer addition: A meta-analysis","authors":"Pengzhao Liu ,&nbsp;Yanrong Lin ,&nbsp;Xiantong Liu ,&nbsp;Zhipeng Li ,&nbsp;Zhen Fan ,&nbsp;Zinan Yan ,&nbsp;Yulin Wu ,&nbsp;Xining Zhao ,&nbsp;Xiaolong Ren ,&nbsp;Xiaoli Chen","doi":"10.1016/j.eja.2025.127628","DOIUrl":"10.1016/j.eja.2025.127628","url":null,"abstract":"<div><div>Organic fertilizer, as a more environmentally friendly alternative to chemical nitrogen fertilizer, has demonstrated potential in enhancing the crop productivity and water use efficiency. However, there is a lack of comprehensive analysis on evapotranspiration, crop yield, as well as water and nitrogen use efficiency in response to organic fertilizer addition. This study evaluated the overall effects of organic fertilizer application on evapotranspiration (ET), crop yield, water use efficiency (WUE) and nitrogen partial productivity (NPFP), based on 560 datasets collected from 124 studies worldwide. The results showed that adding organic fertilizer brought a 9.84 % increase in yield, a 10.22 % increase in WUE and a 2.62 % reduction in NPFP, with no extra ET. In this case, NPFP in organic fertilizer substituting nitrogen fertilizer (OSNF) treatment increased by 10.88 %, while NPFP in nitrogen fertilizer plus organic fertilizer (NPOF) treatment decreased by 21.94 %. Random forest model indicated that the key drivers affecting the efficacy of organic fertilizer were substitution rate (SR), soil organic matter (SOM), pH, organic carbon to nitrogen ratio, and mean annual temperature. Additionally, the organic fertilizer’ efficacy is better with applying poultry manure in higher soil bulk density (&gt;1.4 g cm^–3), low soil total nitrogen (&lt; 5 g kg^–1) and SOM (&lt;10 g kg^–1) exert. Crop yield, WUE, and NPFP all exhibited a quadratic relationship with organic fertilizer substitution rate. The maximum response ratios of yield, WUE, and NPFP were 12.3 %, 11.2 %, and 11.9 %, respectively, with substitution rates of 47.3 %, 46.4 %, and 43.8 %, respectively. Fertilization strategies can be adapted and optimized according to soil characteristics and climate conditions. Our findings offer valuable insights for the development of sustainable green fertilization strategies to enhance crop production.</div></div>","PeriodicalId":51045,"journal":{"name":"European Journal of Agronomy","volume":"168 ","pages":"Article 127628"},"PeriodicalIF":4.5,"publicationDate":"2025-04-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143783120","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing water use efficiency in wheat: 40 Years of genetic gains in the Argentine Pampas","authors":"Facundo Curin ,&nbsp;María E. Otegui ,&nbsp;Fernanda G. González","doi":"10.1016/j.eja.2025.127620","DOIUrl":"10.1016/j.eja.2025.127620","url":null,"abstract":"<div><div>Understanding the impact of breeding on water use and productivity, particularly in underexplored environments where pre-anthesis rainfall is often limited, is crucial for increasing wheat grain yield (GY) under future climatic scenarios. The aim of this work was to study the effects of breeding on GY and its physiological determinants, including water use and water use efficiency among cultivars released to the Argentine market over a 40-year period. Ten Argentinean cultivars classified into early cycle (EC) and late cycle (LC), based on time to anthesis, were grown across three seasons (2015, 2016, and 2017) under irrigated, rainfed, and water restricted conditions. Breeding increased GY by enhancing the harvest index (HI) in LC cultivars and the total biomass (BT) in EC cultivars. These improvements were accompanied by increases in water use efficiency for grain yield production (WUE_GY,ETc) of 1.11 % and 0.45 % per year among LC and EC cultivars, respectively, and a 0.29 % per year improvement in water use efficiency for biomass production (WUE_BT,ETc) among EC cultivars. Crop evapotranspiration (ET_C) decreased by 0.16 % per year exclusively in LC cultivars. Transpiration use efficiency (TE) improved by 0.55 % and 0.45 % per year for LC and EC cultivars, respectively, driven by increases in radiation use efficiency (RUE) of 0.46 % and 0.24 % per year. Notably, no changes in canopy conductance (g_C) were observed. Finally, the proportion of water transpired before anthesis (pPRE_ANT T_C) decreased by 0.45 % per year in LC cultivars. The results of this study indicate that genetic improvement in Argentina has enhanced GY by increasing WUE and RUE, rather than total ET_C. This is a particularly beneficial outcome for rainfed farming systems with limited water availability</div></div>","PeriodicalId":51045,"journal":{"name":"European Journal of Agronomy","volume":"168 ","pages":"Article 127620"},"PeriodicalIF":4.5,"publicationDate":"2025-04-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143761104","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Productivity of saline soils can be improved by integrating soil–crop system management","authors":"Qi Miao ,&nbsp;Shuaibing Wu ,&nbsp;Junchao Li ,&nbsp;Jishi Zhang ,&nbsp;Yunhong Wang ,&nbsp;Gang Wu ,&nbsp;Yixiang Sun ,&nbsp;Hao Ying ,&nbsp;Zhenling Cui","doi":"10.1016/j.eja.2025.127618","DOIUrl":"10.1016/j.eja.2025.127618","url":null,"abstract":"<div><div>Appropriate field management strategies are essential for improving crop productivity and promoting agricultural sustainability in saline–alkaline soils. To explore the potential of synergistic soil improvement and optimal crop management for agricultural production, this study evaluated crop yield, resource utilization, and soil quality in a field experiment with monoculture spring maize cultivation in a coastal saline soil region from 2015 to 2019. Compared with farmers’ practices, improved soil management (gypsum/cattle manure), crop management (variety/density), and a combination of both (ISCM) significantly decreased the sodium content, sodium removal rate, and pH by 51.5–60.2, 9.5–35.2, and 9.9 %, respectively, which increased soil organic carbon and total nitrogen by 31.9 and 22.2 %, respectively. Overall, ISCM was the most prominent in processing. As a result, ISCM increased yield and partial factor productivity of nitrogen by 37.2 and 94.5 %, respectively, and reduced N fertilizer use by 28.6 %. Therefore, soil–crop management, particularly the ISCM strategy, can successfully improve the saline soil environment, increase agricultural productivity, and achieve sustainability under intensive farming practices on saline land.</div></div>","PeriodicalId":51045,"journal":{"name":"European Journal of Agronomy","volume":"168 ","pages":"Article 127618"},"PeriodicalIF":4.5,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143738982","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A new dataset to elucidate inaccurate temperature thresholds masking a hidden source of risk uncertainties in common wheat","authors":"Mael Aubry ,&nbsp;Renan Le Roux,&nbsp;Marie Launay,&nbsp;Iñaki García de Cortázar-Atauri,&nbsp;Carina Furusho-Percot","doi":"10.1016/j.eja.2025.127623","DOIUrl":"10.1016/j.eja.2025.127623","url":null,"abstract":"<div><div>Temperatures outside the optimum range for each phenological phase may wreak cereal yield losses. However, there is no consensus on temperature thresholds for assessing thermal stress in wheat. The wide spread use of these thresholds in the scientific literature is hardly ever considered in climate risk assessments. Our review updates the seminal work of Porter and Gawith (1999) by revising thermal stress thresholds, according to the geographical localisation, wheat type, phenological phases and processes involved. A group of 122 publications on the impact of temperature on wheat crops reveals high variability and inconsistencies in the thresholds. We disentangle some of the factors generating this variability and show that when the sample is filtered by geographical localisation, type of common wheat (winter or spring), and biological process (growth or development), the variability decreases within a range of 35–65 % depending on the stress indicator. This variability leads to considerable uncertainty in the estimation of agroclimatic risks to wheat crops. Indeed, a 2°C difference in the stress threshold (25°C vs. 27°C) for the grain-filling phase reduced the heat stress risk (number of days above the threshold) by up to 45 % on average over a 30-year period in France. These results help interpret previous studies by considering how the chosen threshold position within the range of possible values may lead to under or overestimation. We provide the full database from this review, including metadata defining the validity range of each study’s threshold. We recommend that future studies avoid deterministic thresholds and instead use a range of values to capture uncertainty and minimize conflicting conclusions.</div></div>","PeriodicalId":51045,"journal":{"name":"European Journal of Agronomy","volume":"168 ","pages":"Article 127623"},"PeriodicalIF":4.5,"publicationDate":"2025-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143738981","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Direct measurements unveil overestimations in nitrous oxide emissions from legume-based strip intercropping","authors":"Jinchuan Zhang ,&nbsp;Xin Qian ,&nbsp;Qi Liu ,&nbsp;Yi Xu ,&nbsp;Shengquan Yang ,&nbsp;Shaoyong Meng ,&nbsp;Yuanquan Chen ,&nbsp;Yadong Yang ,&nbsp;Leanne Peixoto ,&nbsp;Zhaohai Zeng ,&nbsp;Jørgen E. Olesen ,&nbsp;Huadong Zang","doi":"10.1016/j.eja.2025.127619","DOIUrl":"10.1016/j.eja.2025.127619","url":null,"abstract":"<div><div>Accurate evaluation of greenhouse gas (GHG) emissions from legume<strong>-</strong>based intercropping is key to supporting the agricultural green transition. However, current methods often ignore the direct emissions from the interspecific interaction row of legume/cereal intercropping. Here, a two-year field experiment was conducted to evaluate the difference between direct measurement and theoretical estimation of N₂O emissions and soil respiration from maize/peanut and maize/soybean strip intercropping. Results showed that theoretical estimation overestimated the N₂O emissions from interaction row by 12 – 15 % compared to direct measurement (<em>P</em> &lt; 0.05). This overestimation occurred mainly during the N₂O emissions peak (lasting 5 – 7 days) after topdressing N fertilization, which accounted for 21 – 49 % of the overestimation. Soil NH₄⁺ and NO₃^- concentrations in the interaction row were 33 – 129 % and 26 – 96 % lower than the weighted average of the crop rows following topdressing N fertilization (<em>P</em> &lt; 0.05). However, the soil respiration of the interaction row did not differ between direct measurement and theoretical estimation, which may indicate N is the key interaction elements for intercropped legume and cereal rather than carbon. Furthermore, the uncertainty in theoretical estimation of N₂O emissions from legume-based intercropping was increases with the decrease of strip width, which mainly due to the increasing contribution of interaction zone. In conclusion, theoretical estimation can lead to an overestimation of N₂O emissions, underscoring the necessity for direct measurement, especially as strip width narrows with increased proportion of interspecific interactions zones.</div></div>","PeriodicalId":51045,"journal":{"name":"European Journal of Agronomy","volume":"168 ","pages":"Article 127619"},"PeriodicalIF":4.5,"publicationDate":"2025-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143734933","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Balancing economic and environmental goals in winter barley: the role of nitrogen timing and genotypes","authors":"Josephine Bukowiecki,&nbsp;Till Rose,&nbsp;Henning Kage,&nbsp;Klaus Sieling","doi":"10.1016/j.eja.2025.127626","DOIUrl":"10.1016/j.eja.2025.127626","url":null,"abstract":"<div><div>Despite a relatively high share of cropping area in northern European agriculture, little is known about how nitrogen (N) fertilization in autumn and spring affects winter barley grain yield and the mechanisms involved. Based on a field trial from 2016 to 2024, the study examined the effect of eight N treatments (two autumn and four spring N levels) and two winter barley genotypes (hybrid vs. line), focusing on green area index (GAI), photosynthetically active radiation (PAR) interception, grain yield and N balance. Results suggest that yield increases from N fertilization were due to higher PAR interception, which boosted grain numbers per square meter. However, the effect of autumn N was minimal when spring N fertilization was sufficient. Although hybrids intercepted less PAR than lines, they used it more efficiently, consistently yielding more. To optimize yield and N use efficiency, we recommend using hybrid genotypes with spring-only N fertilization. Autumn N fertilization leads to higher N balances, making it unsuitable for meeting European environmental goals.</div></div>","PeriodicalId":51045,"journal":{"name":"European Journal of Agronomy","volume":"168 ","pages":"Article 127626"},"PeriodicalIF":4.5,"publicationDate":"2025-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143724289","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Accurate estimation of winter-wheat leaf water content using continuous wavelet transform-based hyperspectral combined with thermal infrared on a UAV platform","authors":"Ning Yang ,&nbsp;Zhitao Zhang ,&nbsp;Junrui Zhang ,&nbsp;Xiaofei Yang ,&nbsp;Hao Liu ,&nbsp;Junying Chen ,&nbsp;Jifeng Ning ,&nbsp;Shikun Sun ,&nbsp;Liangsheng Shi","doi":"10.1016/j.eja.2025.127624","DOIUrl":"10.1016/j.eja.2025.127624","url":null,"abstract":"<div><div>Timely and accurate assessment of crop water status using unmanned aerial vehicle (UAV) imagery is helpful for precision irrigation and field management. The aim of this study is to investigate the application potential of continuous wavelet transform (CWT)-based hyperspectral combined with thermal infrared image data for the estimation of leaf water content (LWC) in winter wheat. This study evaluates the performance of convolutional neural networks (CNN) for feature extraction and long short-term memory (LSTM) networks for sequential data processing in LWC estimation. A UAV platform carrying hyperspectral and thermal infrared sensors was used to collect high spatial resolution images of winter wheat under different water treatments over two years. The LWC was collected simultaneously. The original (OR) and CWT-transformed canopy spectral and textural features, as well as canopy temperature indicators, were extracted from the UAV-based images. On this basis, the LWC estimation model was established using the CNN and LSTM model. The results showed that the combination of thermal features with spectral and texture features significantly improves model performance compared to models built on a single data. The CWT-transformed spectral features improved LWC estimation compared to the original spectrum, with the third scale (CWT3) yielding the best results. Moreover, the CWT-transformed texture at multi-decomposition scales proved to be effective for estimating LWC. Compared to other models, the LSTM model (T-ST_CWT3-LSTM), built by thermal feature fusion with CWT3-based spectral and texture features, achieved the best LWC estimation results, with R² of 0.827 and 0.836, RMSE of 2.575 % and 1.822 %, and MAE of 2.041 % and 1.434 % for 2022 and 2023, respectively. In addition, the robustness of the T-ST_CWT3-LSTM model was successfully verified at different growth stages. Overall, the CWT technique and multi-feature fusion approach provide a valuable technical reference for real-time crop water status monitoring, supporting improved precision irrigation practices and sustainable crop management.</div></div>","PeriodicalId":51045,"journal":{"name":"European Journal of Agronomy","volume":"168 ","pages":"Article 127624"},"PeriodicalIF":4.5,"publicationDate":"2025-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143725273","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}