The food processing industry faces critical challenges related to food waste and food loss due to the deterioration of perishable products derived from agricultural and livestock sources. In 2022, approximately 1.05 billion tons of food were wasted globally, resulting in economic losses of nearly USD 1 trillion. In Indonesia, food waste reached 23–48 million tons annually, causing economic losses of IDR 213–551 trillion and generating approximately 1,702.9 megatons of CO₂ equivalent emissions. These conditions highlight the need for integrated economic and environmental decision-making in food supply chains. This study proposes a novel three-echelon inventory model consisting of a farmer (growing items), a processor, and a retailer (perishable products). The model simultaneously incorporates growing effects, product deterioration, preservation technology investment, carbon emissions under carbon tax regulation, and alternative credit payment schemes. The model aims to maximize the profit for each echelon as well as the entire supply chain. Three main decision variables are considered: the replenishment cycle (T), shipment frequency (n), and preservation technology investment level (ε). The optimal solution is obtained using a numerical iterative procedure, followed by sensitivity analysis to examine the robustness of the results. The results reveal a trade-off between profitability and environmental impact. A credit period longer than the replenishment cycle maximizes profit but leads to the highest emissions, while a shorter credit period significantly reduces emissions (by about 74
This study develops a robust and interpretable framework for modeling the yield trajectories of date palm (Phoenix dactylifera L.) across Saudi Arabia under three climate change scenarios (CMIP6: SSP1-2.6, SSP2-4.5, and SSP5-8.5) projected through 2100. The framework uses neural networks, decision trees, and ensemble machine learning techniques in R, with ten key agroclimatic independent variables. Projected climate diagnostics indicate a significant intensification of heat stress, with mean temperatures rising to 28.7 °C and an increase in tropical nights (> 32 °C) to an average of 34.3 nights per year under the SSP5-8.5 scenario, posing physiological threats to pollination, respiration, and fruit set. Concurrently, the standardized precipitation evapotranspiration index (SPEI12) values indicate increased drought exposure, while rising precipitation extremes contribute to hydrological instability. A regression tree model trained on historical data from 1961 to 2024 indicates that drought thresholds (SPEI12 < − 0.2), minimum temperature, and the frequency of tropical nights consistently influence yield. Comparative feature analyses across decision trees, linear models, and SHAP-based neural models confirm the dominant roles of thermal accumulation and nocturnal warmth, with tree-derived splits providing clarity for field-level interventions. Scenario-based yield projections suggest stability under SSP1-2.6, a gradual decline under SSP2-4.5, and near collapse under SSP5-8.5. These outcomes highlight the need for adjustments to zoning, repositioning of cultivars, and the implementation of adaptive land-use strategies. Embedded within a Climate–Yield Resilience Dashboard, these insights translate climate diagnostics into actionable policy triggers, supporting precision agriculture and stakeholder planning aligned with Saudi Vision 2030. This framework offers an integrative platform for resilience-based crop management, effectively bridging ecophysiological knowledge with scalable adaptation strategies in dryland agroecosystems.
Snowmelt plays a vital role in sustaining river flows and regulating water resources in mountain ecosystems, particularly under changing climatic conditions. This study aims to quantify the contribution of snowmelt to streamflow by estimating snow cover and snow water equivalent (SWE) in the Greater Caucasus Basin of Azerbaijan. MODIS snow-cover and FLDAS land-surface products were analyzed using Random Forest, with data processed on the Google Earth Engine (GEE) platform to support computationally efficient and consistent snow-data processing. On the held-out product-derived test data, the RF models reproduced the dominant variability in SCA and SWE, with RMSE values of 16.3 km² and 2.3 kg m⁻², respectively; these metrics represent product-based agreement rather than an independent improvement in absolute accuracy over MODIS or FLDAS. MODIS and FLDAS products indicated peak snow cover in December 2023 (1,310 km²) and maximum SWE in February 2023 (23 kg m⁻²). During 2013–2023, mean March–June net SWE depletion was 20.48 kg m⁻², equivalent to 0.168 kg m⁻² d⁻¹. Model-attributed snowmelt runoff represented 35.70
The present work explores the dynamic structures for various solitary wave solutions related to wave propagation in shallow water by analyzing the coupled Drinfel’d–Sokolov–Wilson equation. The proposed equation provides important insight for understanding the complex behavior that arise in shallow water wave system and represents an important mathematical framework for describing nonlinear wave interactions. For obtaining solitary wave solutions, the (G'/G,1/G) -expansion method and the extended trial function method are utilized, which yield trigonometric, hyperbolic, rational, and elliptic function solutions. Moreover, obtained solutions are analyzed using 3D surface plots, line profiles and contour plots to demonstrate their propagation behavior and wave characteristics. The developed solutions provide deeper understanding of complex wave dynamics and gives a strong foundation for validating mathematical simulations concerning wave propagation in various nonlinear phenomena. The obtained solitary wave solutions reveal a variety of wave patterns, including dark soliton, kink soliton, dark–bright soliton, bright singular soliton and periodic wave solution. The resulting solitary wave solutions and their corresponding physical phenomena may be helpful in understanding the different kinds of wave propagation arise in plasma physics, nonlinear optics, shallow water system and other scientific fields.
Accurate climate data are essential for sustainable development and climate resilience, yet missing data remain a major challenge in data-sparse regions such as Burkina Faso. This study quantifies the performance gain of hybrid classification–regression machine learning (ML) models for reconstructing missing rainfall time series, relative to raw satellite data imputation. Unlike simple bias correction, the approach combines five satellite products with ML, learning from the diversity among datasets rather than from any single dataset’s bias; a two-step regression–classification scheme forms its final stage. Once trained, the model requires no in situ data, an approach termed the crossed ML approach. The reconstruction uses five satellite products (TAMSAT, RFE, ARC, CHIRP, CHIRPS), geographic coordinates, climatic zone, and month of year, together with in situ observations from Burkina Faso’s national rain gauge network (2000–2020). Regression models (linear regression, decision trees, random forests, boosted trees, multilayer perceptrons) are coupled with classification models for rainfall event detection, yielding 25 crossed ML models compared against 5 uncrossed ML models and 5 raw satellite datasets, evaluated using RMSE, MAE, ME, PBIAS, POD, and FAR. ML models outperform most satellite estimates, particularly in error reduction. The best models reduce the annual median RMSE by about 2.5 relative to most satellite datasets, improve the coefficient of determination by over 0.5, and lower the false alarm rate by more than 0.6. While uncrossed ML models yield higher detection probability, crossed ML models better balance detection and false alarms. These findings highlight ML’s potential to improve rainfall reconstruction and data availability.
