We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Two stage inversion method for microplastics emission estimation
AI summary Read the abstract
This study applied a two-stage inversion algorithm to estimate the spatial and temporal emission patterns of microplastics and microfibers across the western United States, using deposition measurements and atmospheric transport modeling. The approach successfully reconstructed emission source patterns from sparse measurement data. This kind of inverse modeling is useful for identifying major microplastic emission hotspots from limited monitoring networks.
The goal of this contribution is to explore two-stage inversion algorithm for spatio-temporal emission estimation (2D and time) from deposition measurements of microplastics and microfibers from Western USA. We consider the linear inversion model formulated as y = M x , where y is the measurement vector, M is source-receptor-sensitivity matrix computed using Lagrangian particle dispersion model FLEXPART, and x is the unknown emission vector from given spatial element. The inverse problem is typically ill-conditioned due to the measurements sparsity, hence, we propose two stage algorithm for inversion of this type. First, we run the inversion algorithm for the whole spatial domain, hence, we obtain averaged emission from each spatial element of the considered spatial domain. Second, we use the estimated emission from the first step (common for all spatial elements) as a prior emission in the second step where the inversion problem is considered for each spatial element separately. We demonstrate that this approach regularizes the inversion problem of spatio-temporal emission from sparse measurements, concretely on microplastics and microfibers emission estimation in Western USA.
More Papers Like This
Spatio-temporal inversion of atmospheric microplastics emmisions using block-coordinate descent method
AI summary Read the abstract
Researchers developed an inversion modeling method to estimate where airborne microplastics originate based on deposition measurements collected across the western United States. The spatial optimization approach identified likely emission hotspots over broad geographic areas. Better source attribution of atmospheric microplastics can guide pollution reduction strategies for wind-transported plastic particles.
Atmospheric microplastics emissions estimation and uncertainty quantification using Gibbs sampler
AI summary Read the abstract
Researchers developed a Bayesian inverse modeling approach using a Gibbs sampler to estimate spatiotemporal atmospheric microplastic emissions from concentration measurements, finding that incorporating agricultural, road dust, and ocean emission patterns as regularization priors substantially reduces uncertainty in emission estimates.
Sources and fate of atmospheric microplastics revealed from inverse and dispersion modelling: From global emissions to deposition
AI summary Read the abstract
Researchers combined atmospheric observations and inverse modeling to estimate global microplastic emissions at 9.6 megatons per year, then used dispersion modeling to trace sources and deposition patterns from emissions to atmospheric fallout worldwide.
Global Emissions of Atmospheric Microplastics Revealed From Inverse Modelling
AI summary Read the abstract
This study used inverse modeling to estimate global atmospheric emissions of microplastics, finding that airborne transport is a significant pathway in the global plastic cycle. Results suggest substantial emissions contribute to long-range atmospheric transport of microplastic particles.
Estimation of spatio-temporal source of microplastics using Bayesian Neural networks
AI summary Read the abstract
Researchers applied a spatial Bayesian neural network to solve the ill-posed inverse problem of estimating spatio-temporal sources of airborne microplastics, combining Western USA deposition measurements with the FLEXPART dispersion model to improve source estimation compared to conventional Bayesian methods.
Research digests by email
When a large batch of papers lands in the Atlas, we read through it and send a short write-up of what stood out.