Volume 16, Issue 5, December 2025
Bioethanol Production from Brewer’s Spent Grain, Cassava, and Yam Peels: A Review
Onuora Okorie
Department of Chemical Engineering, Enugu State University of Science and Technology, Enugu, Nigeria
Abstract- The increasing global energy demand and the environmental burden of fossil fuels have intensified interest in renewable alternatives such as bioethanol. This study reviews the conversion of agro-industrial residues specifically yam peels, cassava peels, and brewer’s spent grain (BSG) into bioethanol, emphasizing yield variations, process optimization, and sustainability implications. Differences in ethanol yields reported across studies are attributed to the diverse proximate and chemical compositions of the feedstocks, as well as to variations in pretreatment and pre-analysis processing. Pretreatment plays a vital role in exposing cellulose and hemicellulose, enabling efficient hydrolysis and fermentation. High ethanol yields are strongly linked to feedstocks with elevated starch content, low protein, and low dry matter. Advances in experimental design and process optimization are critical for improving conversion efficiency, reducing production costs, and enhancing the economic feasibility of commercial-scale applications. Beyond energy generation, valorizing agricultural wastes contributes to circular economy strategies by transforming low-value byproducts into high-value energy resources, while simultaneously addressing waste management, energy security, and climate change mitigation. This review highlights the challenges and opportunities in lignocellulosic biomass valorization, underscoring the need for continued technological innovation to achieve sustainable biofuel production.
Keywords- Bioethanol, Biomass, Pretreatment Methods, Hydrolysis, Fermentation and Bioprocess Modeling
Download full paper PDF format (Page: 1-12)
An Improved Predictive Model for Assessing the Impact of Deforestation qnd CO₂ Emissions on Flood Hazards in Pakistan
Muhammad Hasnain Abbas Khan
Department of Data Science, University of Engineering & Technology (UET), Lahore, Punjab, Pakistan
Abstract- Climate change, deforestation, and rising CO₂ emissions have significantly increased the frequency and severity of floods in Pakistan. Traditional forecasting models often fail to account for critical environmental variables such as land-use changes, forest degradation, and atmospheric carbon concentrations. This study introduces a novel multi-dimensional machine learning (ML) framework that systematically integrates environmental degradation indicators, deforestation rates, CO₂ emissions, and land use changes, with traditional hydrological and meteorological variables for flood prediction in Pakistan. Addressing a critical methodological gap in regional flood forecasting, it presents the first comprehensive integration of these factors using advanced ensemble ML techniques. A diverse dataset comprising 5,500 records from 2010–2023, sourced from national and international repositories, was compiled and preprocessed through median imputation, StandardScaler normalization, feature selection, and SMOTE balancing. The proposed hybrid boosted ML model and stacking-based ensemble framework demonstrate significant technical innovation and impact, achieving 95.73% accuracy, F1-scores exceeding 0.97, ROC-AUC of 0.982, and PR-AUC of 0.98. Through systematic evaluation of twelve algorithms, this framework establishes new benchmarking standards in flood prediction research. Feature importance analysis reveals CO₂ emissions (correlation: 0.79), rainfall (0.78), and deforestation rates (0.74) as dominant predictors, quantitatively confirming the critical influence of environmental degradation on flood hazards—a key insight previously unaddressed in regional studies. This research advances both scientific understanding and practical disaster management by enhancing model interpretability and delivering actionable insights for evidence-based early warning systems and policy recommendations. It directly addresses challenges such as data imbalance, feature uncertainty, and multi-source integration, offering a robust and scalable solution for flood risk prediction. The methodology further provides a foundation for future work, including geographic expansion, real-time environmental data incorporation, advanced ensemble learning, and mobile application development, contributing toward climate-resilient disaster management at both national and global scales.
Keywords- Flood Prediction, Machine Learning, Climate Change, Deforestation, CO₂ Emissions, Rainfall, Environmental Indicators, Stacking Classifier, Early Warning Systems, Flood Risk Management and Disaster Preparedness
Download full paper PDF format (Page: 13-22)
Approaches to Elevation Measurement
Nagi Zomrawi Mohammed, Moawia Alameen
Al-Baha University, Al-Baha, KSA, Sudan University of Science and Technology, Khartoum, Sudan
Omdurman Islamic University, Omdurman, Sudan
Abstract- The conventional definition of levelling is the process of rolling to eliminate height differences. Thus, when the height of one point is known, the heights of the other points can be calculated. Based on the height difference technique or method adopted, levelling can be classified into many types, such as ordinary levelling, precise, trigonometric, barometric, tacheometric, GPS, etc. To determine the reduced level or height of points, it is necessary to define a datum upon which the heights of points are referred. The most important reference datums are the Geoid, which can be represented by the mean sea level and the Ellipsoid, which defines the mathematical surface of the Earth. The mean sea level that physically represents the base of orthometric heights is rising due to climate change. This rise is caused primarily by global warming, which adds water from melting land-based ice sheets and glaciers and the expansion of seawater as it warms. This future continuous rise leads to changes in ground-reduced levels based on the mean sea level. This research tried to discuss methods of height determination that can be in direct contact with objects or at a distance. In addition to the reference datums adopted, which may often be either the geoid or the ellipsoid. The work concluded that geodetic heights based on the ellipsoid as a reference datum could recently be easily determined through GPS, and they may be transferred to orthometric heights since accurate measurements for a geoidal model are available. Direct methods for height determination are suitable for a limited number of points, whereas indirect ones could be used to collect data for large areas. The accuracy of the latter depends largely on the resolution and the number of ground control points used. Since the mean sea level is continuing to change, it is not coincidence with the geoid, where the world geodetic datum (WGS84) ellipsoid is now the dominant geodetic model. Digital level is now replacing the optical one, where trigonometric levelling can be carried out directly with a total station.
Keywords-Barometric Heighting, Ellipsoidal Height, Geodetic Height, Global Positioning System (GPS), Levelling, Orthometric Height, Photogrammetry, Remote Sensing, Side Airborne Looking Radar (SALR) and Triangulation
Download full paper PDF format (Page: 23-30)
Evaluating Spatio-Temporal Changes of Flood Risk Area in the Case of Chemoga Watershed, Amhara Region, Ethiopia
Mekoyet Fekadu, Hamere Yohannes, Beka Adugna Jima
Department of Civil and Environmental Engineering, Master of Science in Geodesy and Geomatics Program (Specialization in Geomatics), Addis Ababa Institute of Technology (AAiT), Addis Ababa University
Agricultural Machinery Engineering program, Department of Natural Resource Management, College of Agriculture and Natural Resource Management at Salale University, Fitche, Ethiopia
Abstract- Flooding is a frequent natural disaster, often caused by rivers exceeding their storage capacity and inundating nearby low-lying areas. This study analyzed flood risk using a 30×30 m digital elevation model (DEM), along with land use/land cover (LULC), precipitation, and soil data. LULC changes were assessed for 1993, 2003, 2013, and 2023 using a semi-automatic plugin in Quantum GIS, with flood risk trends analyzed through Multi-Criteria Decision-Making Analysis. In 1993, agricultural land (41.56%) and water bodies (30.51%) were the most dominant. By 2003, settlements (29.52%) and forests (23.53%) were highest. In 2013, bare land (40.99%) and agricultural land (26.86%) were prevalent. In 2023, agricultural land accounted for 54.14%, with settlements and bare land each at approximately 16.5%. Classification accuracy ranged from 91.4% to 93.48% across the years. High flood risk areas increased over time: 66.56% (1993), 77.29% (2003), 77.09% (2013), and 79.37% (2023). Conversely, low-risk areas declined from 1.83% in 1993 to just 0.38% in 2023, highlighting a growing vulnerability to flooding.
Keywords- Chemoga Watershed, Flood Risk Area and Land Use Land Cover Change
Download full paper PDF format (Page: 31-45)
