RMSE Values Research Articles

MSDFNet: multi-scale detail feature fusion encoder–decoder network for self-supervised monocular thermal image depth estimation

Abstract Currently available thermal image depth estimation methods are difficult to efficiently extract fine multi-scale feature information from thermal images and suffer from the problem of blurring details at the edges of the estimated depth map. To address these challenges, this paper proposes MSDFNet, a multi-scale detail feature fusion encoder–decoder network, for self-supervised monocular thermal image depth estimation. The model is based on a channel expansion hourglass residual lightweight feature encoder, which can capture rich and fine-grained multi-scale feature information with low computational effort. MSDFNet utilizes a detail feature weight evaluation decoder to fuse cross-scale features and reevaluate the importance of each feature, thereby emphasizing critical edge information at multiple scales. Additionally, MSDFNet incorporates a depth consistency loss function, which provides self-supervised signals for the detailed features of thermal images and improves the optimization of network performance. The method is applied to the ViViD++ and MS2 datasets and achieves state-of-the-art depth estimation performance compared to existing state-of-the-art algorithms. In the Indoor Dark scenario of the ViViD++ dataset, the Abs Rel, Sq Rel, RMSE, and RMSE log error metric values of MSDFNet are reduced by 6.71%, 11.92%, 9.09%, and 5.73%, respectively, while the accuracy metric values δ < 1.25 i , i = 1 , 2 , 3 were improved by 4.18%, 1.13%, and 0.2%, respectively. In addition, MSDFNet proves its excellent generalization ability on the MS2 dataset. The Abs Rel and RMSE error values in the night scene are reduced by 45.6% and 30.09%, respectively, and the accuracy δ < 1.25 i , i = 1 , 3 is improved by 20.95% and 1.33%, respectively. The Abs Rel and RMSE values in the rainy day scenario are reduced by 1.33% and 1.21%, respectively, and the accuracy δ < 1.25 i , i = 1 , 3 is improved by 0.24% and 0.83%, respectively.

A Hybrid CNN-LSTM Model for Predicting Energy Consumption and Production Across Multiple Energy Sources

This study utilizes a robust dataset provided by Energy Exchange Istanbul (EXIST), a leading authority in energy data, which contains hourly energy consumption and production data from 01/01/2018 to 31/12/2023 across Turkey. Various machine learning and deep learning methods such as linear regression (LR), random forest (RF), support vector machines (SVR), convolutional neural networks (CNN), long short-term memory networks (LSTM), and the proposed hybrid CNN-LSTM model are applied to predict energy consumption and production more accurately. This study transforms time series data into a regression problem using the sliding window method. The experimental results show that the hybrid CNN-LSTM model outperforms the other models in forecasting total energy consumption and natural gas, hydro dam, lignite, hydro river, wind, and fuel oil production. The CNN-LSTM model achieved the lowest RMSE and MAE values and the highest R² scores. The success of the proposed hybrid approach is due to the combination of CNN's ability to identify local patterns and LSTM's ability to learn long-term dependencies. This study demonstrates the hybrid CNN-LSTM model's effectiveness in accurately forecasting energy consumption and production. It makes an important contribution to more efficient use of energy resources.

Simulating the Vegetation Gross Primary Productivity by the Biome-BGC Model in the Yellow River Basin of China

In terrestrial ecosystems, the quantification of carbon absorption is primarily represented by the gross primary productivity (GPP), which signifies the initial substances and energy acquired by the ecosystem. The GPP also serves as the foundation for the carbon cycle within the entire terrestrial ecosystem. The Biome-BGC model is a widely used biogeochemical process model for simulating the stocks and fluxes of water, carbon, and nitrogen between ecosystems and the atmosphere. However, it is the abundance of eco-physiological parameters that lead to challenges in calibrating the model. The parameter optimization method of coupling the differential evolution algorithm (DE) with the Biome-BGC model was used to calibrate and validate the eco-physiological parameters of the seven typical vegetation types in the Yellow River Basin (YRB). And then we used the calibrated parameters to simulate the GPP by way of grid-based simulation. Finally, we conducted model adaptability testing and spatiotemporal analysis of GPP variations in the YRB. The results of the validation (R2, RMSE) were: temperate grasses (0.94, 24.33 g C m−2), alpine meadows (0.94, 18.13 g C m−2), shrubs (0.94, 29.20 g C m−2), evergreen needle leaf forests (0.96, 27.88 g C m−2), deciduous broad leaf forests (0.94, 32.09 g C m−2), one crop a year (0.96, 16.19 g C m−2), and two crops a year (0.90, 38.15 g C m−2). After adaptability testing, the average R2 value between the simulated GPP values and the GPP product values in the YRB was 0.85, and the average RMSE value was as low as 50.92 g C m−2. Overall, the model exhibited strong simulation accuracy. Therefore, after calibrating the model with the DE algorithm, the Biome-BGC model could effectively adapt to the ecologically complex YRB. Moreover, it was able to accurately estimate the GPP, which establishes a foundation for analyzing the spatiotemporal trends of the GPP in the YRB. This study provides a reference for optimizing Biome-BGC model parameters and simulating diverse vegetation types on a large scale.

Theoretical and Experimental Investigation of Thermal Conductivity of Unsaturated Soils Amended with a Sustainable Biochar

This study investigates the thermal conductivity of unsaturated kaolin soil amended with biochar to promote sustainable geotechnical engineering. Biochar from agricultural waste offers the dual benefits of carbon sequestration and sustainable waste management. Experimental measurements were conducted for kaolin soil with 0% (control) and 10% biochar under varying moisture contents. Peach pit biochar increased thermal conductivity by 2–3% at 30–40% saturation and 40–50% at higher saturation as compared to the bare soil. Reed biochar decreased thermal conductivity by 1–2% at lower saturation but increased it by 55–60% at higher saturation. Applewood biochar increased thermal conductivity by 35–50% at moderate saturation, decreased beyond 50% water content, and had minimal variation at lower saturation. Further, the existing empirical models (such as Kersten and the Johansen model, Wiener’s model, and Mickley’s model) for predicting the thermal conductivity of materials were validated using the measured results of biochar-amended soils. Adding 10% biochar reduces thermal conductivity by 34.8%, and the Haigh model (2012) fits best with high accuracy and lower RMSE values than models such as Kersten and Johansen, which appears to be less reliable in case of biochar-amended soils. With an addition of biochar, the R2 values of the models decreased from a range of 0.8 to 0.9 to a range of 0.4–0.6, indicating the need for better model adaptation. Wiener bounds accurately predicted thermal conductivity at low saturation levels but varied greatly at higher ones. The most variable sample was peach pit biochar, highlighting the need to refine predictive models for material-specific differences. These findings provide a foundation for developing improved predictive models and integrating biochar into sustainable geotechnical and geothermal systems.

Application of Mont Carlo Simulation and Artificial Neural Network Model to Probabilistic Health Risk Assessment in Fluoride-Endemic Areas

Groundwater contamination with fluoride is a considerable public health concern that affects millions of people worldwide. The rapid growth of urbanization has led to increase in groundwater contamination. The health risk assessment focuses on both acute and chronic health consequences as it investigates the extent and effects of fluoride exposure through contaminated groundwater. Fluoride exposure, especially in endemic locations, has serious health consequences, including dental and skeletal fluorosis. An accurate assessment of these hazards is essential for public health planning and mitigation actions. The present study uses Monte Carlo Simulation (MCS) and an Artificial Neural Network (ANN) model to perform a Probabilistic Health Risk Assessment on populations in fluoride-endemic areas. Analysis of the results of the study reveals that the concentration of fluoride ranged from 0.58 to 3.80 mg/L with an average of 2.30 mg/L across the Kasganj district, which was higher than permissible limits given by BIS and WHO. The highest value of hazard quotient of 3.29 for Children is found to be in the Durga Colony area, while the lowest value of the hazard quotient of 0.31 for adults is found to be in the Nadrai Gate area. The assessment of health risks revealed a high probability of non-carcinogenic disease from the consumption of groundwater containing fluoride. The ANN model has the R2 value of 0.9989 in training and 0.9870 in testing while RMSE value in training and testing was 0.02230 and 0.0267. The findings suggest that before being used, the groundwater in Kasganj, Uttar Pradesh, India, needs to be treated and made drinkable. The results emphasize the critical need for ongoing monitoring, public education initiatives, and implementing feasible mitigating techniques to lower fluoride exposure. The findings show that this hybrid model is excellent at addressing the numerous uncertainties associated with fluoride use, hence improving the reliability of health risk estimates in fluoride-endemic locations. The results offer vital information to help policymakers and local health officials create focused measures to safeguard public health in Kasganj.

Modeling of the intelligent interface of interaction between a pharmacy and a children's clinic

Introduction. Economic barriers, uneven territorial distribution of pharmacy organizations, and insufficient efficiency of drugs supply chains hinder the accessibility of drugs and the timely receipt of pharmaceutical care (PC) by patients worldwide. The digital transformation in healthcare, accelerated by the COVID-19 crisis, necessitates the adaptation of PC practices to meet modern patient needs. Objective of the study. To develop and validate a procedure for the interaction between a clinic and a pharmacy to facilitate data exchange through an interface, enabling the pharmacy to receive data for forming an assortment of drugs adapted to the characteristics and behavioral trends of the main target audience. Material and methods. The study utilized anonymized data on drugs prescriptions by physicians from the medical information system of a network of medical organizations in Moscow for the period from January 2018 to December 2023. Data preprocessing was conducted, followed by the training of a machine learning model using the LightGBM algorithm. The predictive performance of the model was assessed using MAE and RMSE metrics. Results. An analytical interface for the interaction between the clinic and the pharmacy was developed, incorporating a predictive model for forming the drugs assortment. The model effectively accounts for seasonal trends, patient demographic characteristics, and other key factors influencing drugs demand. The average MAE and RMSE values were 1.27 and 1.68, respectively, indicating high model accuracy. Conclusion. Implementing the developed interface allows the pharmacy to form drugs assortment tailored to the real needs of patients, contributing to optimized inventory management, reduced risk of shortages and overstocking, enhanced accessibility of PC for children, and increased economic efficiency of the pharmacy. The integration of big data technologies and machine learning opens new prospects for the personalization of medical and pharmaceutical care.

Comparative Time Series Analysis of SARIMA, LSTM, and GRU Models for Global SF6 Emission Management System

To maintain the balance of the atmosphere, the amount of change in greenhouse gas emissions must be under control. In order to create a management system and take forward-looking steps in this regard, there is no concrete data other than prediction models today. The success of prediction methods is better understood by comparing multiple methods. This research estimates the changes in the emissions of Sulfur Hexafluoride gas (SF6) in the atmosphere using Seasonal Autoregressive Integrated Moving Average Model (SARIMA), Long-Short Term Memory Neural Network (LSTM) and Gated Recurrent Unit (GRU) forecast models and compares their accuracies. Focusing on monthly SF6 emission values between 1998; 2023, time series analysis was performed to predict future emission figures. The actual values and forecast results were compared and evaluated with performance criteria such as R2, RMSE, NSE, MAE and MAPE%. The findings of this research highlight a continious upward trend in SF6 emissions and project that emission levels could approximately double from current levels by 2050. During the analysis process, all three methods performed well in estimating global SF6 gas emissions. The LSTM model generally outperformed SARIMA and GRU models, having the lowest MAPE (0.003%), MAE (0.0003), RMSE (0.0003), and R2 (1) values. It also exhibited very high predictive success with an NSE value of 0.9991. Therefore, it was determined to be the most suitable estimation method with the least error. The aim of this study is to contribute scientifically to the reduction strategies of SF6 emissions.

From data to dynamics: Reconstructing soliton collision phenomena in optical fibers using a convolutional autoencoder

In this study, a convolutional autoencoder is constructed to extract and reconstruct the dynamical processes of soliton collisions in optical fibers. The model demonstrates exceptional reconstruction capabilities, accurately capturing the evolution of optical event horizons and reproducing nonlinear phenomena such as complex frequency conversions and energy exchange processes. The reconstruction results show high consistency with the numerical simulations, with RMSE values of 0.0220 and 0.0174 in the temporal and frequency domains, respectively. Additionally, by adjusting the training parameters of the convolutional autoencoder model, its reconstruction performance for nonlinear dynamic processes was further validated. This method is expected to provide a different perspective for studying nonlinear phenomena in optical fibers while reducing the consumption of computational resources.

Short-term Traffic Flow Prediction：A Method of MEA-LSTM Model Based on Chaotic Characteristics Analysis

To effectively improve the accuracy of short-term traffic flow prediction, an improved Long Short-Term Memory (LSTM) method is proposed using the Mind Evolution Algorithm (MEA). Firstly, to address the issues of abnormal and missing traffic flow data, a Neighborhood Stacked Denoising AutoEncoder (NSDAE) is used for data repair. Then, the maximum Lyapunov exponent is used to determine the chaotic characteristics. Meanwhile, based on Bayesian estimation theory, the features of three-parameter sequences are fused in high-dimensional space using phase space reconstruction technique to obtain reconstructed multi-parameter fused traffic flow data. Finally, by taking advantage of the fact that MEA can divide the data into several subpopulations for optimal search separately, a prediction model based on MEA to improve LSTM is proposed. The results show that compared to the other two traditional data restoration methods, the NSDAE has higher accuracy, with the lowest average values of RMSE, MAE, and MAPE. Through the phase space reconstruction technique, the feature fusion of three parameters of traffic flow is realized in high-dimensional space, which makes up for the insufficiency of a single time-series data that cannot comprehensively levy the characteristics of traffic flow. The MEA-LSTM model outperforms the LSTM model in terms of prediction accuracy, computational efficiency, and generalization ability, and its RMSE, MEA, and MAPE are reduced by 24.3%, 28.9%, and 30.1%, respectively.

Validation of ERA5-Land-based reconstructed air temperature and near-surface ground temperature on James Ross Island

ABSTRACT This study evaluates the accuracy of ERA5-Land reanalysis products by comparing them with continuous data from James Ross Island, Antarctic Peninsula. ERA5-Land shows lower variability in air and ground temperatures than measurements from three automatic weather stations (AWSs) on James Ross Island: Johann Gregor Mendel (AWS-JGM), Abernethy Flats (AWS-AF), and Johnson Mesa (AWS-JM). Differences in mean annual air temperature (MAAT) and mean annual ground temperature at 5 cm depth (MAGT5) are noted between ERA5-Land and AWSs for the periods 2007/08-2020/21 and 2011/12-2020/21. Strong correlations (0.88–0.92) exist between ERA5-Land and AWSs, with the highest coefficients and lowest RMSE values for AWS-JM (2.7 °C for AT and 4.9 °C for GT5). Validation statistics reveal seasonal variations in ERA5-Land temperature bias with larger discrepancies in summer and low differences in winter. ERA5-Land’s snow depth estimation indicates unrealistic permanent snow cover between 3–24 meters, affecting ground temperature bias and underestimation. Despite its benefits, ERA5-Land’s inaccuracies in ground temperature data limit its direct use for permafrost research. Corrections were applied to enhance the use of ERA5-Land data for detailed permafrost studies in the JRI area.

Entropy-informed multi-stage sampling design for soil pollution mapping in heavy metal contaminated areas

The accuracy of soil heavy metal pollution mapping is heavily reliant on the sampling strategies utilized in both the preliminary and detailed survey stages of site investigations. This study introduces an entropy-informed multi-stage sampling design (EIMSD) method that leverages preliminary survey data as background information and utilizes relative entropy to progressively select sampling points in detailed surveys. Results indicate that the EIMSD method outperforms the grid sampling design (GSD) and conventional sampling design (CSD) methods across both hypothetical and real-world study areas. This superiority is evidenced by a notable rise in R2 values, ranging from 6.4% to 60.4% and a decrease in RMSE values from 16.7% to 54.0% relative to GSD, and a similar trend with an increase in R2 values from 6.7% to 44.1% and a reduction in RMSE values from 16.5% to 39.7% when compared to CSD. This study also investigates the optimal configurations for EIMSD, focusing on the number of detailed sampling points per stage (Nadd) and the ratio of preliminary-to-detailed survey sample sizes (Np/Nd) when the sum of Np and Nd is held constant. Our findings highlight that adding one detailed sampling point per stage (Nadd = 1) is the most effective. For areas with strong spatial variability, a larger Np/Nd value of approximately 3/2 is recommended, whereas a ratio close to 1 is apt for areas with moderate variability. Conversely, for areas with weak variability, a smaller Np/Nd value of about 2/3 is advised. EIMSD provides a more detailed and accurate map of soil heavy metal contamination, facilitating more targeted and effective remediation strategies.

EKLT: Kolmogorov-Arnold attention-driven LSTM with Transformer model for river water level prediction

Water level prediction is crucial for water resource management and flood warnings. Hybrid LSTM-based methods are widely used but face the following challenges: (1) The attention mechanism based on the universal approximation theorem (UAT) is the mainstream method for improving performance, but its essence is infinite approximation of functions, and the accuracy is difficult to improve; (2) LSTM struggles with modeling complex, long-term dependencies. To address these problems, we apply empirical mode decomposition (EMD) and propose input-spatial attention based on the Kolmogorov-Arnold theorem (KAT) for precise water level feature representation. Cascaded LSTM and Transformer structures capture long-term dependencies. Finally, temporal attention is embedded to achieve accurate water level prediction. Experiments show the proposed method achieves RMSE, MAE, MAPE and R2 values of 0.1870, 0.1328, 1.1228 and 0.9540 in the Liaohe River of Liaoning Province, China, and 0.3027, 0.1844, 4.6034 and 0.8659 in the Hunhe River, respectively, indicating its strong predictive performance.

A New Approach to Reduce the Position Data of Moving Objects Using Neural Network

ABSTRACTIn this study, an ANN‐based Global Navigation Satellite System (GNSS) location data reduction approach was proposed. As GNSS location data becomes more common, efficient data reduction techniques are needed to reduce transmission, storage, and processing costs. This involves selecting key points from the original trajectory to maintain integrity, eliminating redundancy, and lowering transmission and storage expenses. In this study, we proposed a new method for reducing GNSS location data in both online and offline settings, utilizing an ANN trained with a mathematically generated dataset. ANN has not been used in data reduction in the literature. The approach involves training the ANN with a window size of 3 and a threshold value of 10°, followed by using the trained model for data reduction. Experimental results show that the ANN achieves a reduction rate of around 59.18% compared to the original trajectory. Notably, the ANN yields a significantly lower RMSE compared to a mathematical method, particularly in areas requiring precision. Despite the slightly slower computation times, the ANN remains suitable for real‐time applications, demonstrating its efficacy for GNSS location data reduction. Our study highlights its online capability, reasonable reduction rates, and low RMSE values, distinguishing it from existing literature. This method shows potential for scenarios where balancing reduction rates with data quality is crucial.

Multi-algorithm fusion-based state of energy assessment of retired lithium-ion batteries

The widespread use of lithium-ion batteries has somewhat alleviated the fossil energy crisis. However, when the state of health (SOH) of a lithium-ion battery drops to 80 %, it usually needs to be taken out of service from its original workplace. Accurately evaluating the state of energy (SOE) of retired batteries can effectively avoid unnecessary losses due to accidental failures of retired lithium-ion batteries in other application scenarios. Therefore, this paper proposes to come to the SOE estimation model of retired lithium batteries based on reinforcement learning Q-learning. Firstly, according to the unstable characteristics of retired Li-ion batteries, Q-learning is used to optimize the weight parameters of three algorithms, namely, back-propagation neural network (BPNN), long and short-term memory network (LSTM) and support vector regression (SVR), to establish the Q-BPNN-LSTM-SVR (QBLS) integration algorithm. Then, considering that the three factors of temperature, voltage, and current affect the state of energy of retired lithium-ion batteries, they are selected as inputs to the QBLS algorithm. Finally, the QBLS algorithm estimates the SOE of individual retired batteries and retired battery packs at different temperatures and operating conditions. The results demonstrate that the method has high precision in estimating the SOE of retired Li-ion batteries, with a Max AE less of than 3 %, and the maximum values of MAE and RMSE do not exceed 0.70 % and 0.80 %, respectively. The method above offers a dependable foundation for using and administering retired lithium-ion batteries.

Machine Learning-Based Predictions of Metal and Non-Metal Elements in Engine Oil Using Electrical Properties

This study investigates the influence of six metallic and non-metallic elements (Fe, Cr, Pb, Cu, Al, Si) on the quality of engine oil under normal, cautious, and critical conditions. To achieve this, the research employs the Design of Experiments (DoE) approach, specifically the Box–Behnken Design (BBD) method, for designing experiments. The electrical properties of 70 engine oil samples prepared under varying conditions were analyzed. Machine learning models, including RBF, ANFIS, MLP, GPR, and SVM, were utilized to predict the concentrations of the six pollutants in the lubricant oil samples based on their electrical characteristics. The models’ performance was assessed using RMSE and R2 indicators during train, test, and All stages. The results revealed that the Radial Basis Function (RBF) model exhibited the best overall performance (RMSE = 0.01, R2 = 0.99). The study proceeds with optimizing RBF model parameters, such as hidden size (best = 17), spread (best = 0.4 or higher), and training algorithm (best = trainlm), to estimate each pollutant individually. The generalizability of the model was assessed by reducing the training data percentage and increasing the testing data percentage. The results demonstrated the model’s proper performance for all pollutants in various training sizes (RMSE = 0.01, R2 = 0.99). However, as the training data ratio reduced to 60:40 and 50:50, the model’s performance in estimating Cu deteriorated, resulting in increased RMSE values (10.76 or 11.85) and decreased R2 values (0.89 or 0.87) across the All step. This academic research hopes to contribute to the field of applied studies, considering the inherent complexities of lubricants and the challenges in measuring small-scale electrical properties.

Evaluation of WSR-88D Level III and MRMS Rainfall Estimates Against Rain Gauge Observations at the Savannah River Site

Rainfall data for the Savannah River Site (SRS) has been historically measured by rain gauges. These instruments serve as ground truth for most climatological and weather applications; however, gauge measurements are prone to errors or biases under certain weather conditions. Rainfall estimates from radar reflectivity values have been developed and improved over the years and serve as an alternative or supplement for gauge measurements. This study compares measurements from tipping bucket rain gauges with rainfall estimates from the NOAA NWS WSR-88D Level III hourly rainfall product and the Multi-Radar Multi-Sensor (MRMS) gauge-corrected precipitation estimate. Results show good agreement between radar derived amounts and ground measurements, with MRMS values showing correlation coefficients of 0.60-0.95 and RMSE values of less than 1.0 cm (0.4 in). The WSR-88D Level III estimates result in correlation coefficients of 0.46-0.86 and RMSE values of less than 1.3 cm (0.5 in). Few outliers are observed for each data pair and are evaluated against precipitation classification products (WSR-88D Hybrid Hydrometeor Classification product and MRMS Precipitation Flag product). The rain gauges used in this study are not part of the Hydrometeorological Automated Data System (HADS) network used to correct the MRMS estimates and therefore, results of this work provide an independent validation of the MRMS gauge correction scheme.

Research on Gas Emission Prediction Based on KPCA-ICSA-SVR

In the context of deep mining, the uncertainty of gas emission levels presents significant safety challenges for mines. This study proposes a gas emission prediction model based on Kernel Principal Component Analysis (KPCA), an Improved Crow Search Algorithm (ICSA) incorporating adaptive neighborhood search, and Support Vector Regression (SVR). Initially, data preprocessing is conducted to ensure a clean and complete dataset. Subsequently, KPCA is applied to reduce dimensionality by extracting key nonlinear features from the gas emission influencing factors, thereby enhancing computational efficiency. The ICSA is then employed to optimize SVR hyperparameters, improving the model’s optimization capabilities and generalization performance, leading to the development of a robust KPCA-ICSA-SVR prediction model. The results indicate that the KPCA-ICSA-SVR model achieves the best performance, with RMSE values of 0.17898 and 0.3071 for the training and testing sets, respectively, demonstrating superior robustness and generalization capability.

Two-step hybrid model for monthly runoff prediction utilizing integrated machine learning algorithms and dual signal decompositions

Runoff is pivotal in water resource management and ecological conservation. Current research predominantly emphasizes enhancing the precision of machine learning-based runoff predictions, with limited focus on their physical interpretability. This study introduces an innovative two-step hybrid runoff prediction framework tailored for the headwater region of the Yellow River Basin (YRB) to improve prediction accuracy and elucidate the runoff modeling process. The framework integrates machine learning techniques with dual signal decomposition approaches, incorporating diverse hydrometeorological and geographic indicators. Long Short-Term Memory (LSTM) and eXtreme Gradient Boosting (XGBoost) algorithms were employed to predict monthly runoff generation in sub-basins delineated by the Soil and Water Assessment Tool (SWAT), which were subsequently integrated using a Recurrent Neural Network (RNN) for monthly runoff concentration prediction. Results indicate that the proposed models delivered superior prediction performance compared to the SWAT model (R2 = 0.86, NSE = 0.85), with the LSTM-based two-step hybrid model (R2 = 0.90, NSE = 0.90) outperforming the XGBoost-based model (R2 = 0.89, NSE = 0.88). The dual decomposition method, integrating seasonal-trend decomposition based on loess (STL) and successive variational mode decomposition (SVMD), demonstrated exceptional efficacy in addressing the complexities of hydrometeorological time series. Models decomposed by STL-SVMD exhibited the highest average R2 and NSE values, as well as the lowest RMSE and MAE values in sub-basin runoff calculations. The low standard deviations of performance metrics further underscored the stability of these models across all sub-basins. This study demonstrates the efficacy of the proposed two-step hybrid model for simulating physical runoff processes in the headwater region of the YRB, providing valuable insights for regional hydrological cycle research and hydro-ecological security.

Multispectral Assessment of Net Radiations Using Comprehensive Multi-Satellite Data

Precise estimation of net radiation (Rn) is fundamental to understanding surface energy balance and is critical for accurately determining crop water requirements, especially using remote sensing and geospatial techniques. The core objective of this study is to evaluate multi-satellite-based net radiations on major cropped areas of the Punjab and Sindh provinces of Pakistan. In this study, overlapping scenes from the Moderate Resolution Imaging Spectroradiometer (MODIS), Landsat 8, and Sentinel 2 were used from 2016 to 2020 along with three temperature products MOD11A1, Landsat 8 (brightness temperature), and ERA5. The multi-satellite-based net radiation estimations on overlapping days were compared with the Global Land Data Assimilation System (GLDAS) dataset. The models based on Landsat 8 and Sentinel 2 data exhibited good performance, with a Nash–Sutcliffe Efficiency (NSE) of 68.9%, a mean error (ME) of 13.918 W/m2, and a bias of 50.669 W/m2. The results indicated that Landsat 8 and Sentinel 2 data produced reliable estimations of net radiation, while MODIS data tended to overestimate due to its higher spatial resolution and broader coverage area. Landsat 8-based estimations are good compared to others, as it has good correlation coefficient and lower RMSE values. The study concludes that Landsat 8 provides the most reliable estimates of net radiation for determining crop water requirements, outperforming other datasets in accuracy. The findings underscore the importance of using high-resolution multi-satellite data for precise agricultural water management, recommending its use in future studies and water resource planning in Pakistan.

Short-term wind power prediction based on IBOA-AdaBoost-RVM

This study introduces an innovative model, namely IBOA-AdaBoost-RVM, which leverages the Improved Butterfly Optimization Algorithm (IBOA), Adaptive Boosting (AdaBoost), and Relevance Vector Machine (RVM). This model is used to solve the problem of low precision of wind power prediction. Initially, normalization is applied to reduce the influence of varying data dimensions. Subsequently, input variables are determined through the Pearson correlation method. Lastly, the efficacy of the introduced model is assessed across four distinct seasonal monthly data sets. The observed outcomes indicate that the proposed model outperforms other models in terms of evaluation metrics, with the average R2, RMSE, MAE, and MAPE values across the four datasets being 0.954, 10.403, 7.032, and 0.645, respectively, show that the proposed method has potential in the field of wind power prediction.