Pith. sign in

REVIEW 4 major objections 6 minor 56 references

U3DWind is a five-city, building-resolved 3D low-altitude wind dataset of 720 simulations that turns urban air mobility wind hazards into a public, task-based benchmark.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.5

2026-07-11 18:36 UTC pith:QCNW2376

load-bearing objection Solid multi-city LBM-LES UAM wind resource with five formal tasks and extensive Shanghai baselines; soft spots are deferred validation, Shanghai-only reported numbers, and proxy operational labels, not a broken central claim. the 4 major comments →

arxiv 2607.04495 v1 pith:QCNW2376 submitted 2026-07-05 physics.flu-dyn physics.geo-ph

U3DWind: A Low Altitude Wind Field Dataset and Benchmark for Urban Air Mobility

Shixiong Zhou , Huanxia Wei , Chao Xia , Yingying Xing , Changmin Jiang , Hai Yang , Shuai Jia This is my paper
classification physics.flu-dyn physics.geo-ph
keywords Urban Air MobilityWind FieldUrban Airspace ManagementComputational Fluid DynamicsLattice Boltzmann MethodLarge-Eddy SimulationNoise Propagation Modeling
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

Safe urban air mobility depends on winds, gusts, and building-induced turbulence that reshape routes, vertiport choices, and community noise—but public data for city-scale, data-driven planning have been thin in coverage, realism, and shared tests. This paper introduces U3DWind: 720 building-resolved Lattice Boltzmann large-eddy simulations over Beijing, Shanghai, Guangzhou, Shenzhen, and Hong Kong, spanning 16 inflow directions, three wind speeds, and three seasonal scenarios at 10 m resolution, with full 3D velocity, turbulent kinetic energy, density, and solid masks. It then defines five operational tasks—wind-field prediction, sparse-sensor reconstruction, site wind-exposure ranking, airworthiness risk scoring, and noise propagation—and reports baselines on Shanghai under a held-out inflow split. A sympathetic reader cares because without multi-city, morphology-aware volumes and common metrics, wind-aware UAM work stays locked in isolated CFD runs and single-site studies.

Core claim

The authors claim that U3DWind—a multi-city, building-resolved low-altitude wind-field resource of 720 LBM-LES cases at 10 m, plus five formalized UAM tasks and baselines—fills the public data gap and enables systematic evaluation of wind-induced impacts for urban airspace management and data-driven high-fidelity urban flow simulation.

What carries the argument

U3DWind itself: GPU-accelerated LBM-LES volumes driven by real building footprints, DEM terrain, and NASA POWER climatology, paired with five mathematically specified tasks (surrogate modeling, airworthiness risk, sparse reconstruction, site ranking, noise propagation) and unified metrics.

Load-bearing premise

That time-averaged 10 m simulated winds from simplified power-law inflows and public building maps are faithful enough to serve as ground truth for operational risk scores and community noise decisions.

What would settle it

Co-located mast or lidar measurements of velocity and turbulence in a released city domain that systematically disagree with the matching U3DWind fields, or commercial eVTOL go/no-go logs that do not track the Task 2 worst-of-three exceedance scores, would undercut the claim that the dataset is operational ground truth.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

Share X Bluesky LinkedIn Reddit HN

If this is right

  • Shared multi-city volumes let researchers train and compare neural operators, assimilation methods, and planners under the same inflow and morphology conditions.
  • Vertiport and route planning can rank sites and corridors by building-resolved wind exposure instead of height or comfort heuristics alone.
  • Sparse rooftop and corridor sensors can reconstruct dense fields by projecting onto dataset-derived reduced-order modes.
  • Community noise estimates can replace free-field screening with building-diffraction and wind-modulated footprints.
  • Fast surrogates of the LES fields make city-scale aerodynamic and acoustic queries feasible for digital-twin style operations.

Where Pith is reading between the lines

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

  • The five morphologies and climatologies form a natural cross-city transfer test that single-city LES campaigns cannot provide.
  • The large accuracy gap between prior-based assimilation and pure deep models at low sensor counts implies operational monitoring will stay hybrid for some time.
  • Adding unsteady LES sequences to the same task suite would directly stress gust-aware control and real-time re-routing.
  • Tying Task 2 labels to actual flight logs would convert literature thresholds into operator-defined go/no-go criteria.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The manuscript introduces U3DWind, a multi-city building-resolved low-altitude wind dataset of 720 LBM-LES cases at 10 m resolution over Beijing, Shanghai, Guangzhou, Shenzhen, and Hong Kong, forced by NASA POWER climatology and OSM/Microsoft building masks. It releases 3D3C velocity, TKE, density, and fluid–solid masks, and defines five UAM-oriented benchmark tasks—surrogate wind-field prediction, sparse-sensor reconstruction, site wind-exposure ranking, airworthiness wind-compliance risk scoring, and noise propagation—with extensive Shanghai baselines (Tables 3–9, Figs. 3–8). The central claim is that this resource enables systematic evaluation of wind-induced UAM impacts and serves as an open benchmark for urban airspace management and data-driven urban flow simulation.

Significance. If the resource is released as described, it would fill a genuine gap: public city-scale, building-resolved 3D wind fields with task-oriented UAM benchmarks are scarce. Strengths include the multi-city morphological and climatological diversity (Table 1, Fig. 2), the GPU LBM-LES generation pipeline, explicit task formalisms (Table 2), and a broad, reproducible baseline suite spanning classical, operator-learning, assimilation, ranking, and hybrid acoustic methods. The work is useful to fluid dynamics, transportation, and urban systems communities even if labels remain simulation-proxy rather than flight-log validated.

major comments (4)
  1. [§3.1 Data Generation Pipeline] §3.1 and §5.2: Lattice verification and observational validation are deferred entirely to Supplementary S1, which is not in the main text. For a dataset paper whose central claim is that the fields support operational UAM evaluation, the main manuscript needs at least a concise quantitative summary (e.g., against wind-tunnel or urban mast data) of LBM-LES fidelity at the released 10 m resolution and power-law inlet setup. Without that, reported Task 1/3 accuracy is fidelity to the solver, not to nature.
  2. [§4.2 Task 2 / Table 4] §3.3–§4.2, Table 2 and Table 4: Task 2 risk is the worst exceedance of three fixed thresholds (7.6 / 7.62 / 12.0 m/s) applied to time-averaged ¯u and k. Perfect AUC=1.000 for tree models on the Shanghai IE split is therefore self-consistency of a proxy label, not evidence that the label tracks operational go/no-go. The authors acknowledge that autonomous eVTOL rules are not formalized (§5.2). The abstract and §1 claim that the benchmark enables “systematic evaluation of wind-induced impacts” should be qualified to “simulation-proxy evaluation,” or an external fidelity check (even limited) should be added.
  3. [§3.4 Evaluation Protocols] §3.4 and §4 opening: All reported experiments use only the Shanghai domain and an inflow-extreme (IE) split. The multi-city contribution is load-bearing in the abstract and §1, yet no cross-city transfer, leave-one-city-out, or even multi-city descriptive statistics of model error are shown. Either add a minimal multi-city experiment or clearly reframe the present paper as a Shanghai-primary benchmark with multi-city data release pending full evaluation.
  4. [§5.2 Limitations] §3.2 and §5.2: The release is stationary/time-averaged fields. Tasks 2 and 5 (gust exceedance, community SPL under wind refraction) are sensitive to unsteadiness that the authors themselves flag as future work. The manuscript should state more explicitly which of the five tasks are appropriate for time-averaged fields and which require the planned unsteady sequences, so users do not over-interpret the current labels.
minor comments (6)
  1. [Data Availability] Data Availability: full 10 m data are request-only (>50 TB); only 20 m is promised open after review. State clearly what will be public at acceptance (fields, masks, task labels, code) so the “open benchmark” claim is checkable.
  2. [Table 3] Table 3: several deep baselines report ± std while others do not; clarify seeds/runs. SSIM plateau <0.42 is noted but not linked to a recommended multi-scale or generative fix beyond a brief remark.
  3. [Figure 1] Fig. 1 caption and §1: “invisible turbulence” is informal; prefer “building-induced turbulence not captured by mesoscale products.”
  4. [§3.4] Eqs. (1)–(5) and Table 2: notation for fluid mask Ω_fluid vs χ is slightly inconsistent across tasks; unify.
  5. [§4.5 Task 5] Task 5 reference solver (Maekawa + PE) is described clearly, but third-octave band count and receiver-plane height (~80 m) should be fixed in one place for reproducibility.
  6. [Throughout] Minor typos: “UA V” spacing, “V olocopter”, “Changming” vs “Changmin” in affiliations/CRediT.

Circularity Check

1 steps flagged

No derivation-by-construction circularity; mild self-citation of the authors' LBM-LES generator is normal for a simulation dataset paper and does not force the headline claims.

specific steps
  1. self citation load bearing [§3.1 Data Generation Pipeline; also Abstract and [21]]
    "The U3DWind dataset is generated using LatticeUrbanWind (LUW), our open-source GPU-accelerated lattice Boltzmann large-eddy simulation (LBM-LES) framework [21]. ... For lattice verification as well as validations against observation, see Supplementary Material S1."

    The sole source of the released 3D3C fields is the authors' own LUW framework, cited as prior work by overlapping authors, with main-text validation only deferred to S1. This is load-bearing for data provenance. It is not, however, a uniqueness theorem or fitted ansatz that forces the benchmark results: the five tasks and baseline scores are defined and measured against that generated corpus in the usual simulation-benchmark sense, not reduced to the citation by construction.

full rationale

U3DWind is a dataset-and-benchmark paper, not a closed-form derivation of a physical law. The load-bearing chain is: (i) generate building-resolved fields with the authors' GPU LBM-LES (LatticeUrbanWind), (ii) define five operationally motivated tasks with explicit I/O and metrics, (iii) train/evaluate baselines against those fields or labels. Surrogate accuracy (Tasks 1, 3) is fidelity to the same LBM-LES used as ground truth—standard for simulation benchmarks, not a reduction of a claimed prediction to its fitted inputs. Task 2 risk is the worst exceedance of three fixed external thresholds (7.6 m/s UAM limit, FAR §23.341/§25.341 25 ft/s gust, 12 m/s eVTOL ground ops) applied to ¯u and k; those thresholds are not fitted to maximize Table 4 scores. Task 4 GT is the per-case LBM-LES approach-cylinder P95; Task 5 GT is a stated Maekawa + PE reference solver. The only mild circularity-adjacent element is self-citation of LatticeUrbanWind [21] as the generation engine with validation deferred to Suppl. S1—load-bearing for data provenance, not a uniqueness theorem or ansatz that forces the scientific result. No equation equates a reported prediction to a fitted parameter by construction. Score 1 reflects that single non-forcing self-citation; the central claim (open multi-city benchmark resource) remains independent of circular reduction.

Axiom & Free-Parameter Ledger

5 free parameters · 5 axioms · 3 invented entities

The central claim rests on accepting LBM-LES plus climatological power-law inflows as operationally relevant urban wind truth, on chosen scenario discretizations and risk thresholds, and on treating the five task definitions as the right UAM evaluation axes. No new physical particle or force is postulated; the invented objects are the dataset, framework name, and task suite.

free parameters (5)
  • Power-law exponents α (per city and season)
    Fitted from NASA POWER 10 m and 50 m climatological means; they set the entire vertical inflow shape for every case (Table 1, §3.2).
  • Reference 10 m wind speeds {3, 6, 9} m/s
    Hand-chosen operational tiers for eVTOL envelope; they discretize the dataset and all speed-dependent tasks (§3.2).
  • Task 2 operational thresholds 7.6 / 7.62 / 12.0 m/s and rescaling to r̂∈[0,1]
    Chosen from UAM literature, FAR design-gust numbers, and disclosed eVTOL envelopes; they fully define the risk label (§4.2).
  • Grid resolution 10 m and Task 1 40 m crop/downsample
    Computational choices that set what structures are resolved and what baselines see (§3.1; Table 3 caption).
  • IE hold-out directions 90° and 270° (and related Task 4 extreme-quartile split)
    Author-chosen generalization protocol that determines all reported metrics (§3.4).
axioms (5)
  • domain assumption GPU LBM-LES with the stated LES closure and von Kármán synthetic inlet adequately represents building-resolved urban canopy flow for UAM-relevant statistics.
    Load-bearing for treating fields as ground truth; validation deferred to Suppl. S1 (§3.1).
  • domain assumption 1D power-law profiles from NASA POWER city-centroid reanalysis plus 16 discrete azimuths capture the relevant inflow climatology.
    Defines all 720 cases; ignores full mesoscale heterogeneity and non-power-law profiles (§3.2).
  • domain assumption OSM + Microsoft building footprints and Copernicus GLO-30 DEM, voxelized at 10 m, are accurate enough for aerodynamic and acoustic tasks.
    Geometry errors propagate into wakes, ranking, and diffraction (§3.1).
  • ad hoc to paper Time-averaged stationary fields suffice for the five operational tasks (including airworthiness risk and noise).
    Authors later note missing unsteady gusts and vortex streets as a limitation (§5.2).
  • ad hoc to paper Shanghai IE-split results are a fair primary benchmark for the multi-city resource.
    Cross-city hold-out is reserved for a future release (§4 intro; §5.2).
invented entities (3)
  • U3DWind dataset (720 multi-city LBM-LES cases) no independent evidence
    purpose: Provide public building-resolved 3D low-altitude wind fields for UAM research.
    Primary contribution; independent evidence is the planned release and described generation pipeline, not external measurements.
  • LatticeUrbanWind (LUW) framework no independent evidence
    purpose: GPU LBM-LES engine used to generate the dataset at city scale.
    Cited as authors’ open-source framework [21]; independent evidence would require third-party validation beyond this paper.
  • Five U3DWind benchmark tasks and associated risk/ranking/noise labels no independent evidence
    purpose: Standardize evaluation of prediction, reconstruction, airworthiness risk, vertiport exposure, and noise.
    Task definitions and labels are paper-constructed operational proxies (§3.3–3.4).

pith-pipeline@v1.1.0-grok45 · 30053 in / 3810 out tokens · 41758 ms · 2026-07-11T18:36:26.070685+00:00 · methodology

0 comments
Cite this review

Pith. "Pith review of U3DWind: A Low Altitude Wind Field Dataset and Benchmark for Urban Air Mobility." pith.science (2026). https://pith.science/paper/QCNW2376

@misc{pith2026260704495,
  author       = {Pith},
  title        = {Pith review of: U3DWind: A Low Altitude Wind Field Dataset and Benchmark for Urban Air Mobility},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QCNW2376}},
  note         = {Machine review of arXiv:2607.04495}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Urban Air Mobility (UAM) requires reliable assessment of low-altitude wind hazards, because winds, gusts, and building-induced turbulence have been recognized as critical factors affecting vehicle stability, route feasibility, vertiport siting, and airspace management. While wind-tunnel experiments, computational fluid dynamics (CFD), multiscale downscaling, reduced-order models, and UAV planning datasets have advanced wind-aware analysis, public resources for data-driven, city-scale UAM planning remain limited in geographic coverage, scenario diversity, vertical extent, building realism, and task-oriented benchmarking. To address this gap, we introduce U3DWind, a building-resolved low-altitude wind-field dataset generated using our GPU-accelerated Lattice Boltzmann Method--Large-Eddy Simulation (LBM-LES) framework for rapid urban flow simulation. U3DWind covers five megacities in China: Beijing, Shanghai, Guangzhou, Shenzhen, and Hong Kong. It contains 720 simulations, with 16 inflow directions, three reference wind speeds, and three seasonal atmospheric scenarios (annual, summer, and winter) for each city. At a 10 m grid resolution, the dataset provides three-dimensional three-component (3D3C) velocity, turbulent kinetic energy (TKE), flow density, and fluid--solid masks. To support operationally relevant evaluation, we further define five baseline tasks: wind-field prediction, sparse-sensor wind-field reconstruction, site wind-exposure ranking, airworthiness wind-compliance risk scoring, and noise propagation modeling. As a multi-city, building-resolved 3D urban wind-field dataset, U3DWind enables systematic evaluation of wind-induced impacts in low-altitude traffic scenarios and provides an open benchmark for urban airspace management and data-driven high-fidelity urban flow simulation.

Figures

Figures reproduced from arXiv: 2607.04495 by Changmin Jiang, Chao Xia, Hai Yang, Huanxia Wei, Shixiong Zhou, Shuai Jia, Yingying Xing.

Figure 1
Figure 1. Figure 1: Sectional visualization of the instantaneous wind speed over the Shanghai reference domain in the U3DW [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Geographic footprint of the U3DWind release and per-city annual wind climatology. The base map (Lambert-conformal projection) places the five U3DWind cities on mainland China and Hong Kong; each city marker is paired with a 16-sector annual wind rose drawn at the corresponding inset, with sector mass colour-coded by wind-speed tier (calm, light, moderate, strong). The annual rose is the simple mean of the … view at source ↗
Figure 3
Figure 3. Figure 3: Task 1 qualitative prediction at z = 100 m on a held-out easterly-inflow case over the 384 × 384-cell urban-core crop. Panel a is the LBM-LES ground truth; panels b, c, d show representative members of the three baseline families: the analytical log-law projection produces a spatially flat field with no urban structure, kriging smooths the training-set ensemble into a blurred spatial field that captures br… view at source ↗
Figure 4
Figure 4. Figure 4: Task 2 baseline landscape on the Shanghai release. Markers are color-coded by baseline family. Panel a is the threshold-classification [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Task 3 sensor-count scaling. Relative L2 reconstruction error on the held-out easterly-inflow case as a function of the number of rooftop sensors Ns . Kalman assimilation and Gappy POD saturate the error curve by Ns = 64; the sensor Transformer and graph-attention baselines approach them at Ns = 256 but require an order of magnitude more sensors for comparable error; classical spatial interpolants (inverse… view at source ↗
Figure 6
Figure 6. Figure 6: Task 3 reconstruction at z = 100 m on the held-out easterly-inflow case with Ns = 64 sensors drawn from the rooftop and corridor placement pool. Panel a is the LBM-LES ground truth; panels b, c, d are three reconstructions spanning the methodological axis. Inverse-distance weighting is a geometry-only baseline that produces blob-like interpolants unable to capture building wakes; Gappy POD with K = 6 modes… view at source ↗
Figure 7
Figure 7. Figure 7: Task 4 baseline landscape on the IE split (Shanghai release). Both panels read directly from the same JSON that feeds Table 8; markers [PITH_FULL_IMAGE:figures/full_fig_p017_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Task 5 UAV noise propagation for the easterly inflow case (ANG_90) on the Shanghai U3DW [PITH_FULL_IMAGE:figures/full_fig_p018_8.png] view at source ↗

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

56 extracted references · 2 linked inside Pith

  1. [1]

    Bauranov, J

    A. Bauranov, J. Rakas, Designing airspace for urban air mobility: A review of concepts and approaches, Progress in Aerospace Sciences 125 (2021) 100726

    2021

  2. [2]

    Johnson, C

    W. Johnson, C. Silva, NASA concept vehicles and the engineering of advanced air mobility aircraft, Aeronautical Journal 126 (2022) 59–91

    2022

  3. [3]

    Pons-Prats, T

    J. Pons-Prats, T. Živojinovi ´c, J. Kuljanin, On the understanding of the current status of urban air mobility devel- opment and its future trends, Transportation Research Part E: Logistics and Transportation Review 166 (2022) 102868

    2022

  4. [4]

    E. R. Mueller, P. H. Kopardekar, K. H. Goodrich, Enabling airspace integration for high-density on-demand mobility operations, in: 17th AIAA Aviation Technology, Integration, and Operations Conference, 2017, p. 3086

    2017

  5. [5]

    Reiche, A

    C. Reiche, A. P. Cohen, C. Fernando, An initial assessment of the potential weather barriers of urban air mobility, IEEE Transactions on Intelligent Transportation Systems 22 (2021) 6018–6027. 21

    2021

  6. [6]

    Al Labbad, A

    M. Al Labbad, A. Wall, G. L. Larose, F. Khouli, H. Barber, Experimental investigations into the effect of urban airflow characteristics on Urban Air Mobility applications, Journal of Wind Engineering and Industrial Aerodynamics 229 (2022) 105126

    2022

  7. [7]

    Krawczyk, R

    Z. Krawczyk, R. K. S. S. Vuppala, R. Paul, K. Kara, Evaluating reduced-order urban wind models for simulating flight dynamics of advanced aerial mobility aircraft, Aerospace 11 (10) (2024) 830

    2024

  8. [8]

    Çetin, A

    E. Çetin, A. M. Cano, R. Deransy, S. Tres, C. Barrado, Implementing mitigations for improving societal accep- tance of urban air mobility, Drones 6 (2022) 28

    2022

  9. [9]

    K. A. Pascioni, S. A. Rizzi, Tonal noise prediction of a distributed propulsion unmanned aerial vehicle, in: 2018 AIAA/CEAS Aeroacoustics Conference, 2018, p. 2951

    2018

  10. [10]

    Blocken, T

    B. Blocken, T. Stathopoulos, J. P. A. J. van Beeck, Pedestrian-level wind conditions around buildings: Review of wind-tunnel and CFD techniques and their accuracy for wind comfort assessment, Building and Environment 100 (2016) 50–81

    2016

  11. [11]

    Mochida, I

    A. Mochida, I. Y . F. Lun, Prediction of wind environment and thermal comfort at pedestrian level in urban area, Journal of Wind Engineering and Industrial Aerodynamics 96 (2008) 1498–1527

    2008

  12. [12]

    Tominaga, A

    Y . Tominaga, A. Mochida, R. Yoshie, H. Kataoka, T. Nozu, M. Yoshikawa, T. Shirasawa, AIJ guidelines for practical applications of CFD to pedestrian wind environment around buildings, Journal of Wind Engineering and Industrial Aerodynamics 96 (10–11) (2008) 1749–1761

    2008

  13. [13]

    Toparlar, B

    Y . Toparlar, B. Blocken, B. Maiheu, G. F. van Heijst, A review on the CFD analysis of urban microclimate, Renewable and Sustainable Energy Reviews 80 (2017) 1613–1640

    2017

  14. [14]

    Ricci, M

    A. Ricci, M. Burlando, M. P. Repetto, B. Blocken, Static downscaling of mesoscale wind conditions into an urban canopy layer by a CFD microscale model, Building and Environment 225 (2022) 109626

    2022

  15. [15]

    R. K. S. S. Vuppala, Z. Krawczyk, R. Paul, K. Kara, Modeling advanced air mobility aircraft in data-driven reduced order realistic urban winds, Scientific Reports 14 (2024) 383

    2024

  16. [16]

    Rienecker, V

    H. Rienecker, V . Hildebrand, H. Pfifer, Energy optimal 3D flight path planning for unmanned aerial vehicle in urban environments, CEAS Aeronautical Journal 14 (2023) 621–636

    2023

  17. [17]

    Ebert, J

    C. Ebert, J. Weiss, M. Uijt de Haag, C. Ruwisch, F. J. Silvestre, Trajectory planning in windy urban environment using gappy proper orthogonal decomposition for wind estimates, AIAA Journal 61 (6) (2023) 2640–2651

    2023

  18. [18]

    Blocken, W

    B. Blocken, W. D. Janssen, T. van Hooff, CFD simulation for pedestrian wind comfort and wind safety in urban areas: General decision framework and case study for the Eindhoven university campus, Environmental Modelling & Software 30 (2012) 15–34

    2012

  19. [19]

    Baskar, A

    D. Baskar, A. A. Gorodetsky, A simulated wind-field dataset for testing energy efficient path-planning algorithms for UA Vs in urban environment, in: AIAA A VIATION 2020 FORUM, American Institute of Aeronautics and Astronautics, Reston, V A, 2020

    2020

  20. [20]

    Maronga, S

    B. Maronga, S. Banzhaf, C. Burmeister, T. Esch, R. Forkel, D. Fröhlich, et al., Overview of the PALM model system 6.0, Geoscientific Model Development 13 (2020) 1335–1372

    2020

  21. [21]

    H. Wei, Q. Jia, S. Han, LatticeUrbanWind: Meter-level, hundred-km-scale, minute-computation, billion- grid winds supporting next-gen real-time low-altitude applications,https://github.com/hweifluids/ LatticeUrbanWind, gitHub repository, accessed 1 June 2026 (2026)

    2026

  22. [22]

    Boeing, OSMnx: New methods for acquiring, constructing, analyzing, and visualizing complex street net- works, Computers, Environment and Urban Systems 65 (2017) 126–139

    G. Boeing, OSMnx: New methods for acquiring, constructing, analyzing, and visualizing complex street net- works, Computers, Environment and Urban Systems 65 (2017) 126–139. 22

    2017

  23. [23]

    D. S. Nithya, G. Quaranta, V . Muscarello, M. Liang, Review of wind flow modelling in urban environments to support the development of urban air mobility, Drones 8 (4) (2024) 147

    2024

  24. [24]

    Z. Xie, I. P. Castro, LES and RANS for turbulent flow over arrays of wall-mounted obstacles, Flow, Turbulence and Combustion 76 (2006) 291–312

    2006

  25. [25]

    H. J. S. Fernando, Fluid dynamics of urban atmospheres in complex terrain, Annual Review of Fluid Mechanics 42 (2010) 365–389

    2010

  26. [26]

    M. O. Letzel, M. Krane, S. Raasch, High resolution urban large-eddy simulation studies from street canyon to neighbourhood scale, Atmospheric Environment 42 (38) (2008) 8770–8784

    2008

  27. [27]

    M. F. King, A. Khan, N. Delbosc, H. L. Gough, C. Halios, J. F. Barlow, Modelling urban airflow and natural ventilation using a GPU-based lattice-Boltzmann method, Building and Environment 125 (2017) 273–284

    2017

  28. [28]

    S. Lenz, M. Schönherr, M. Geier, M. Krafczyk, A. Pasquali, A. Christen, Towards real-time simulation of turbulent air flow over a resolved urban canopy using the cumulant lattice Boltzmann method, Journal of Wind Engineering and Industrial Aerodynamics 189 (2019) 151–162

    2019

  29. [29]

    Jacob, P

    J. Jacob, P. Sagaut, Wind comfort assessment by means of large eddy simulation with lattice Boltzmann method in full scale city area, Building and Environment 139 (2018) 110–124

    2018

  30. [30]

    Merlier, J

    L. Merlier, J. Jacob, P. Sagaut, Lattice-Boltzmann large-eddy simulation of pollutant dispersion in complex urban environment, Building and Environment 148 (2019) 634–652

    2019

  31. [31]

    N. H. Ahmad, A. Inagaki, M. Kanda, N. Onodera, T. Aoki, Large-eddy simulation of the gust index in an urban area using the lattice Boltzmann method, Boundary-Layer Meteorology 163 (2017) 447–467

    2017

  32. [32]

    Olauson, ERA5: The new champion of wind power modelling?, Renewable Energy 126 (2018) 322–331

    J. Olauson, ERA5: The new champion of wind power modelling?, Renewable Energy 126 (2018) 322–331

    2018

  33. [33]

    Z. Li, N. Kovachki, K. Azizzadenesheli, B. Liu, K. Bhattacharya, A. Stuart, A. Anandkumar, Fourier neural operator for parametric partial differential equations, in: International Conference on Learning Representations, 2021

    2021

  34. [34]

    G. Wen, Z. Li, K. Azizzadenesheli, A. Anandkumar, S. M. Benson, U-FNO: An enhanced Fourier neural operator-based deep-learning model for multiphase flow, Advances in Water Resources 163 (2022) 104180

    2022

  35. [35]

    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 Computa- tional Physics 378 (2019) 686–707

    2019

  36. [36]

    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 (2021) 1727–1738

    2021

  37. [37]

    L. Lu, P. Jin, G. Pang, Z. Zhang, G. E. Karniadakis, Learning nonlinear operators via DeepONet based on the universal approximation theorem of operators, Nature Machine Intelligence 3 (2021) 218–229

    2021

  38. [38]

    Pathak, S

    J. Pathak, S. Subramanian, P. Harrington, S. Raja, A. Chattopadhyay, M. Mardani, T. Kurth, D. Hall, Z. Li, K. Azizzadenesheli, P. Hassanzadeh, K. Kashinath, A. Anandkumar, FourCastNet: A global data-driven high- resolution weather model using adaptive Fourier neural operators, arXiv preprint arXiv:2202.11214 (2022)

    Pith/arXiv arXiv 2022

  39. [39]

    Raissi, A

    M. Raissi, A. Yazdani, G. E. Karniadakis, Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations, Science 367 (2020) 1026–1030

    2020

  40. [40]

    P. D. Vascik, R. J. Hansman, Constraint identification in on-demand mobility for aviation through an exploratory case study of los angeles, in: 17th AIAA Aviation Technology, Integration, and Operations Conference, 2017, p. 3083. 23

    2017

  41. [41]

    T. Chen, C. Guestrin, XGBoost: A scalable tree boosting system, in: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016, pp. 785–794

    2016

  42. [42]

    Stankevi ˇci¯ut˙e, A

    K. Stankevi ˇci¯ut˙e, A. M. Alaa, M. van der Schaar, Conformal time-series forecasting, in: Advances in Neural Information Processing Systems, V ol. 34, 2021, pp. 6216–6228

    2021

  43. [43]

    S. Bai, J. Z. Kolter, V . Koltun, An empirical evaluation of generic convolutional and recurrent networks for sequence modeling, arXiv preprint arXiv:1803.01271 (2018)

    Pith/arXiv arXiv 2018

  44. [44]

    Vaswani, N

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser, I. Polosukhin, Attention is all you need, in: Advances in Neural Information Processing Systems 30 (NeurIPS 2017), 2017, pp. 5998–6008

    2017

  45. [45]

    W. L. Hamilton, R. Ying, J. Leskovec, Inductive representation learning on large graphs, in: Advances in Neural Information Processing Systems 30 (NeurIPS 2017), 2017, pp. 1024–1034

    2017

  46. [46]

    Fukami, R

    K. Fukami, R. Maulik, N. Ramachandra, K. Fukagata, K. Taira, Global field reconstruction from sparse sensors with Voronoi tessellation-assisted deep learning, Nature Machine Intelligence 3 (2021) 945–951

    2021

  47. [47]

    Zhang, X

    J. Zhang, X. Zhao, Three-dimensional spatiotemporal wind field reconstruction based on physics-informed deep learning, Applied Energy 300 (2021) 117390

    2021

  48. [48]

    J. L. Anderson, An ensemble adjustment Kalman filter for data assimilation, Monthly Weather Review 129 (2001) 2884–2903

    2001

  49. [49]

    Willcox, Unsteady flow sensing and estimation via the gappy proper orthogonal decomposition, Computers & Fluids 35 (2006) 208–226

    K. Willcox, Unsteady flow sensing and estimation via the gappy proper orthogonal decomposition, Computers & Fluids 35 (2006) 208–226

    2006

  50. [50]

    Rohrmeier, W

    K. Rohrmeier, W. Wei, D. Ison, Decoding the Vertiport: Planning for Urban Air Mobility, Journal of Planning Literature 41 (1) (2026) 24–34

    2026

  51. [51]

    S. Lee, N. Cho, Optimal location of urban air mobility (UAM) vertiport using a three-stage geospatial analysis, Future Transportation 5 (2025) 12

    2025

  52. [52]

    Zaheer, S

    M. Zaheer, S. Kottur, S. Ravanbakhsh, B. Póczos, R. Salakhutdinov, A. J. Smola, Deep sets, in: Advances in Neural Information Processing Systems 30 (NeurIPS 2017), 2017, pp. 3391–3401

    2017

  53. [53]

    J. Lee, Y . Lee, J. Kim, A. R. Kosiorek, S. Choi, Y . W. Teh, Set transformer: A framework for attention-based permutation-invariant neural networks, in: International Conference on Machine Learning, 2019, pp. 3744– 3753

    2019

  54. [54]

    C. J. C. Burges, From RankNet to LambdaRank to LambdaMART: An overview, Tech. Rep. MSR-TR-2010-82, Microsoft Research (2010)

    2010

  55. [55]

    Q. Tan, H. Bian, S. Zhong, X. Zhang, Simulation of multi-rotor powered urban aerial mobility noise for envi- ronmental assessment, in: INTER-NOISE and NOISE-CON Congress and Conference Proceedings, 2021

    2021

  56. [56]

    P. Du, X. Liu, X. Fan, J. Wang, Conditional neural field latent diffusion model for generating spatiotemporal turbulence, Nature Communications 15 (2024) 10416. 24

    2024