# arXiv physics.ao-ph: RadarFill fills blind zones in 3-D weather radar mosaics [Canonical reading](https://escapemeridian.com/readings/arxiv-physics-ao-ph-radarfill-fills-blind-zones-in-3-d-weather-dc2652a6) Xin Huang and coauthors present RadarFill, a conditional diffusion framework for filling systematic blind zones in three-dimensional weather radar reflectivity mosaics. Ground-based radar can miss reflectivity at low altitudes, where complete observations for direct verification are unavailable. RadarFill uses partially observed reflectivity and an observation mask, while preserving available measurements as it generates estimates for the gaps. The authors report tests on 4,384 held-out test times and representative convective and typhoon cases, where RadarFill recovered principal precipitation structures and moderate-to-strong echoes under extensive low-level missingness. For real blind zones in Fujian Province, they compared its estimates with 90 matched volume scans from an independent phased-array radar; RadarFill had lower reflectivity errors than conventional interpolation within the principal 1–2 km low-level blind zones. They also report that improvements in two controlled nowcasting experiments persisted into subsequent forecasts, especially at low altitudes and short lead times. These results offer complementary evidence from controlled synthetic missingness, independent radar observations and downstream nowcasting. The authors say they support the potential of observation-constrained conditional generation for filling real 3-D radar blind zones. Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims. ## Original source arXiv physics.ao-ph · Article Original source date: Nov 3, 2026 (2026-11-03). Date basis: Date basis not provided. Added to Escape Meridian: 2026-11-10T05:31:49.276Z. First observed: 2026-11-04T05:12:31.307Z. Retrieved: 2026-11-04T05:30:48.781Z. - [RadarFill: Observation-Constrained Gap Filling of Systematic Blind Zones in 3-D Weather Radar Reflectivity Mosaics](https://arxiv.org/abs/2611.04768) ## Limits The source says low-altitude missing reflectivity lacks corresponding complete observations for direct verification. The real blind-zone evaluation described in the abstract is in Fujian Province and uses 90 matched volume scans from an independent phased-array radar. The abstract describes controlled synthetic missingness and two controlled nowcasting experiments; it does not establish performance beyond the reported evaluations. ## Open question How well would RadarFill perform on real blind zones beyond the Fujian evaluation and the 90 matched volume scans? ## Source-stated The authors present RadarFill as a conditional diffusion framework that fills gaps in three-dimensional radar reflectivity mosaics while preserving available measurements. > conditions on partially observed reflectivity and an observation mask, enforces available observations throughout reverse sampling, and preserves valid measurements in the final field. - [Supporting source](https://arxiv.org/abs/2611.04768) ## Source-stated In tests on held-out times and representative storm cases, the authors report that RadarFill recovered principal precipitation structures and moderate-to-strong echoes under extensive low-level missingness. > Across 4,384 held-out test times and representative convective and typhoon cases, RadarFill recovers principal precipitation structures and moderate-to-strong echoes under extensive low-level missingness - [Supporting source](https://arxiv.org/abs/2611.04768) ## Source-stated In the real blind-zone evaluation in Fujian Province, RadarFill had lower reflectivity errors than conventional interpolation within the principal low-level blind zones. > Within the principal 1-2 km low-level blind zones, RadarFill yields lower reflectivity errors than conventional interpolation - [Supporting source](https://arxiv.org/abs/2611.04768) ## Source-stated The authors report that improvements in RadarFill-initialized fields persisted into subsequent forecasts in two controlled nowcasting experiments, with the largest benefits at low altitudes and short lead times. > In two controlled nowcasting experiments, improvements in RadarFill-initialized fields persist into subsequent forecasts, with the largest benefits at low altitudes and short lead times. - [Supporting source](https://arxiv.org/abs/2611.04768) ## Source-stated The authors characterize their results as complementary evidence supporting the potential of observation-constrained conditional generation for gap filling in real 3-D radar blind zones. > These results provide complementary evidence from controlled synthetic missingness, independent radar observations, and downstream nowcasting, supporting the potential of observation-constrained conditional generation for gap filling in real 3-D radar blind zones. - [Supporting source](https://arxiv.org/abs/2611.04768) ## Source and contributor context Huang, Xin, Sun, Haofei, Han, Wei, Yang, Yunfan, Chen, Huaguan, Song, Xingtao, Xiao, Dong · Article author These are credits stated by the source; they do not establish a profile match or authorship of every passage. ## Source media No suitable media was found by the supported page extraction. ## Terms and references Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement. ## arXiv arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed. - [Explore this context](/entities/arxiv) - [Reference](https://arxiv.org/) - [Reference](https://info.arxiv.org/help/api/index.html) ## Editorial provenance Version 1. Publication ID: 6aeba1ac-a1d2-4882-8304-f2e11d84b7b8. Prepared by gpt-6-luna; grounding review by gpt-6-luna. Source check: passed by anthropic/claude-opus-5.5 on 2026-11-10T05:31:49.276Z. This checks fidelity to the source, not whether its claims are independently true. Editorial selection: accept; reasons: material\_development. Observation: e2f19cac5db02535d232d7319813192ed9776ad541cd4f549471dba285866155. Source text hash: 74e771a17ef33aaeec6e50f69dd2500ac60d0fcb18e4cd89ab9d95c7f5104102. ## Revision and source-check history ### Version 1 — 2026-11-10T05:31:49.276Z Xin Huang and coauthors present RadarFill, a conditional diffusion framework for filling systematic blind zones in three-dimensional weather radar reflectivity mosaics. Ground-based radar can miss reflectivity at low altitudes, where complete observations for direct verification are unavailable. RadarFill uses partially observed reflectivity and an observation mask, while preserving available measurements as it generates estimates for the gaps. The authors report tests on 4,384 held-out test times and representative convective and typhoon cases, where RadarFill recovered principal precipitation structures and moderate-to-strong echoes under extensive low-level missingness. For real blind zones in Fujian Province, they compared its estimates with 90 matched volume scans from an independent phased-array radar; RadarFill had lower reflectivity errors than conventional interpolation within the principal 1–2 km low-level blind zones. They also report that improvements in two controlled nowcasting experiments persisted into subsequent forecasts, especially at low altitudes and short lead times. These results offer complementary evidence from controlled synthetic missingness, independent radar observations and downstream nowcasting. The authors say they support the potential of observation-constrained conditional generation for filling real 3-D radar blind zones. ## Original source arXiv physics.ao-ph · Article Original source date: Nov 3, 2026 (2026-11-03). Date basis: Date basis not provided. Added to Escape Meridian: 2026-11-10T05:31:49.276Z. First observed: 2026-11-04T05:12:31.307Z. Retrieved: 2026-11-04T05:30:48.781Z. - [RadarFill: Observation-Constrained Gap Filling of Systematic Blind Zones in 3-D Weather Radar Reflectivity Mosaics](https://arxiv.org/abs/2611.04768) ## Limits The source says low-altitude missing reflectivity lacks corresponding complete observations for direct verification. The real blind-zone evaluation described in the abstract is in Fujian Province and uses 90 matched volume scans from an independent phased-array radar. The abstract describes controlled synthetic missingness and two controlled nowcasting experiments; it does not establish performance beyond the reported evaluations. ## Open question How well would RadarFill perform on real blind zones beyond the Fujian evaluation and the 90 matched volume scans? ## Source-stated The authors present RadarFill as a conditional diffusion framework that fills gaps in three-dimensional radar reflectivity mosaics while preserving available measurements. > conditions on partially observed reflectivity and an observation mask, enforces available observations throughout reverse sampling, and preserves valid measurements in the final field. - [Supporting source](https://arxiv.org/abs/2611.04768) ## Source-stated In tests on held-out times and representative storm cases, the authors report that RadarFill recovered principal precipitation structures and moderate-to-strong echoes under extensive low-level missingness. > Across 4,384 held-out test times and representative convective and typhoon cases, RadarFill recovers principal precipitation structures and moderate-to-strong echoes under extensive low-level missingness - [Supporting source](https://arxiv.org/abs/2611.04768) ## Source-stated In the real blind-zone evaluation in Fujian Province, RadarFill had lower reflectivity errors than conventional interpolation within the principal low-level blind zones. > Within the principal 1-2 km low-level blind zones, RadarFill yields lower reflectivity errors than conventional interpolation - [Supporting source](https://arxiv.org/abs/2611.04768) ## Source-stated The authors report that improvements in RadarFill-initialized fields persisted into subsequent forecasts in two controlled nowcasting experiments, with the largest benefits at low altitudes and short lead times. > In two controlled nowcasting experiments, improvements in RadarFill-initialized fields persist into subsequent forecasts, with the largest benefits at low altitudes and short lead times. - [Supporting source](https://arxiv.org/abs/2611.04768) ## Source-stated The authors characterize their results as complementary evidence supporting the potential of observation-constrained conditional generation for gap filling in real 3-D radar blind zones. > These results provide complementary evidence from controlled synthetic missingness, independent radar observations, and downstream nowcasting, supporting the potential of observation-constrained conditional generation for gap filling in real 3-D radar blind zones. - [Supporting source](https://arxiv.org/abs/2611.04768) ## Source and contributor context Huang, Xin, Sun, Haofei, Han, Wei, Yang, Yunfan, Chen, Huaguan, Song, Xingtao, Xiao, Dong · Article author These are credits stated by the source; they do not establish a profile match or authorship of every passage. ## Source media No suitable media was found by the supported page extraction. ## Terms and references Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement. ## arXiv arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed. - [Explore this context](/entities/arxiv) - [Reference](https://arxiv.org/) - [Reference](https://info.arxiv.org/help/api/index.html) ## Editorial provenance Version 1. Publication ID: 6aeba1ac-a1d2-4882-8304-f2e11d84b7b8. Prepared by gpt-6-luna; grounding review by gpt-6-luna. Source check: passed by anthropic/claude-opus-5.5 on 2026-11-10T05:31:49.276Z. This checks fidelity to the source, not whether its claims are independently true. Editorial selection: accept; reasons: material\_development. Observation: e2f19cac5db02535d232d7319813192ed9776ad541cd4f549471dba285866155. Source text hash: 74e771a17ef33aaeec6e50f69dd2500ac60d0fcb18e4cd89ab9d95c7f5104102. ### passed — 2026-11-10T05:31:49.276Z — anthropic/claude-opus-5.5 Publication: 6aeba1ac-a1d2-4882-8304-f2e11d84b7b8 attribution — pass: The draft correctly credits Xin Huang and coauthors, as listed on the arXiv physics.ao-ph page, and attributes results with phrasing such as 'the authors report.' One small drift: the third paragraph opens 'These results offer complementary evidence' in an unattributed voice. The next sentence restores attribution ('The authors say'), so this does not mislead. Smallest repair: 'The authors describe these results as complementary evidence...' > Authors:Xin Huang, Haofei Sun, Wei Han, Yunfan Yang, Huaguan Chen, Xingtao Song, Dong Xiao summary — pass: The draft leads with the development: a model that fills gaps in 3-D radar mosaics where low-altitude coverage is missing. It explains the problem in plain words (radar misses low altitudes, and there are no complete observations to check against). It then gives the evidence and the forecasting consequence. Technical terms such as conditional diffusion, reflectivity and nowcasting are kept exactly, for tooltip treatment rather than paraphrase. The title is a fair description of the paper's aim. The question raises a genuine open issue, generalization beyond Fujian, without leading. > Ground-based weather radar mosaics contain systematic blind zones, especially at low altitudes, where missing reflectivity lacks corresponding complete observations for direct verification. date — pass: originalPublished 2026-11-03 matches the v1 submission date in the arXiv submission history. The feed date of 2026-11-04 is correctly not used. > \[Submitted on 3 Nov 2026\] support — pass: Each claim matches the abstract: the 4,384 held-out test times, convective and typhoon cases, 90 matched phased-array volume scans in Fujian, lower errors than interpolation in the 1-2 km zones, and nowcasting gains at low altitudes and short lead times. The draft keeps the authors' hedged wording ('potential') and adds no benefits or real-world reliability claims beyond the source. The limitations accurately note that performance outside these evaluations is not established. > Within the principal 1-2 km low-level blind zones, RadarFill yields lower reflectivity errors than conventional interpolation, and the gap-filled precipitation structures show broad agreement with independent observations. media — pass: The selected media is the verified arXiv abstract page for this paper, which is the direct original source. No better source media was found. ## Recorded references Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter. No public references recorded here yet. [Reference data (JSON)](/readings/dc2652a67333ac805fc0b155d066bb2f86b88b7a882d2ed2a9ffad11a358af6b/references.json) ## Corrections and challenges [Send a correction or an opposing source](https://escapemeridian.com/corrections/) [JSON record](https://escapemeridian.com/readings/arxiv-physics-ao-ph-radarfill-fills-blind-zones-in-3-d-weather-dc2652a6.json)