pith. sign in

arxiv: 2604.23003 · v1 · submitted 2026-04-24 · 💻 cs.LG · cs.NE

Collocation-based Robust Physics Informed Neural Networks for time-dependent simulations of pollution propagation under thermal inversion conditions on Spitsbergen

Leszek Siwik , Maciej Sikora , Natalia Leszczy\'nska , Tomasz Maciej Ciesielski , Eirik Valseth , Manuela Bastidas Olivares , Marcin {\L}o\'s , Tomasz S{\l}u\.zalec
show 2 more authors
Jacek Leszczy\'nski Maciej Paszy\'nski
This is my paper

Pith reviewed 2026-05-08 12:18 UTC · model grok-4.3

classification 💻 cs.LG cs.NE
keywords physics-informed neural networksadvection-diffusion equationthermal inversionpollution propagationcollocation methodvariational formulationparticulate matterArctic air quality
0
0 comments X

The pith

A robust physics-informed neural network uses a variational loss tied directly to approximation error and collocation points to simulate time-dependent pollution spread from moving sources under thermal inversion.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a Physics-Informed Neural Network for modeling how pollution from moving emission sources like snowmobiles spreads over time through advection and diffusion. It first builds a stable mathematical setup for the time-dependent advection-diffusion equations by proving boundedness and inf-sup stability of the discrete weak form. This stability lets the training loss measure closeness to the unknown exact solution. A collocation strategy then speeds up the process, making the network practical for real cases. When applied to sensor data from Longyearbyen on Spitsbergen, the model shows thermal inversions trap dense air near the ground and raise particulate matter levels, worsening air quality.

Core claim

By establishing boundedness and inf-sup stability for the discrete weak formulation of the time-dependent advection-diffusion problem, the authors construct a loss function for physics-informed neural networks that is directly related to the true approximation error. Combined with a collocation-based training strategy, this framework efficiently simulates pollution propagation from moving emission sources. In the case of snowmobile traffic in Longyearbyen, the simulations confirm that thermal inversion conditions trap dense air masses near the ground, leading to significantly higher particulate matter concentrations.

What carries the argument

The robust variational framework for the discrete weak formulation of the time-dependent advection-diffusion problem, which supplies boundedness and inf-sup stability so the loss function directly bounds the difference to the unknown exact solution.

If this is right

  • The framework enables simulation of pollution from any moving emission sources without traditional mesh generation.
  • Collocation reduces training time while preserving the direct error relation in the loss.
  • Thermal inversion effects on pollutant buildup can be quantified using real sensor data from Arctic locations.
  • Time-dependent air quality predictions become feasible in environments with complex terrain and weather.
  • The method supports repeated runs to explore different emission scenarios and inversion strengths.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same stability construction could extend to other time-dependent transport problems where traditional PINNs lack error control.
  • Coupling the network with live sensor feeds might allow short-term forecasting of pollution spikes during inversions.
  • The approach could test how changes in emission patterns or temperature profiles alter accumulation without new code for each case.
  • Results may help evaluate whether restricting vehicle traffic during inversions would measurably improve measured air quality.

Load-bearing premise

The robust variational framework establishes boundedness and inf-sup stability for the discrete weak formulation of the time-dependent advection-diffusion problem, allowing the loss function to relate directly to the true unknown approximation error.

What would settle it

A side-by-side comparison of the neural network output against an independent high-resolution numerical solution or additional field measurements of particulate matter during documented thermal inversion events in Longyearbyen; systematic large deviations would show the error bound or dynamics are not captured.

Figures

Figures reproduced from arXiv: 2604.23003 by Eirik Valseth, Jacek Leszczy\'nski, Leszek Siwik, Maciej Paszy\'nski, Maciej Sikora, Manuela Bastidas Olivares, Marcin {\L}o\'s, Natalia Leszczy\'nska, Tomasz Maciej Ciesielski, Tomasz S{\l}u\.zalec.

Figure 1
Figure 1. Figure 1: The convergence of the PINN (a) and CRVPINN (b) training. view at source ↗
Figure 2
Figure 2. Figure 2: Snapshots of the snowmobiles pollution CRVPINN simulation. view at source ↗
Figure 3
Figure 3. Figure 3: Snowmobiles and the air quality sensor used to take the pollution concentration measure view at source ↗
Figure 4
Figure 4. Figure 4: Mobile measurements of NO2 from snowmobiles versus current Directive [40] and future policy [41]. In addition to the moving snowmobile measurements, we perform stationary measurements near Longyearbyen2 and the results are presented in Figures 7-9. As one may observe, the ambient concentration of (NO2) measured oscillated between 0 and 10 µg/m3 , the concentration of (PM2.5) measured oscillated between 0 a… view at source ↗
Figure 5
Figure 5. Figure 5: Mobile measurements of PM2.5 from snowmobiles versus current Directive [ view at source ↗
Figure 6
Figure 6. Figure 6: Mobile measurements of PM10 from snowmobiles versus current Directive [ view at source ↗
Figure 7
Figure 7. Figure 7: Stationary measurements of NO2 versus current Directive [40] and future policy [41] view at source ↗
Figure 8
Figure 8. Figure 8: Stationary measurements of PM2.5 versus current Directive [ view at source ↗
Figure 9
Figure 9. Figure 9: Stationary measurements of O3 versus current Directive [40] and future policy [41] view at source ↗
Figure 10
Figure 10. Figure 10: Stationary measurements of PM10 versus current Directive [ view at source ↗
read the original abstract

In this paper, we propose a Physics-Informed Neural Network framework for time-dependent simulations of pollution propagation originating from moving emission sources. We formulate a robust variational framework for the time-dependent advection-diffusion problem and establish the boundedness and inf-sup stability of the corresponding discrete weak formulation. Based on this mathematical foundation, we construct a robust loss function that is directly related to the true approximation error, defined as the difference between the neural network approximation and the (unknown) exact solution. Additionally, a collocation-based strategy is introduced to speed up neural network training. As a case study, we investigate pollution propagation caused by snowmobile traffic in Longyearbyen, Spitsbergen, supported by detailed in-field measurements collected using dedicated sensors. The proposed framework is applied to analyze the effects of thermal inversion on pollutant accumulation. Our results demonstrate that thermal inversion traps dense and humid air masses near the ground, significantly enhancing particulate matter (PM) concentration and worsening local air quality.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript proposes a collocation-based robust Physics-Informed Neural Network framework for time-dependent simulations of pollution propagation from moving emission sources. It formulates a variational framework for the advection-diffusion problem, asserts boundedness and inf-sup stability of the discrete weak formulation, and constructs a loss function directly related to the true (unknown) approximation error. A collocation strategy is used to accelerate training. The approach is applied as a case study to snowmobile traffic pollution in Longyearbyen, Spitsbergen, incorporating in-field sensor measurements to analyze thermal inversion effects on particulate matter concentrations.

Significance. If the stability and error-equivalence claims are rigorously established, the work would provide a valuable theoretical foundation for reliable PINN approximations of time-dependent advection-diffusion problems with moving sources, addressing a common weakness in standard PINN loss designs. Integration with real in-field data from Spitsbergen adds practical relevance for environmental modeling applications.

major comments (2)
  1. [Abstract / Variational Framework] Abstract and the section on the robust variational framework: The central claim that the robust variational formulation establishes boundedness and inf-sup stability for the discrete weak form of the time-dependent advection-diffusion problem (allowing the loss to directly control the true approximation error) is load-bearing. However, no explicit inf-sup constant is derived, and no argument is given that the neural-network trial space satisfies the Fortin condition uniformly in time for the moving-source operator. Without this verification, the loss-to-error relation remains formal.
  2. [Case Study / Results] Spitsbergen case study and results section: The conclusion that thermal inversion traps air masses and significantly enhances PM concentrations rests on in-field measurements and model outputs that lack quantitative validation, error bars, or direct comparisons to reference solutions or baseline models. This undermines the strength of the empirical claims about local air quality impacts.
minor comments (1)
  1. [Abstract] The abstract refers to 'detailed in-field measurements' without specifying how these data enter the training loss or validation; this integration should be clarified with a dedicated paragraph or equation.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed comments. We address each major point below, providing the strongest honest defense of the manuscript while indicating where revisions will be made to improve clarity and rigor.

read point-by-point responses

  1. Referee: [Abstract / Variational Framework] Abstract and the section on the robust variational framework: The central claim that the robust variational formulation establishes boundedness and inf-sup stability for the discrete weak form of the time-dependent advection-diffusion problem (allowing the loss to directly control the true approximation error) is load-bearing. However, no explicit inf-sup constant is derived, and no argument is given that the neural-network trial space satisfies the Fortin condition uniformly in time for the moving-source operator. Without this verification, the loss-to-error relation remains formal.

    Authors: The manuscript derives boundedness and inf-sup stability for the discrete weak formulation of the time-dependent advection-diffusion problem in the variational framework section, which underpins the loss design. We agree that an explicit inf-sup constant is not computed, as it would require a full analysis of the moving-source operator that depends on the specific source trajectory. For the neural-network trial space, the collocation strategy with dense sampling is intended to ensure a discrete inf-sup condition holds in practice. In the revised manuscript we will add a dedicated subsection with a proof sketch for uniform-in-time stability and numerical verification of the loss-to-error relation on manufactured solutions. revision: partial

  2. Referee: [Case Study / Results] Spitsbergen case study and results section: The conclusion that thermal inversion traps air masses and significantly enhances PM concentrations rests on in-field measurements and model outputs that lack quantitative validation, error bars, or direct comparisons to reference solutions or baseline models. This undermines the strength of the empirical claims about local air quality impacts.

    Authors: The case study is grounded in real in-field sensor measurements from Longyearbyen, and the model outputs show clear correlation with observed PM spikes during inversion events. We acknowledge the absence of error bars and direct baseline comparisons in the current version. Because no closed-form reference solution exists for the full moving-source scenario, we will add sensor uncertainty error bars, a quantitative comparison against a standard (non-robust) PINN on the same data, and relative error metrics on a simplified 2D manufactured-solution test case to strengthen the empirical support. revision: yes

Circularity Check

0 steps flagged

No circularity: stability claim and loss construction are presented as derived within the paper

full rationale

The paper states that it formulates a robust variational framework for the time-dependent advection-diffusion problem and establishes boundedness and inf-sup stability of the discrete weak formulation, then builds a loss directly related to the approximation error. No step reduces a claimed prediction or result to a fitted parameter, self-citation chain, or definitional renaming by construction. The central methodological step is an asserted mathematical derivation rather than an input-output equivalence, and the Spitsbergen case study applies the framework without evidence that any output is forced by the inputs themselves. This is the common case of a self-contained derivation whose validity may be debated on other grounds but does not exhibit circular reduction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim depends on the existence of a bounded and inf-sup stable discrete weak formulation for the advection-diffusion problem and on the neural network's capacity to minimize a loss that approximates the true solution error.

axioms (1)
  • domain assumption The time-dependent advection-diffusion problem admits a bounded and inf-sup stable discrete weak formulation.
    Invoked to justify the robust loss function construction.

pith-pipeline@v0.9.0 · 5528 in / 1206 out tokens · 43711 ms · 2026-05-08T12:18:32.424119+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

46 extracted references · 46 canonical work pages

  1. [1]

    A. V. G. Fabio Capello, Clinical Handbook of Air Pollution-Related Diseases, Springer Cham, 2019

    work page 2019

  2. [2]

    M. J. Wolf, D. C. Esty, H. Kim, M. L. Bell, S. Brigham, Q. Nortonsmith, S. Zaharieva, Z. A. Wendling, A. de Sherbinin, J. W. Emerson, New insights for tracking global and local trends in exposure to air pollutants, Environmental Science & Technology 56 (7) (2022) 3984–3996

    work page 2022

  3. [3]

    URLhttps://www.who.int/data/gho/data/themes/air-pollution

    World Health Organization, Air pollution data portal, accessed: 2024-11-25 (2024). URLhttps://www.who.int/data/gho/data/themes/air-pollution

    work page 2024

  4. [4]

    World Health Organization, WHO global air quality guidelines: particulate matter (PM2.5 and PM10), ozone, nitrogen dioxide, sulfur dioxide and carbon monoxide, World Health Organization, 2021

    work page 2021

  5. [5]

    Guarnieri, J

    M. Guarnieri, J. R. Balmes, Outdoor air pollution and asthma, The Lancet 383 (9928) (2014) 1581–1592

    work page 2014

  6. [6]

    A. I. Tiotiu, P. Novakova, D. Nedeva, H. J. Chong-Neto, S. Novakova, P. Steiropoulos, K. Kowal, Impact of air pollution on asthma outcomes, In- ternational Journal of Environmental Research and Public Health 17 (17) (2020)

    work page 2020

  7. [7]

    C. H. Li, K. Sayeau, A. K. Ellis, Air pollution and allergic rhinitis: Role in symptom exacerbation and strategies for management, Journal of Asthma and Allergy 13 (2020) 285–292

    work page 2020

  8. [8]

    Dámato, R

    G. Dámato, R. Pawankar, C. Vitale, M. Lanza, A. Molino, A. Stanziola, A. San- duzzi, A. Vatrella, M. DÁmato, Climate change and air pollution: Effects on respiratory allergy, Allergy, Asthma & Immunology Research 8 (5) (2016) 391–395

    work page 2016

  9. [9]

    R.-R. Duan, K. Hao, T. Yang, Y. Cui, Air pollution and chronic obstructive pulmonary disease, Chronic Diseases and Translational Medicine 06 (04) (2020) 260–269

    work page 2020

  10. [10]

    Hsu, C.-D

    H.-T. Hsu, C.-D. Wu, M.-C. Chung, T.-C. Shen, T.-J. Lai, C.-Y. Chen, R.-Y. Wang, C.-J. Chung, The effects of traffic-related air pollutants on chronic obstruc- tive pulmonary disease in the community-based general population, Respiratory Research 22 (2021). 24

    work page 2021

  11. [11]

    M. C. Turner, D. Krewski, W. R. Diver, C. A. P. III, R. T. Burnett, M. Jerrett, J. D. Marshall, S. M. Gapstur, Ambient air pollution and cancer mortality in the cancer prevention study ii, Environmental Health Perspectives 125 (8) (2017)

    work page 2017

  12. [12]

    M. C. Turner, Z. J. Andersen, A. Baccarelli, W. R. Diver, S. M. Gapstur, C. A. Pope III, D. Prada, J. Samet, G. Thurston, A. Cohen, Outdoor air pollution and cancer: An overview of the current evidence and public health recommendations, CA: A Cancer Journal for Clinicians 70 (6) (2020) 460–479

    work page 2020

  13. [13]

    A. A. Martoni, Air Pollution and Cancer, Springer International Publishing, 2018, pp. 445–457

    work page 2018

  14. [14]

    Franchini, P

    M. Franchini, P. M. Mannucci, Air pollution and cardiovascular disease, Throm- bosis Research 129 (3) (2012) 230–234

    work page 2012

  15. [15]

    Rajagopalan, S

    S. Rajagopalan, S. G. Al-Kindi, R. D. Brook, Air pollution and cardiovascu- lar disease: Jacc state-of-the-art review, Journal of the American College of Cardiology 72 (17) (2018) 2054–2070

    work page 2018

  16. [16]

    Bourdrel, M.-A

    T. Bourdrel, M.-A. Bind, Y. Bejot, O. Morel, J.-F. Argacha, Cardiovascular effects of air pollution, Archives of Cardiovascular Diseases 110 (11) (2017) 634–642

    work page 2017

  17. [17]

    P. Fu, K. K. L. Yung, Air pollution and alzheimer’s disease: A systematic review and meta-analysis, Journal of Alzheimer’s Disease 77 (2) (2020) 701–714

    work page 2020

  18. [18]

    Peters, Ambient air pollution and alzheimer’s disease: the role of the compo- sition of fine particles, Environmental Sciences 120 (3) (2023)

    A. Peters, Ambient air pollution and alzheimer’s disease: the role of the compo- sition of fine particles, Environmental Sciences 120 (3) (2023)

    work page 2023

  19. [19]

    Murata, L

    H. Murata, L. M. Barnhill, J. M. Bronstein, Air pollution and the risk of parkinson’s disease: A review, Movement Disorders 37 (5) (2022) 894–904

    work page 2022

  20. [20]

    D. Kwon, K. C. Paul, Y. Yu, K. Zhang, A. D. Folle, J. Wu, J. M. Bronstein, B. Ritz, Traffic-related air pollution and parkinson’s disease in central california, Environmental Research 240 (2024)

    work page 2024

  21. [21]

    Borroni, A

    E. Borroni, A. C. Pesatori, V. Bollati, M. Buoli, M. Carugno, Air pollution exposure and depression: A comprehensive updated systematic review and meta-analysis, Environmental Pollution 292 (2022)

    work page 2022

  22. [22]

    Gładka, T

    A. Gładka, T. Zatonski, J. Rymaszewska, Association between the long-term exposure to air pollution and depression, Advances in Clinical and Experimental Medicine 31 (10) (2022) 1139–1152. 25

    work page 2022

  23. [23]

    C. G. Zundel, P. Ryan, C. Brokamp, A. Heeter, Y. Huang, J. R. Strawn, H. A. Marusak, Air pollution, depressive and anxiety disorders, and brain effects: A systematic review, NeuroToxicology 93 (2022) 272–300

    work page 2022

  24. [24]

    Braithwaite, S

    I. Braithwaite, S. Zhang, J. B. Kirkbride, D. P. J. Osborn, J. F. Hayes, Air pollution (particulate matter) exposure and associations with depression, anxiety, bipolar, psychosis and suicide risk: A systematic review and meta-analysis, Environmental Health Perspectives 127 (12) (2019)

    work page 2019

  25. [25]

    C. Lim, G. Thurston, Air pollution, oxidative stress, and diabetes: a life course epidemiologic perspective, Current Diabetes Reports 19 (58) (2019)

    work page 2019

  26. [26]

    L. A. Paul, R. T. Burnett, J. C. Kwong, P. Hystad, A. van Donkelaar, L. Bai, M. S. Goldberg, E. Lavigne, R. Copes, R. V. Martin, A. Kopp, H. Chen, The impact of air pollution on the incidence of diabetes and survival among prevalent diabetes cases, Environment International 134 (2020)

    work page 2020

  27. [27]

    C.-Y. Hu, X. Gao, Y. Fang, W. Jiang, K. Huang, X.-G. Hua, X.-J. Yang, H.-B. Chen, Z.-X. Jiang, X.-J. Zhang, Human epidemiological evidence about the association between air pollution exposure and gestational diabetes mellitus: Systematic review and meta-analysis, Environmental Research 180 (2020)

    work page 2020

  28. [28]

    Sikora, A

    M. Sikora, A. Oliver-Serra, L. Siwik, N. Leszczyńska, T. M. Ciesielski, E. Valseth, J. Leszczyński, A. Paszyńska, M. Paszyński, Graph grammars and physics informed neural networks for simulating of pollution propagation on spitzbergen, Applied Soft Computing 182 (2025) 113394

    work page 2025

  29. [29]

    Timlin, A

    U. Timlin, A. Meyer, T. Nordström, A. Rautio, Permafrost thaw challenges and life in svalbard, Current Research in Environmental Sustainability 4 (2022) 100122

    work page 2022

  30. [30]

    Pacyna, E

    J. Pacyna, E. Pacyna, Environmental Determinants of Human Health, Springer Nature, Cham, Switzerland, 2018

    work page 2018

  31. [31]

    Andersson Stavridis, S

    M. Andersson Stavridis, S. B. Røed, B. B. Hansen, Øyvind Mikkelsen, T. M. Ciesielski, B. M. Jenssen, Tracing the footprints of arctic pollution: Spatial variations in toxic and essential elements in svalbard reindeer (rangifer tarandus platyrhynchus) faeces, Science of The Total Environment 906 (2024) 167562

    work page 2024

  32. [32]

    M.Łoś, T.Służalec, P.Maczuga, A.Vilkha, C.Uriarte, M.Paszyński, Collocation- based robust variational physics-informed neural networks (crvpinns), Computers & Structures 316 (2025) 107839. 26

    work page 2025

  33. [33]

    Raissi, P

    M. Raissi, P. Perdikaris, G. E. Karniadakis, Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations, Journal of Computational physics 378 (2019) 686–707

    work page 2019

  34. [34]

    S. Cai, Z. Mao, Z. Wang, M. Yin, G. E. Karniadakis, Physics-informed neural networks (PINNs) for fluid mechanics: A review, Acta Mechanica Sinica 37 (12) (2021) 1727–1738

    work page 2021

  35. [35]

    Y. Chen, L. Lu, G. E. Karniadakis, L. Dal Negro, Physics-informed neural networks for inverse problems in nano-optics and metamaterials, Optics Express 28 (8) (2020) 11618–11633.doi:10.1364/OE.384875

    work page doi:10.1364/oe.384875 2020

  36. [36]

    Kharazmi, Z

    E. Kharazmi, Z. Zhang, G. E. Karniadakis, hp-vpinns: Variational physics- informed neural networks with domain decomposition, Computer Methods in Applied Mechanics and Engineering 374 (2021) 113547

    work page 2021

  37. [37]

    Rojas, P

    S. Rojas, P. Maczuga, J. Muñoz-Matute, D. Pardo, M. Paszyński, Robust variational physics-informed neural networks, Computer Methods in Applied Mechanics and Engineering 425 (2024) 116904

    work page 2024

  38. [38]

    URLhttps://www.aqmd.gov/aq-spec/sensordetail/airly

    South Coast Air Quality Management District, Airly sensor detail, accessed: 2024-11-25 (2024). URLhttps://www.aqmd.gov/aq-spec/sensordetail/airly

    work page 2024

  39. [39]

    URLhttps://airly.org/en/features/air-quality-sensors/

    Airly, Air quality sensors, accessed: 2024-11-25 (2024). URLhttps://airly.org/en/features/air-quality-sensors/

    work page 2024

  40. [40]

    directive 2008/50/ec of the european parliament and of the council of 21 may 2008 on ambient air quality and cleaner air for europe (2008)

    Document 32008l0050. directive 2008/50/ec of the european parliament and of the council of 21 may 2008 on ambient air quality and cleaner air for europe (2008). URL https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri= CELEX:32008L0050

    work page 2008

  41. [41]

    proposal for a directive of the european parliament and of the council on ambient air quality and cleaner air for europe (recast) (2022)

    Directorate-general for environment. proposal for a directive of the european parliament and of the council on ambient air quality and cleaner air for europe (recast) (2022). URL https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=COM% 3A2022%3A542%3AFIN

    work page 2022

  42. [42]

    Barabasz, S

    B. Barabasz, S. Migórski, R. Schaefer, M. Paszyński, Multi-deme, twin adaptive strategy hp-hgs, Inverse Problems in Science and Engineering 19 (1) (2011) 3–16. 27

    work page 2011

  43. [43]

    Gajda-Zagórska, R

    E. Gajda-Zagórska, R. Schaefer, M. Smołka, M. Paszyński, D. Pardo, A hy- brid method for inversion of 3d dc resistivity logging measurements, Natural computing 14 (3) 355–374

    work page

  44. [44]

    Woźniak, M

    M. Woźniak, M. Łoś, M. Paszyński, L. Dalcin, V. M. Calo, Parallel fast isogeo- metric solvers for explicit dynamics, Computing and Informatics 36 (2) (2017) 423–448

    work page 2017

  45. [45]

    Paszyński, On the parallelization of self-adaptive hp-finite element methods part i.composite programmablegraph grammar model, Fundamenta Informaticae 93 (4) (2009) 411–434

    M. Paszyński, On the parallelization of self-adaptive hp-finite element methods part i.composite programmablegraph grammar model, Fundamenta Informaticae 93 (4) (2009) 411–434

    work page 2009

  46. [46]

    Paszyński, On the parallelization of self-adaptive hp-finite element methods part ii

    M. Paszyński, On the parallelization of self-adaptive hp-finite element methods part ii. partitioning communication agglomeration mapping (pcam) analysis, Fundamenta Informaticae 93 (4) (2009) 435–457. 28

    work page 2009