Publications
2025 |
|
|
Lee M, Bayazid AR, Rosenthal L, Khan RH, Arku R, Barratt B, Quayyum Z, Baumgartner J. Spatiotemporal patterns in air pollution and sound in Dhaka, Bangladesh Scientific Reports, 15 , pp. 32852, 2025. @article{M2025,
title = {Spatiotemporal patterns in air pollution and sound in Dhaka, Bangladesh}, author = {Lee M, Bayazid AR, Rosenthal L, Khan RH, Arku R, Barratt B, Quayyum Z, Baumgartner J. }, url = {https://www.nature.com/articles/s41598-025-12815-9}, doi = {10.1038/s41598-025-12815-9}, year = {2025}, date = {2025-09-25}, journal = {Scientific Reports}, volume = {15}, pages = {32852}, abstract = {High air pollution and sound in Dhaka pose major health risks, yet limited high-resolution data hinders epidemiologic assessments and interventions. We measured fine particulate matter (PM₂.₅), black carbon (BC), and sound (dBA) at 70 locations across Dhaka during the 2023 dry (Jan-Mar) and wet (Jul-Sep) seasons, with continuous monitoring at 8 fixed sites and short-term sampling at 62 rotating sites (3 days/season). Over 1.7 million minutes of PM2.5 and 1.2 million minutes of sound were collected, enabling detailed spatial (land use features) and temporal (seasonal, weekly, daily, and diurnal) analyses. Air pollution was four times higher in the dry season, with commercial/industrial areas and transportation corridors having the highest levels, particularly at night. Sound levels varied less temporally, and were highest in transportation corridors, mixed-use areas, and commercial hubs. Air pollution and sound across the city exceeded international guidelines. These findings provide critical evidence to inform targeted policy and interventions.}, keywords = {urban environments}, pubstate = {published}, tppubtype = {article} } High air pollution and sound in Dhaka pose major health risks, yet limited high-resolution data hinders epidemiologic assessments and interventions. We measured fine particulate matter (PM₂.₅), black carbon (BC), and sound (dBA) at 70 locations across Dhaka during the 2023 dry (Jan-Mar) and wet (Jul-Sep) seasons, with continuous monitoring at 8 fixed sites and short-term sampling at 62 rotating sites (3 days/season). Over 1.7 million minutes of PM2.5 and 1.2 million minutes of sound were collected, enabling detailed spatial (land use features) and temporal (seasonal, weekly, daily, and diurnal) analyses. Air pollution was four times higher in the dry season, with commercial/industrial areas and transportation corridors having the highest levels, particularly at night. Sound levels varied less temporally, and were highest in transportation corridors, mixed-use areas, and commercial hubs. Air pollution and sound across the city exceeded international guidelines. These findings provide critical evidence to inform targeted policy and interventions.
|
|
|
Metzler AB, Nathvani R, Sharmanska V, Bai W, Moulds S, Owoo NS, Fynn IEM, Muller E, Dufitimana E, Akara GK, Owusu G, Agyei-Mensah S, Ezzati M. Science of The Total Environment, 988 (179739), 2025. @article{AB2025,
title = {Unsupervised deep clustering of high-resolution satellite imagery reveals phenotypes of urban development in Sub-Saharan Africa}, author = {Metzler AB, Nathvani R, Sharmanska V, Bai W, Moulds S, Owoo NS, Fynn IEM, Muller E, Dufitimana E, Akara GK, Owusu G, Agyei-Mensah S, Ezzati M.}, doi = {10.1016/j.scitotenv.2025.179739}, year = {2025}, date = {2025-08-01}, journal = {Science of The Total Environment}, volume = {988}, number = {179739}, abstract = {Sub-Saharan Africa and other developing regions have urbanized extensively, leading to complex urban features with varying presence and types of roads, buildings and vegetation. We use a novel hierarchical deep learning framework and high-resolution satellite images to characterize multidimensional urban environments in multiple cities. Application of the model to images from Accra, Dakar, and Dar es Salaam identified areas with analogous patterns of building density, roads and vegetation. These included dense settlements within the metropolitan boundary (20\textendash54% of urban area), peri-urban intermix of natural and built environment (21\textendash44%), natural vegetation (9\textendash13%) and agricultural land (8\textendash15%). Kigali, with its mountainous geography and post-colonial expansion, exhibited unique urban characteristics including a sparser urban core (23%) and significant wildland-urban intermix (19% of vegetation). Other notable clusters were water (2% of area of Accra) and empty land (8\textendash10% of Accra and Dakar). Our results demonstrate that unlabeled satellite images with unsupervised deep learning can be used for consistent and coherent near-real-time urban monitoring, particularly in regions where traditional data are scarce.}, keywords = {urban environments}, pubstate = {published}, tppubtype = {article} } Sub-Saharan Africa and other developing regions have urbanized extensively, leading to complex urban features with varying presence and types of roads, buildings and vegetation. We use a novel hierarchical deep learning framework and high-resolution satellite images to characterize multidimensional urban environments in multiple cities. Application of the model to images from Accra, Dakar, and Dar es Salaam identified areas with analogous patterns of building density, roads and vegetation. These included dense settlements within the metropolitan boundary (20–54% of urban area), peri-urban intermix of natural and built environment (21–44%), natural vegetation (9–13%) and agricultural land (8–15%). Kigali, with its mountainous geography and post-colonial expansion, exhibited unique urban characteristics including a sparser urban core (23%) and significant wildland-urban intermix (19% of vegetation). Other notable clusters were water (2% of area of Accra) and empty land (8–10% of Accra and Dakar). Our results demonstrate that unlabeled satellite images with unsupervised deep learning can be used for consistent and coherent near-real-time urban monitoring, particularly in regions where traditional data are scarce.
|
|
|
Clark SN, Arku RE, Ezzati M, Bennett J, Nathvani R, Alli AS, Nimo J, Moses JB, Baah S, Hughes A, Agyei-Mensah S, Owusu G, Toledano M, Brauer M. Scientific Reports, (15), pp. 21403, 2025. @article{SN2025,
title = {Moving beyond the noise: geospatial modelling of urban sound environments in a sub-Saharan African city}, author = {Clark SN, Arku RE, Ezzati M, Bennett J, Nathvani R, Alli AS, Nimo J, Moses JB, Baah S, Hughes A, Agyei-Mensah S, Owusu G, Toledano M, Brauer M. }, doi = {10.1038/s41598-025-06537-1}, year = {2025}, date = {2025-07-01}, journal = {Scientific Reports}, number = {15}, pages = {21403}, abstract = {Cities encompass a mixture of artificial, human, animal, and nature-based sounds, which through long and short-term exposures, can impact on physical and mental health. Yet, most epidemiological research has focused on only transportation noise, leaving a significant gap in understanding the health impacts of other urban sound types, especially in sub-Saharan Africa (SSA). We conducted a large-scale measurement campaign in Accra, Ghana, collecting audio recordings and sound levels from 129 locations between April 2019-June 2020. We classified sound types with a neural network model and then used Random Forest land use regression to predict prevalences of different sound types citywide. We then developed a composite metric integrating sound levels with the prevalence of sound types. Road traffic sounds dominated the urban core, while human and animal sounds were prominent in high-density and peri-urban areas, respectively. Our high-resolution approach provides a comprehensive characterization of the complexity of urban sounds in a major SSA city, paving the way for new epidemiological studies on the health impacts of exposure to diverse sound sources in the future.}, keywords = {urban environments}, pubstate = {published}, tppubtype = {article} } Cities encompass a mixture of artificial, human, animal, and nature-based sounds, which through long and short-term exposures, can impact on physical and mental health. Yet, most epidemiological research has focused on only transportation noise, leaving a significant gap in understanding the health impacts of other urban sound types, especially in sub-Saharan Africa (SSA). We conducted a large-scale measurement campaign in Accra, Ghana, collecting audio recordings and sound levels from 129 locations between April 2019-June 2020. We classified sound types with a neural network model and then used Random Forest land use regression to predict prevalences of different sound types citywide. We then developed a composite metric integrating sound levels with the prevalence of sound types. Road traffic sounds dominated the urban core, while human and animal sounds were prominent in high-density and peri-urban areas, respectively. Our high-resolution approach provides a comprehensive characterization of the complexity of urban sounds in a major SSA city, paving the way for new epidemiological studies on the health impacts of exposure to diverse sound sources in the future.
|
|