This week’s digest covers environmental sustainability, marine biology, and AI impacts, featuring both African-led and global studies.
📊 This week at a glance
🌍 African-led research
Offshore wind sites in Vietnam need multi-variable screening, not just wind maps.
Using 32 years of data, researchers screened four sites for wind, waves, and water depth, finding that wave and depth conditions vary significantly and must be considered for viable wind farm placement. This changes the approach from wind-only resource assessment to integrated metocean screening, which is crucial for planning offshore wind in African coastal regions with similar data needs.
A hybrid AI model predicts bridge scour more accurately than traditional equations.
The study combined data augmentation with a genetic algorithm-optimized neural network to predict local scour (erosion around bridge piers), outperforming conventional empirical formulas. This offers a more reliable tool for designing safe bridge foundations in data-scarce regions, which is vital for infrastructure resilience in Africa.
Drought and war compounded food and fuel insecurity in Nairobi’s informal settlements.
In 2022, a severe drought and the Russian-Ukrainian war led to reduced food security and cleaner cooking fuel use among 701 households using pay-as-you-go LPG. This highlights how overlapping crises can push vulnerable populations to rely on dirtier fuels, informing policies for energy and food resilience in African cities.
Macroplastics in South African rivers serve as artificial habitats for aquatic insects.
In four headwater streams, macroinvertebrates colonized plastic substrates similarly to natural ones, indicating that plastics can alter river ecosystems by providing new surfaces for life. This changes our understanding of plastic pollution’s ecological role in Afrotropical rivers, with implications for river management and biodiversity conservation.
A parasitic isopod is linked to poorer health in Caribbean damselfish.
Fish infected with the isopod Anilocra chromis showed reduced body condition and other health metrics compared to uninfected fish. This provides evidence of the parasite’s negative impact on host fitness, which is important for understanding fish population dynamics in tropical marine environments.
Machine learning models can optimize waste-to-energy recovery on university campuses.
This study (abstract limited) applies machine learning to predict energy recovery from municipal solid waste, offering a data-driven approach for sustainable waste management in institutional settings. This could help African universities reduce waste and generate energy, though details are limited.
🔬 Global breakthroughs
Shipping emissions mapped in space and time to guide low-carbon pathways.
Using multi-source data, researchers quantified ship carbon emissions and evaluated strategies for reducing them, providing a detailed spatiotemporal picture. This supports targeted emission reduction policies in the maritime industry, which is relevant for African ports and shipping routes.
A new accounting method integrates carbon and biodiversity footprints for organizations.
The study develops a metric that combines an organization’s carbon and biodiversity impacts, enabling more holistic environmental accounting. This changes how companies can measure their ecological footprint, offering a tool for African businesses to align with global sustainability goals.
Dire wolves were a distinct lineage, not close relatives of gray wolves.
Paleogenomic analysis of two dire wolf specimens shows they diverged from other canids millions of years ago, with limited interbreeding with gray wolves. This revises the evolutionary history of these extinct predators, providing insights into canid evolution that can inform conservation of modern species.
Training AI models has environmental impacts beyond carbon emissions.
A life cycle assessment of training on Nvidia A100 GPUs found significant impacts in water use, mineral depletion, and toxicity, not just carbon. This broadens the scope of AI sustainability assessments, urging African institutions to consider full environmental costs when adopting AI technologies.
All papers are open access. Explore more Environmental Science research on FRELIP · discover open scholarship at frelip.org and search 36,000+ open works at search.frelip.org. FRELIP — born in Nigeria, built for African scholarship, serving the world.