Search

Surface waterbodies are heavily exposed to pollutants caused by natural disasters and human activities. Empowering sensor technologies in water quality monitoring, sufficient measurements have become available to develop machine learning (ML) models. Numerous ML models have quickly been adopted to predict water quality indicators in various surface waterbodies. This paper reviews 78 recent articles from 2022 to October 2024, categorizing water quality models utilizing ML into three groups: Point-to-Point (P2P), which estimates the current target value based on other measurements at the same time point; Sequence-to-Point (S2P), which utilizes previous time series data to predict the target value at one time point ahead; and Sequence-to-Sequence (S2S), which uses previous time series data to forecast sequential target values in the future. The ML models used in each group are classified and compared according to water quality indicators, data availability, and model performance. Widely used strategies for improving performance, including feature engineering, hyperparameter tuning, and transfer learning, are recognized and described to enhance model effectiveness. The interpretability limitations of ML applications are discussed. This review provides a perspective on emerging ML for surface water quality models. [ABSTRACT FROM AUTHOR]

Purpose: Evaluation of the added value of Dietary Approaches to Stop Hypertension (DASH) and Mediterranean diet scores on the prediction model of the World Health Organization (WHO) to predict 10-year cardiovascular disease (CVD) mortality using the Golestan Cohort Study data. Methods: A total of 44,648 participants (25,268 women and 18,531 men) were included in the final analysis. To assess the external validity of the non-laboratory risk model of WHO, the Area Under the Curve (AUC) and calibration plot methods were used. The multivariate Cox proportional hazards regression analysis was used to evaluate the association of 10-year CVD mortality risk with DASH and Mediterranean scores and their components. The added value of each significant variables was evaluated by the concordance C-statistic and integrated discrimination improvement (IDI). Statistical significance was defined as p-value < 0.05. Results: DASH and Mediterranean diet scores were not significant predictors of 10-year CVD mortality in both genders (p > 0.05). However, sodium and total vegetable in both genders and added sugar in women were significant predictors for 10-year stroke mortality (p < 0.05). Sodium intake in women and monounsaturated fatty acid (MUFA) to saturated fatty acid (SFA) ratio in men had significant associations with 10-year mortality of myocardial infarction/coronary heart disease (MI/CHD). Calculation of IDI showed that none of the evaluated nutritional indices/variables could significantly improve the WHO model performance and predictive ability. Conclusion: Inclusion of DASH and Mediterranean diet scores and their components did not improve WHO risk prediction model performance and predictive ability to predict 10-year CVD mortality. [ABSTRACT FROM AUTHOR]

Thermospheric density, a crucial factor in spacecraft operations, poses a significant challenge in accurate determination due to the intricate coupling of the thermosphere‐ionosphere system. Despite capturing long‐term trends, the widely used empirical models, which are used for low Earth orbit spacecraft operations, often fail to reproduce small‐scale or short‐term variations in the thermosphere. This work presents a novel approach for model improvement. The background model employed is our recently developed empirical model, which blends numerical simulations and satellite measurements. The model residuals are compiled into longitude and latitude bins, and the corresponding basic modes of describing variabilities are subsequently derived via the PCA method. The attendant amplitudes exhibit significant local time and seasonal dependencies, motivating optimized parameterization, that is, static calibration. Moreover, the present study reveals that the most dominant mode correlates to land‐sea contrasts and manifests global synchronicity upon excluding local time and seasonal dependences. This establishes the real‐time model adjustment foundation by exploiting limited calibration observations, which is referred to as dynamic calibration. The results from the dynamic calibration model indicated considerably improved thermospheric representation, especially for small‐scale or short‐term variations, toward a better thermospheric prediction. Plain Language Summary: The thermosphere is the upper atmosphere layer within the altitude range of about 100–800 km. The associated air drag caused in the upper atmosphere is critical for spacecraft operations, especially for the low Earth‐orbiting satellites. The specification or the forecast of the upper atmospheric variations attracts decades‐long efforts from space physicists. However, model uncertainties and systematic biases persist due to insufficient data coverage or missing physical processes in the upper atmosphere. At the same time, lower atmospheric forcing also plays an important role in reshaping the upper atmosphere behaviors while still challenging to characterize, resulting in the most current models primarily driven by solar activities. This work introduces a data assimilation‐based approach to capture density residuals between the background model and satellite observations. The results indicate a considerable model improvement for thermospheric predictions, especially for small‐scale or short‐term variations. Key Points: Small‐scale variabilities of thermospheric densities unresolved in the background model uncertainties are exploredThermospheric density residuals between background model and satellite observations exhibit pronounced local time and seasonal dependenceStatic and dynamic calibration models are developed and demonstrate considerable model improvement [ABSTRACT FROM AUTHOR]

Titanium alloys are widely used in the manufacture of gas turbines' compressor blades. Elucidating their mechanical behavior and strength under damaged conditions is the key to evaluating the equipment's reliability. However, the conventional Johnson-Cook (J-C) constitutive model has limitations in describing the dynamic response of titanium alloy materials under the impact of a high strain rate. In order to solve this problem, the mechanical behavior of a TC4 titanium alloy under high strain rate and different temperature conditions was analyzed by combining experiments and numerical simulations. In this study, the parameters of the J-C model were analyzed in detail, and an improved J-C constitutive model is proposed, based on the new mechanism of the strain rate strengthening effect and the temperature softening effect, which improves the accuracy of the description of strain sensitivity and temperature dependence. Finally, the VUMAT subroutine of ABAQUS software was used for numerical simulation, and the predictive ability of the improved model was verified. The simulation results showed that the maximum prediction error of the traditional J-C model was 23.6%, while the maximum error of the improved model was reduced to 5.6%. This indicates that the improved J-C constitutive model can more accurately predict the mechanical response of a titanium alloy under an impact load and provides a theoretical basis for the study of the mechanical properties of titanium alloy blades under subsequent conditions of foreign object damage. [ABSTRACT FROM AUTHOR]

Coastal areas face severe corrosion issues, posing significant risks and economic losses to equipment, personnel, and the environment. YOLO v5, known for its speed, accuracy, and ease of deployment, has been employed for the rapid detection and identification of marine corrosion. However, corrosion images often feature complex characteristics and high variability in detection targets, presenting significant challenges for YOLO v5 in recognizing and extracting corrosion features. To improve the detection performance of YOLO v5 for corrosion image features, this study investigates two enhanced models: EfficientViT-NWD-YOLO v5 and Gold-NWD-YOLO v5. These models specifically target improvements to the backbone and neck structures of YOLO v5, respectively. The performance of these models for corrosion detection is analyzed in comparison with both YOLO v5 and NWD-YOLO v5. The evaluation metrics including precision, recall, F1-score, Frames Per Second (FPS), pre-processing time, inference time, non-maximum suppression time (NMS), and confusion matrix were used to evaluate the detection performance. The results indicate that the Gold-NWD-YOLO v5 model shows significant improvements in precision, recall, F1-score, and accurate prediction probability. However, it also increases inference time and NMS time, and decreases FPS. This suggests that while the modified neck structure significantly enhances detection performance in corrosion images, it also increases computational overhead. On the other hand, the EfficientViT-NWD-YOLO v5 model shows slight improvements in precision, recall, F1-score, and accurate prediction probability. Notably, it significantly reduces inference and NMS time, and greatly improves FPS. This indicates that modifications to the backbone structure do not notably enhance corrosion detection performance but significantly improve detection speed. From the application perspective, YOLO v5 and NWD-YOLO v5 are suitable for routine corrosion detection applications. Gold-NWD-YOLO v5 is better suited for scenarios requiring high precision in corrosion detection, while EfficientViT-NWD-YOLO v5 is ideal for applications needing a balance between speed and accuracy. The findings can guide decision making for corrosion health monitoring for critical infrastructure in coastal areas. [ABSTRACT FROM AUTHOR]

The movement and deformation of rock strata and the ground surface is a dynamic deformation process that occurs as underground mining progresses. Therefore, the dynamic prediction of three-dimensional surface deformation caused by underground mining is of great significance for assessing potential geological disasters. Synthetic aperture radar interferometry (InSAR) has been introduced into the field of mine deformation monitoring as a new mapping technology, but it is affected by many factors, and it cannot monitor the surface deformation value over the entire mining period, making it impossible to accurately predict the spatiotemporal evolution characteristics of the surface. To overcome this limitation, we propose a new dynamic prediction method (InSAR-DIB) based on a combination of InSAR and an improved Boltzmann (IB) function model. Theoretically, the InSAR-DIB model can use information on small dynamic deformation during mining to obtain surface prediction parameters and further realize a dynamic prediction of the surface. The method was applied to the 1613 (1) working face in the Huainan mining area. The results showed that the estimated mean error of the predicted surface deformation during mining was between 80.2 and 112.5 mm, and the estimated accuracy met the requirements for mining subsidence monitoring. The relevant research results are of great significance, and they support expanding the application of InSAR in mining areas with large deformation gradients. [ABSTRACT FROM AUTHOR]

Copyright of Remote Sensing for Natural Resources is the property of Remote Sensing for Natural Resources Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Abstract Thermospheric density, a crucial factor in spacecraft operations, poses a significant challenge in accurate determination due to the intricate coupling of the thermosphere‐ionosphere system. Despite capturing long‐term trends, the widely used empirical models, which are used for low Earth orbit spacecraft operations, often fail to reproduce small‐scale or short‐term variations in the thermosphere. This work presents a novel approach for model improvement. The background model employed is our recently developed empirical model, which blends numerical simulations and satellite measurements. The model residuals are compiled into longitude and latitude bins, and the corresponding basic modes of describing variabilities are subsequently derived via the PCA method. The attendant amplitudes exhibit significant local time and seasonal dependencies, motivating optimized parameterization, that is, static calibration. Moreover, the present study reveals that the most dominant mode correlates to land‐sea contrasts and manifests global synchronicity upon excluding local time and seasonal dependences. This establishes the real‐time model adjustment foundation by exploiting limited calibration observations, which is referred to as dynamic calibration. The results from the dynamic calibration model indicated considerably improved thermospheric representation, especially for small‐scale or short‐term variations, toward a better thermospheric prediction.

The simulation of tropical precipitation has been a challenge to climate models. The multi-model ensemble mean of CMIP6 models only show limited improvement relative to the CMIP3 and CMIP5 models. However, a simple ensemble mean may mask the improvement of individual models. Here we evaluated 20 CMIP6 models and their corresponding earlier version in CMIP5. The results show that the CMIP6 models is significantly improved in the tropical precipitation compared to their counterparts in CMIP5, and the alleviation of bias mainly happened in the top ten model pairs with largest RMSE reduction (TOP10). For the mean of TOP10, the antisymmetric (symmetric) bias mode in CMIP5 is significantly reduced by 55% (78%). Further energetics evaluation shows that, the CMIP5-to-CMIP6 decrease of antisymmetric bias in the bulk of TOP10 models is accompanied by more realistic southward cross-equatorial atmospheric energy transport ( AET EQ ), which is mainly contributed by the better representation of the extra-tropical surface turbulent flux ( STF ). On the other hand, the decrease in the symmetric bias of TOP10 models is associated with the enlargement of the negative bias in the seasonal contrast of AET EQ , which is caused by the alleviation of the biases in the seasonal contrast of extra-tropical STF . Our analysis revealed the improvement in the simulation of tropical precipitation in CMIP6, and we pointed out that the improvement is associated with the model-generational changes in the simulated atmospheric energy balances. [ABSTRACT FROM AUTHOR]

Plant diseases pose a significant threat to global agricultural productivity and food safety. Early detection and accurate identification of these diseases are essential for effective disease management strategies. Traditional plant disease identification mainly relies on manual observation and experienced expert judgement, which has the disadvantages of being time‐consuming, labour‐intensive and low efficiency. Given the above problems, this study proposes a method for identifying apple leaf diseases based on a convolutional neural network combining hybrid attention and bidirectional long short‐term memory (BiLSTM). Appropriate apple leaf disease samples were selected from multiple public data sets to form an experimental data set. Then, the data set is imported into the improved convolutional neural network for training. Based on the original ResNet18 model, a new convolutional neural network, AppleNet, is constructed by adding a hybrid attention module and modifying the classifier structure. The experimental results show that the average recognition accuracy of AppleNet is 94.66%, which is 2.47% higher than that of the ResNet18 network. In addition, the training time of the model is only slightly increased. The ablation experiment further verified the effectiveness of the model modification. Compared with other advanced models in recognition accuracy and model training time, the superiority of AppleNet is confirmed. This study verifies that deep learning has great potential and application prospects in plant disease identification and provides a new technical solution for intelligent and convenient plant disease identification. [ABSTRACT FROM AUTHOR]

Track fastener defect detection is an essential component in ensuring railway safety operations. Traditional manual inspection methods no longer meet the requirements of modern railways. The use of deep learning image processing techniques for classifying and recognizing abnormal fasteners is faster, more accurate, and more intelligent. With the widespread use of unmanned aerial vehicles (UAVs), conducting railway inspections using lightweight, low-power devices carried by UAVs has become a future trend. In this paper, we address the characteristics of track fastener detection tasks by improving the YOLOv4-tiny object detection model. We improved the model to output single-scale features and used the K-means++ algorithm to cluster the dataset, obtaining anchor boxes that were better suited to the dataset. Finally, we developed the FPGA platform and deployed the transformed model on this platform. The experimental results demonstrated that the improved model achieved an mAP of 95.1% and a speed of 295.9 FPS on the FPGA, surpassing the performance of existing object detection models. Moreover, the lightweight and low-powered FPGA platform meets the requirements for UAV deployment. [ABSTRACT FROM AUTHOR]

The Sichuan-Xizang Railway is an important part of the railway network in China, and geological disasters, such as mountain floods and landslides, frequently occur in this region. Precipitation is an important cause of these disasters; therefore, accurate simulation of the precipitation in this region is highly important. In this study, the descriptions for uncertain processes in the cloud microphysics scheme are improved; these processes include cloud droplet activation, cloud-rain autoconversion, rain accretion by cloud droplets, and the entrainment-mixing process. In the default scheme, the cloud water content of different sizes corresponds to the same cloud droplet concentration, which is inconsistent with the actual content; this results in excessive cloud droplet size, unreasonable related conversion rates of microphysical process (such as cloud-rain autoconversion), and an overestimation of precipitation. Our new scheme overcomes the problem of excessive cloud droplet size. The processes of cloudrain autoconversion and rain accretion by cloud droplets are similar to the stochastic collection equation, and the mixing mechanism of cloud droplets is more consistent with that occurred during the actual physical process in the cloud. Based on the new and old schemes, multiple precipitation processes in the flood season of 2021 along the Sichuan-Xizang Railway are simulated, and the results are evaluated using ground observations and satellite data. Compared to the default scheme, the new scheme is more suitable for the simulation of cloud physics, reducing the simulation deviation of the liquid water path and droplet radius from 2 times to less than 1 time and significantly alleviating the overestimation of precipitation intensity and range of precipitation center. The average root-mean-square error is reduced by 22%. Our results can provide a scientific reference for improving precipitation forecasting and disaster prevention in this region. [ABSTRACT FROM AUTHOR]

Copyright of Acta Scientiarum Naturalium Universitatis Sunyatseni / Zhongshan Daxue Xuebao is the property of Sun-Yat-Sen University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Titanium alloys are widely used in the manufacture of gas turbines’ compressor blades. Elucidating their mechanical behavior and strength under damaged conditions is the key to evaluating the equipment’s reliability. However, the conventional Johnson-Cook (J-C) constitutive model has limitations in describing the dynamic response of titanium alloy materials under the impact of a high strain rate. In order to solve this problem, the mechanical behavior of a TC4 titanium alloy under high strain rate and different temperature conditions was analyzed by combining experiments and numerical simulations. In this study, the parameters of the J-C model were analyzed in detail, and an improved J-C constitutive model is proposed, based on the new mechanism of the strain rate strengthening effect and the temperature softening effect, which improves the accuracy of the description of strain sensitivity and temperature dependence. Finally, the VUMAT subroutine of ABAQUS software was used for numerical simulation, and the predictive ability of the improved model was verified. The simulation results showed that the maximum prediction error of the traditional J-C model was 23.6%, while the maximum error of the improved model was reduced to 5.6%. This indicates that the improved J-C constitutive model can more accurately predict the mechanical response of a titanium alloy under an impact load and provides a theoretical basis for the study of the mechanical properties of titanium alloy blades under subsequent conditions of foreign object damage.

Coastal areas face severe corrosion issues, posing significant risks and economic losses to equipment, personnel, and the environment. YOLO v5, known for its speed, accuracy, and ease of deployment, has been employed for the rapid detection and identification of marine corrosion. However, corrosion images often feature complex characteristics and high variability in detection targets, presenting significant challenges for YOLO v5 in recognizing and extracting corrosion features. To improve the detection performance of YOLO v5 for corrosion image features, this study investigates two enhanced models: EfficientViT-NWD-YOLO v5 and Gold-NWD-YOLO v5. These models specifically target improvements to the backbone and neck structures of YOLO v5, respectively. The performance of these models for corrosion detection is analyzed in comparison with both YOLO v5 and NWD-YOLO v5. The evaluation metrics including precision, recall, F1-score, Frames Per Second (FPS), pre-processing time, inference time, non-maximum suppression time (NMS), and confusion matrix were used to evaluate the detection performance. The results indicate that the Gold-NWD-YOLO v5 model shows significant improvements in precision, recall, F1-score, and accurate prediction probability. However, it also increases inference time and NMS time, and decreases FPS. This suggests that while the modified neck structure significantly enhances detection performance in corrosion images, it also increases computational overhead. On the other hand, the EfficientViT-NWD-YOLO v5 model shows slight improvements in precision, recall, F1-score, and accurate prediction probability. Notably, it significantly reduces inference and NMS time, and greatly improves FPS. This indicates that modifications to the backbone structure do not notably enhance corrosion detection performance but significantly improve detection speed. From the application perspective, YOLO v5 and NWD-YOLO v5 are suitable for routine corrosion detection applications. Gold-NWD-YOLO v5 is better suited for scenarios requiring high precision in corrosion detection, while EfficientViT-NWD-YOLO v5 is ideal for applications needing a balance between speed and accuracy. The findings can guide decision making for corrosion health monitoring for critical infrastructure in coastal areas.

The movement and deformation of rock strata and the ground surface is a dynamic deformation process that occurs as underground mining progresses. Therefore, the dynamic prediction of three-dimensional surface deformation caused by underground mining is of great significance for assessing potential geological disasters. Synthetic aperture radar interferometry (InSAR) has been introduced into the field of mine deformation monitoring as a new mapping technology, but it is affected by many factors, and it cannot monitor the surface deformation value over the entire mining period, making it impossible to accurately predict the spatiotemporal evolution characteristics of the surface. To overcome this limitation, we propose a new dynamic prediction method (InSAR-DIB) based on a combination of InSAR and an improved Boltzmann (IB) function model. Theoretically, the InSAR-DIB model can use information on small dynamic deformation during mining to obtain surface prediction parameters and further realize a dynamic prediction of the surface. The method was applied to the 1613 (1) working face in the Huainan mining area. The results showed that the estimated mean error of the predicted surface deformation during mining was between 80.2 and 112.5 mm, and the estimated accuracy met the requirements for mining subsidence monitoring. The relevant research results are of great significance, and they support expanding the application of InSAR in mining areas with large deformation gradients.

Soil erosion is an important cause of global land degradation, and accurate monitoring of it is essential. The Soil and Water Assessment Tool (SWAT), a distributed hydrological model, is an advanced technique for predicting soil erosion at watershed scale. However, as the erosion framework was established in gently sloping land, SWAT is limited in predicting soil erosion in some highland and mountainous regions. Therefore, this study suggested a method to integrate the sediment transport theoretical formula that can reflect the morphology of gully regions into SWAT to obtain SWAT‐S to enhance the calculation performance of sediment load, and the SWAT‐S was evaluated according to the coefficient of determination (R2), Nash‐Sutcliffe coefficient (NSE), Percent‐Bias (P‐BIAS) and root mean square errors (RMSE)‐observations SD ratio (RSR) in the Yanhe basin on the Chinese Loess Plateau. The results showed that SWAT‐S is more successful in reproducing the monthly sediment load, with R2, NSE, |P‐BIAS| and RSR were changed by 5.08%, 17.65%, −2.92% and −10.00% in the calibration, as well as by 1.18%, 10.39%, 45.45% and −18.75% in the validation of the SWAT‐S compared to SWAT. Meanwhile, SWAT‐S estimates 2.66 × 106 t more sediment than SWAT during the June–September flood season and better matches observed data. In total, the revised SWAT can improve the performance of sediment estimation, which is beneficial for the wider application of the model in more regions of the world. [ABSTRACT FROM AUTHOR]

Abstract Clouds are the primary source of uncertainty in the prediction of climate change. To reduce the uncertainty of cloud simulations and overcome this difficulty in prediction, many climate modeling centers are now developing a new type of climate model, the global nonhydrostatic atmospheric model, which reduces the uncertainty arising from a cumulus parameterization by computing clouds explicitly using a cloud microphysics scheme. Among the global nonhydrostatic atmospheric models used in recent intercomparison studies, NICAM aims to project climate change by improving our understanding of cloud changes due to warming and related physical processes. NICAM is the first global nonhydrostatic model and was developed by our research team. This review summarizes the outcomes of a recent major five-year research program in Japan for studying climate using NICAM, as well as providing an overview of current issues regarding the use of global kilometer-scale simulations in high-resolution climate modeling.

ABSTRACT Accurate wood volume predictions are critical in hyperdiverse forests because each species has specific size and shape traits. Although generic models at a multispecies level were widely used in Amazonian managed forests, they are subject to more significant bias due to interspecific variability. We used an extensive database of wood volume collected in managed forests to test the hypothesis that generic models violate the independence assumption due to that predictions vary with species-specific size. Our hypothesis was proved as residuals of the generic model were conditioned to species and specific size. The multilevel models were more accurate both in fitting and validation procedures, and accounted for variance derived from species and specific size, providing a more reliable prediction. However, we found that the size-specific models have a similar predictive ability to species-specific models for new predictions. This implies more practical estimates in hyperdiverse forests where fitting species-specific models can be complex. The findings are crucial for sustainable forest management as they allow for more reliable wood volume estimates, leading to less financial uncertainty and preventing damage to forest stocks through under or over-exploitation.

This study aims to find a pattern of correcting mathematical understanding errors in solving mathematical problems for junior high school students. The method used in this research is qualitative. Data collection using triangulation technique: observation, test results, and interviews. Interviews were conducted on three subjects for one month at MTs. The instrument used is a mathematical problem-solving test. This study found patterns of improvement in mathematical problems in geometry, namely the area and perimeter of rectangles and lines. This study found that the pattern of mathematical understanding errors can be corrected by applying patterns such as reminding the material of rectangular shapes, parallelograms, algebraic forms, and units of measure, writing down everything that is known in the problem, and making an example of a variable if it cannot be written directly. The pattern of improving mathematical understanding can solve the mathematics problems of junior high school students. Furthermore, this pattern of mathematical understanding errors can be applied by teachers to improve math problems and embed geometric concepts in mathematics learning. [ABSTRACT FROM AUTHOR]

Plant diseases are common factors restricting crop yield and quality. For individuals without training, it is still difficult to identify various diseases accurately because many leaf diseases are similar in colour, shape and other aspects. Requiring an expert diagnosis can cause delays and lead to increasing costs. Therefore, this paper proposes an improved convolutional neural network for recognition of tomato leaf diseases. The experimental dataset was from the publicly available dataset PlantVillage. Based on the original ResNet18 model, a new convolutional neural network, TomatoNet, was constructed by adding a squeeze‐and‐excitation module and modifying the classifier structure. The results show that the average recognition accuracy of the TomatoNet network is 99.63%, which is 0.53% higher than the ResNet18 network. In addition, the recognition accuracy improved from 88.97% to 98.35% after the improvement of the AlexNet network. Finally, the superiority of TomatoNet was verified by comparison with other advanced models. This experiment verifies the feasibility of a deep learning algorithm for plant leaf disease recognition, which can provide a more efficient and convenient solution for detecting plant leaf disease. [ABSTRACT FROM AUTHOR]

【Objective】 The soil conservation service curve number (SCS-CN) method is an empirical model developed in the 1950s to estimate event-based rainfall-runoff response. We propose an optimized SCS-CN in this paper and apply it to calculate runoffs in different underlying surfaces in Yanqing District of Beijing. 【Method】 Seven different underlying surfaces were selected, and long-term measured rainfall and runoff from each is used to establish and validate the improved SCS-CN model. 【Result】 ①The results calculated using the standard SCS-CN model are higher than the measured values from all seven underlying surfaces. ②The improved SCS-CN model is more accurate than the standard SCS model to reproduce the measured data. ③The improved SCS-CN model is most accurate for estimating runoffs in fallow land, followed by forest land and cultivated land. 【Conclusion】 The standard SCS-CN model is not suitable for calculating runoff in the seven underlying surfaces in Yanqing District of Beijing, and the proposed SCS-CN model significantly improves it.

Clouds are the primary source of uncertainty in the prediction of climate change. To reduce the uncertainty of cloud simulations and overcome this difficulty in prediction, many climate modeling centers are now developing a new type of climate model, the global nonhydrostatic atmospheric model, which reduces the uncertainty arising from a cumulus parameterization by computing clouds explicitly using a cloud microphysics scheme. Among the global nonhydrostatic atmospheric models used in recent intercomparison studies, NICAM aims to project climate change by improving our understanding of cloud changes due to warming and related physical processes. NICAM is the first global nonhydrostatic model and was developed by our research team. This review summarizes the outcomes of a recent major five-year research program in Japan for studying climate using NICAM, as well as providing an overview of current issues regarding the use of global kilometer-scale simulations in high-resolution climate modeling. [ABSTRACT FROM AUTHOR]

As a new generation of global climate models, monsoon simulation in Coupled Model Intercomparison Project (CMIP) Phase 6 models is of great concern to climate modeling community. Using 21 CMIP3 models, 28 CMIP5 models and 38 CMIP6 models, we show evidence that the long-standing dry biases in South Asia (SA) are resulted from less rainfall with both less frequency and intensity in a shortened monsoon season. By evaluating several key metrics, we identify that the monsoon rainfall simulation in CMIP6 models has improved in both of the multimodel ensemble mean (MME) and individual models, consistent with the improvements in monsoon annual cycle and rainfall characteristics. Further analyses and sensitivity experiments show that the cold SST biases all year round over the northern Indian Ocean (NIO) are important sources for the persistent dry biases in the CMIPs' models. The cooling effect of SST biases on the tropospheric temperature becomes increasingly prominent since the boreal spring, weakening the baroclinity of monsoon circulation via the thermal wind relationship and eventually resulting in insufficient monsoon rainfall. Comparison across the three generation CMIP models also confirms that the improvement of SA summer rainfall simulation in CMIP6 MME benefits from the reduction of NIO SST biases. This study highlights the importance of improving SST simulation in reducing the monsoon rainfall biases. [ABSTRACT FROM AUTHOR]

Explainable Artificial Intelligence (XAI) is an emerging research field bringing transparency to highly complex and opaque machine learning (ML) models. Despite the development of a multitude of methods to explain the decisions of black-box classifiers in recent years, these tools are seldomly used beyond visualization purposes. Only recently, researchers have started to employ explanations in practice to actually improve models. This paper offers a comprehensive overview over techniques that apply XAI practically to obtain better ML models, and systematically categorizes these approaches, comparing their respective strengths and weaknesses. We provide a theoretical perspective on these methods, and show empirically through experiments on toy and realistic settings how explanations can help improve properties such as model generalization ability or reasoning, among others. We further discuss potential caveats and drawbacks of these methods. We conclude that while model improvement based on XAI can have significant beneficial effects even on complex and not easily quantifiable model properties, these methods need to be applied carefully, since their success can vary depending on a number of factors, such as the model and dataset used, or the employed explanation method. [Display omitted] • We show how and why XAI improves several model properties in toy experiments. • We develop a unifying theoretical framework for XAI-based model improvement. • We extensively review and discuss existing approaches that use XAI to improve models. • We provide experiment-supported practical recommendations for XAI-based augmentation. [ABSTRACT FROM AUTHOR]

Plant species recognition is an important research area in image recognition in recent years. However, the existing plant species recognition methods have low recognition accuracy and do not meet professional requirements in terms of recognition accuracy. Therefore, ShuffleNetV2 was improved by combining the current hot concern mechanism, convolution kernel size adjustment, convolution tailoring, and CSP technology to improve the accuracy and reduce the amount of computation in this study. Six convolutional neural network models with sufficient trainable parameters were designed for differentiation learning. The SGD algorithm is used to optimize the training process to avoid overfitting or falling into the local optimum. In this paper, a conventional plant image dataset TJAU10 collected by cell phones in a natural context was constructed, containing 3000 images of 10 plant species on the campus of Tianjin Agricultural University. Finally, the improved model is compared with the baseline version of the model, which achieves better results in terms of improving accuracy and reducing the computational effort. The recognition accuracy tested on the TJAU10 dataset reaches up to 98.3%, and the recognition precision reaches up to 93.6%, which is 5.1% better than the original model and reduces the computational effort by about 31% compared with the original model. In addition, the experimental results were evaluated using metrics such as the confusion matrix, which can meet the requirements of professionals for the accurate identification of plant species. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Irrigation & Drainage is the property of Journal of Irrigation & Drainage Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In the engine room of intelligent ships, visual recognition is an essential technical precondition for automatic inspection. At present, the problems of visual recognition in marine engine rooms include missing detection, low accuracy, slow speed, and imperfect datasets. For these problems, this paper proposes a marine engine room equipment recognition model based on the improved You Only Look Once v5 (YOLOv5) algorithm. The channel pruning method based on batch normalization (BN) layer weight value is used to improve the recognition speed. The complete intersection over union (CIoU) loss function and hard-swish activation function are used to enhance detection accuracy. Meanwhile, soft-NMS is used as the non-maximum suppression (NMS) method to reduce the false rate and missed detection rate. Then, the main equipment in the marine engine room (MEMER) dataset is built. Finally, comparative experiments and ablation experiments are carried out on the MEMER dataset to verify the strategy's efficacy on the model performance boost. Specifically, this model can accurately detect 100.00% of diesel engines, 95.91% of pumps, 94.29% of coolers, 98.54% of oil separators, 64.21% of meters, 60.23% of reservoirs, and 75.32% of valves in the actual marine engine room. [ABSTRACT FROM AUTHOR]

We proposed a method to improve the forecasts of landfalling tropical cyclones (LTCs) by constraining the "cloud physics" with Global Precipitation Measurement (GPM) satellite observations. Eight typical LTCs that are well observed by GPM satellite in the Asia‐Pacific region from 2015 to 2021 are selected to verify the feasibility of this method. Using a cloud‐resolving model, the LTCs are simulated for 3 days with both the original and modified microphysics scheme for comparison. The improvement of LTC forecasts is evaluated in terms of hydrometeor structure, amplitude, and location. Most notably, the structure forecast of condensed water improved up to 32% on average for all LTCs. The location forecast and amplitude forecast of condensed water also improved to varying degrees. Moreover, it is found that the error of LTC forecasts was reduced even more by using microphysics constraints from GPM observation than that by assimilating GPM data directly in other research. Plain Language Summary: The skill of Landfalling tropical cyclone (LTC) forecasts has increased in recent years but lacks to some degree in accuracy, refinement, and discrimination. It is still difficult for operational forecasters to accurately predict the spatial structure and precise location of condensed water in a LTC, which is exactly the need for disaster precaution and reduction. Herein, we find an approach to improve operational LTC forecasts through the fusion of Global Precipitation Measurement (GPM) measurements into cloud‐resolving models. By using GPM satellite data as constraints on cloud microphysical processes, the LTC forecasts produce significant improvements especially for the structure forecast of LTC hydrometeors. In addition, the errors of 89‐GHz brightness temperature reduce by ∼2.36 K in the forecast of Category 4 LTC Lan (2017), and ∼2.31 K in the forecast of Category 5 LTC Goni (2020), respectively. Such degree of improvements is similar to or even greater than the magnitude of direct GPM satellite data assimilation in LTC forecasts, so that constraining the microphysics model with GPM observations could be a substantial benefit for LTC predictions. Key Points: Direct constraints on cloud microphysics model are established using Global Precipitation Measurement observationGreat improvement is achieved in forecasting the spatial structure of landfalling tropical cyclone (LTC) hydrometeorsEight typical LTC cases are investigated to ensure robust improvement of their forecasts [ABSTRACT FROM AUTHOR]

Extreme high-temperature stress (HTS) associated with climate change poses potential threats to wheat grain yield and quality. Wheat grain protein concentration (GPC) is a determinant of wheat quality for human nutrition and is often neglected in attempts to assess climate change impacts on wheat production. Crop models are useful tools for quantification of temperature impacts on grain yield and quality. Current crop models either cannot simulate or can simulate only partially the effects of HTS on crop N dynamics and grain N accumulation. There is a paucity of observational data on crop N and grain quality collected under systematic HTS scenarios to develop algorithms for model improvement as well as evaluate crop models. Two-year phytotron experiments were conducted with two wheat cultivars under HTS at anthesis, grain filling, and both stages. HTS significantly reduced total aboveground N and increased the rate of grain N accumulation, while total aboveground N and the rate of grain N accumulation were more sensitive to HTS at anthesis than at grain filling. The observed relationships between total aboveground N, rate of grain N accumulation, and HTS were quantified and incorporated into the WheatGrow model. The new HTS routines improved simulation of the dynamics of total aboveground N, grain N accumulation, and GPC by the model. The improved model provided better estimates of total aboveground N, grain N accumulation, and GPC under HTS (the normalized root mean square error was reduced by 40%, 85%, and 80%, respectively) than the original WheatGrow model. The improvements in the model enhance its applicability to the assessment of climate change effects on wheat grain quality by reducing the uncertainties of simulating N dynamics and grain quality under HTS.

Accurately assessing the health risks posed by major contaminants is essential for protecting groundwater. However, the complexity of pollution sources and the uncertainty of parameters pose challenges for quantitative health risk assessment. In this study, a source-oriented groundwater risk evaluation process was improved by screening key pollutants, employing a combined hydrochemical and positive matrix factorization (PMF) approach for source apportionment, and incorporating two-dimensional Monte Carlo simulation for risk characterization. The application of this process to groundwater assessment in Central Jiangxi Province identified NO 3 -, F-, Se and Mn as the key pollutants. The pollution sources were anthropogenic activities, rock dissolution, regional geological processes, and ion exchange. Anthropogenic sources contributed 36.8 % and 28.8 % of the pollution during the wet season and dry season, respectively, and accounted for more than half of the health risks. NO 3 - from anthropogenic sources was the primary controlling pollutant. Additionally, the risk assessment indicated that children were at the highest health risk during the dry season, with ingestion rate suggested to be controlled below 1.062 L·day−1 to make the health risk within an acceptable range. The improved assessment methodology could provide more accurate results and recommended intakes. [Display omitted] • Hydrochemistry and PMF methods were combined for pollution source analysis. • 2D-MCS was incorporated to improve the accuracy of health risk assessment. • NO 3 -, F-, Se and Mn were the key pollutants. • Children were identified as a high-risk group and dry season as a high-risk period. • NO 3 - from anthropogenic and F- from natural were the main sources of risk. [ABSTRACT FROM AUTHOR]

As one of the research highlights of physics-level modelling, human mobility has generated numerous universal laws over the past decades. However, the majority of research has concentrated on the intracity networks, leaving the intercity mobility systems with insufficient attention despite its increasingly crucial role in the development of urban agglomerations. Related research gaps further extend to the limited understanding of spatiotemporal heterogeneity in intercity human mobility. To bridge these gaps, our study systematically validated and improved the modelling framework of intercity mobility flows utilizing real-world data sources. Specifically, building upon the nationwide Location-based Services (LBS) datasets in China, the applicability of classic human mobility models, including gravity model and intervening opportunities-class models, was extensively explored in the intercity domain by developing fitting models that incorporated multi-class urban attributes. Then, we contributed to proposing improved models that consider the diverse attraction effect of the origin and potential destinations. Moreover, our research scope was expanded to incorporate spatiotemporal heterogeneity through model comparisons among various city sets during both regular period and holiday. The findings suggested that our improved models effectively enhance the modelling accuracy while strengthening the explanatory power. They especially demonstrate a balanced performance even when handling datasets with spatiotemporal heterogeneity. Consequently, this study provides valuable insights into understanding intercity human mobility from the intrinsic mechanism of opportunity attraction. Our models hold practical significance in accurately modelling intercity mobility flows utilizing observable urban attributes and spatial layouts, further providing effective tools for preemptive traffic management. [ABSTRACT FROM AUTHOR]

As more and more attention is paid to the formation of models of new service development, it is noticed that the compositions of a new model are usually based only on examples of previous models that were presented in the scientific literature. Although the practical application of the proposed models in any organizational condition is often verified by empirical studies, the revisions of the validity of model compositions are still quite fragmented. The lack of such research does not give the possibility for any further improvement of the model or elimination of the revealed shortcomings and it makes the proper preparation for testing the proposed model in a company complicated. In order to evaluate the suitability of the composition of the newly formed model of service development in the conditions of modern service companies, an empirical evaluation was performed by using the methods of abstraction, comparative analysis, synthesis, and expert survey. According to the results of the research, the suggestions for the improvement of the new service development model were made and guidelines for further research were formulated.

To better understand how climate change might influence global canola production, scientists from six countries have completed the first inter-comparison of eight crop models for simulating growth and seed yield of canola, based on experimental data from six sites across five countries. A sensitivity analysis was conducted with a combination of five levels of atmospheric CO2 concentrations, seven temperature changes, five precipitation changes, together with five nitrogen application rates. Our results were in several aspects different from those of previous model inter-comparison studies for wheat, maize, rice, and potato crops. A partial model calibration only on phenology led to very poor simulation of aboveground biomass and seed yield of canola, even from the ensemble median or mean. A full calibration with additional data of leaf area index, biomass, and yield from one treatment at each site reduced simulation error of seed yield from 43.8 to 18.0%, but the uncertainty in simulation results remained large. Such calibration (with data from one treatment) was not able to constrain model parameters to reduce simulation uncertainty across the wide range of environments. Using a multi-model ensemble mean or median reduced the uncertainty of yield simulations, but the simulation error remained much larger than observation errors, indicating no guarantee that the ensemble mean/median would predict the correct responses. Using multi-model ensemble median, canola yield was projected to decline with rising temperature (2.5–5.7% per °C), but to increase with increasing CO2 concentration (4.6–8.3% per 100-ppm), rainfall (2.1–6.1% per 10% increase), and nitrogen rates (1.3–6.0% per 10% increase) depending on locations. Due to the large uncertainty, these results need to be treated with caution. We further discuss the need to collect new data to improve modelling of several key physiological processes of canola for increased confidence in future climate impact assessments. [ABSTRACT FROM AUTHOR]

Soil erosion is a major and costly environmental hazard worldwide. Obtaining a thorough understanding of the soil erosion process is essential for erosion control. The high erodibility of the saline‐sodic soil in coastal areas causes rills to develop rapidly and erosion patterns change substantially during the rainfall process. Understanding the temporal and spatial dynamics of the soil surface morphology is crucial for revealing the mechanism behind the severe soil erosion process, but simulating these dynamics with many existing empirical or physical models is difficult. Cellular automata (CA) provides a simple alternative approach with high computational efficiency for modelling the complex patterns of soil erosion evolution using local interactions rules. However, previously established CA models poorly simulated the lateral erosion. By dividing the runoff amount into a discrete variable of every constituent cell and considering the runoff flow velocity when calculating the flow direction, the CA model developed here, based on discrete physical variables (CADV), could simulate soil erosion rate dynamics and rill morphological development with low relative errors. The variations in the simulated sediment yield rates during rainfall events were consistent with the experimental measurements, and the relative errors of cumulative sediment yields between the simulated and measured values were lower than 5% from the 38th to 60th min (the end of rainfall event) during the rainfall. The morphological characteristics of slope erosion with relative errors in rill length, rill width, and rill depth were lower than 5% at specific times during rainfall events. Specifically, regarding the rill width parameter, the relative errors of the maximum and average rill widths were both lower than 5% when the rainfall event ended at the 60th min. The results show that the CADV is an appropriate tool for describing the soil erosion process and improving our understanding of the erosion process characteristics of saline‐sodic soil slopes in a coastal region of China. Highlights: Cellular automata based on discrete variables (CADV) improves soil erosion simulation.CADV simulates soil loss and rill morphological development with low relative errors.CADV improves insight into the mechanism of soil erosion in coastal area.CADV provides a new approach to improve soil erosion modelling with cellular automata. [ABSTRACT FROM AUTHOR]

Evapotranspiration (ET) is a key component of the hydrological cycle, but traditional monitoring approaches are always based on measurements, which cannot satisfy the requirements of research on a regional scale. Hence, ET estimation by remote sensing is essential. MOD16 is a remote-sensing model based on the P-M equation and has good applicability. However, it describes soil moisture indirectly by RH, etc., which may cause uncertainties in ET estimating, so this study attempts to utilize the NDWI as a supplement to soil moisture information and makes improvements on the MOD16 model (with the resultant new model being named MOD16-sm). Specific work includes two aspects: one is model verification through making comparisons between ET estimates and measurements, and the other is a model application effect test analyzing the spatiotemporal characteristics of ET and exploring how ET responds to climate and land-use changes. Model verification indicated that the accuracy of the improved MOD16-sm model increased, with a higher R2 of 0.71, a lower RMSE 0.9 mm, and a lower MAE 0.91 mm, and that the improved MOD16-sm model was convincing. The application effect test of the MOD16-sm model showed that the average relative change rate of annual ET was 1.7%, showing an upward trend, and areas with growth trends of ET also had high vegetation coverage. As for the impacts of climate and land-use changes on ET, ET was positively correlated with precipitation, whereas it had no relevant correlation with air temperature in most areas, and the ET of all land-use types displayed significant increasing trends resulting from climate change. The application effect test demonstrated that ET estimates by the improved MOD16-sm model were reasonable. [ABSTRACT FROM AUTHOR]

The increasing frequency and intensity of low-temperature events in temperate and cold rice production regions threaten rice yields under climate change. While process-based crop models can project climate impacts on rice yield, their accuracy under low-temperature conditions has not been well-evaluated. Our six-year chamber experiments revealed that low temperatures reduce spikelet fertility from panicle initiation to flowering, grain number per spike during panicle development, and grain weight during grain filling. We examined the algorithms of spikelet fertility response to temperature used in crop models. Results showed that simulation performance is poor for crop yields if the same function was used at different growth stages outside the booting stage. Indeed, we replaced a parameter spikelet fertility algorithm of the ORYZA model and developed the function of estimating grain number per spike and grain weight. After that, the improved equation algorithm was applied to 10 rice growth models. New functions considered the harmful effects of low temperatures on rice yield at different stages. In addition, the threshold temperatures of the cold tolerance were set for different rice varieties. The improved algorithm enhances the model's ability to simulate rice yields under climate change, providing a more reliable tool for adapting rice production to future climatic challenges., (© The Author(s) 2024. Published by Oxford University Press on behalf of the Society for Experimental Biology. All rights reserved. For commercial re-use, please contact reprints@oup.com for reprints and translation rights for reprints. All other permissions can be obtained through our RightsLink service via the Permissions link on the article page on our site—for further information please contact journals.permissions@oup.com.)

Strawberry is a kind fruit with high nutritional value and economic value, but powdery mildew, black mildew, and other diseases decreased the quality of strawberry. In this paper, an enhanced AlexNet (E-AlexNet) was proposed for strawberry quality evaluation. Firstly, strawberry images shot from laboratory and field were collected, then, a series of pre-processing, such as data augmentation and balance were performed on the data set. At last, the images were imported into the E-AlexNet for training. Our improvement is as follows: (1) The convolution kernel's size has been modified; (2) A single convolutional layer is divided into three convolutional layers with different convolution kernels; (3) Using the BN layer and L2 regularization. The network's accuracy and order of magnitude can be improved by the above operation. Finally, the accuracy influence of the neuron volume in the final fully connected layer was discussed. Result shows that the average recognition accuracy of the original AlexNet network with original data set is 84.50%, and the accuracy of E-Alexnet network is 90.70%, after augmentation, the average recognition accuracy of AlexNet is 89.34% and E-Alexnet is 95.75%. The E-AlexNet is superior to the original AlexNet before and after data augmentation. The proposed model is also compared with other classical models, and the experimental results show that the proposed improved method is feasible. The network improvement method proposed in this paper is significant, and the E-AlexNet network has a broad application prospect in strawberry quality evaluation. [ABSTRACT FROM AUTHOR]

As more and more attention is paid to the formation of models of new service development, it is noticed that the compositions of a new model are usually based only on examples of previous models that were presented in the scientific literature. Although the practical application of the proposed models in any organizational condition is often verified by empirical studies, the revisions of the validity of model compositions are still quite fragmented. The lack of such research does not give the possibility for any further improvement of the model or elimination of the revealed shortcomings and it makes the proper preparation for testing the proposed model in a company complicated. In order to evaluate the suitability of the composition of the newly formed model of service development in the conditions of modern service companies, an empirical evaluation was performed by using the methods of abstraction, comparative analysis, synthesis, and expert survey. According to the results of the research, the suggestions for the improvement of the new service development model were made and guidelines for further research were formulated. [ABSTRACT FROM AUTHOR]

The level of spatial autocorrelation (SAC) present in model residuals can have a detrimental influence on statistical inferences in geographical ecology. There has been significant progress in understanding the nature and causes of residual SAC and developing new methods to reduce the SAC. However, ecologists have not yet found a suitable answer as to when the level of residual autocorrelation is likely to magnify and whether its shift through spatial regression (i.e. the difference in the residual SAC produced by non‐spatial and spatial approaches) can be predicted using a set of readily available variables. In this paper, I reanalyzed the outcomes reported by Bini et al. (2009), who originally compared the differences in the standardized regression coefficients resulting from non‐spatial ordinary least squares and various spatially explicit methods. Although these researchers were unable to identify reliable predictors of the coefficient shifts, in the present study, I observed that the level of residual autocorrelation significantly and linearly increased with an increase in the magnitude of the SAC inherently possessed by, especially, response and explanatory variables. Moreover, the amount of shift in the residual SAC varied as a linear function of the response and explanatory variables. These findings imply that, in general, specific conditions exist under which there is an increase in the magnitude of the residual SAC, that such a shift can be predicted by the nature and structure of the data sets involved, and that, after all, the residual SAC varies systematically regardless of the selection of the data type, statistical method and spatial scale. I suggest that the level of the SAC present in the input variables can directly indicate how much improvement (i.e. reduction in the residual SAC) a non‐spatial model will experience when a proper spatial approach is employed. [ABSTRACT FROM AUTHOR]