Pith. sign in

REVIEW 4 major objections 5 minor 40 references

GEOID-Flood: A Large-Scale Multi-Modal Benchmark Dataset for Flood Segmentation

T0 review · 4 major / 5 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read A new benchmark combining bi-temporal radar, co-registered optical imagery, elevation, and permanent-water labels shows that models trained on it transfer to unseen 2026 floods better than four existing datasets.

desk verdict GEOID-Flood is a valuable, well-documented dataset that deserves refereeing, but the headline transfer numbers are conditional on an unaudited permanent-water layer and missing error bars. read the letter →

arxiv 2608.02315 v1 pith:2QA26NU6 submitted 2026-08-03 cs.CV

Gaetano Chiriaco , Luca Barco , Andrea Bragagnolo , Claudio Rossi , Edoardo Arnaudo This is my paper · ORCID
classification cs.CV
keywords floodsegmentationbenchmarkdatasetSentinel-1SARSentinel-2opticalgeospatialfoundationmodelspermanentwatermulti-modalfusiontransfergeneralization
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper introduces GEOID-Flood, a flood-segmentation benchmark built from emergency-management activations spanning 219 flood events in 65 countries over a decade. Each tile pairs pre- and post-event Sentinel-1 radar (in two processing formats), a pre-event Sentinel-2 optical composite, elevation data, and a three-class label separating background, permanent water, and flooded water. The central claim is that this combination, especially the dedicated permanent-water layer, makes models trained on it generalize to unseen floods better than models trained on any of four existing public flood datasets. The paper also finds that geospatial foundation models give only a modest advantage over conventional encoders, and that end-to-end optical–radar fusion with fine-tuning resolves transient flooding best. This matters because flood mapping needs benchmarks that measure whether representations transfer across regions and sensors, not just within a single event or region.

What carries the argument

The load-bearing component is the permanent-water layer. Because open water looks similar to flooded water in a single radar acquisition, the benchmark's three-class labels depend on a binary permanent-water mask produced by a lightweight 41K-parameter decoder trained on 788 labeled surface-water tiles from 2019, using annual geospatial embedding fields, and applied globally to all 219 events. The annual aggregation of embeddings is what makes permanent water distinguishable from transient flood signals; this layer is claimed to be cleaner at 10 m than the 30 m static water product that earlier datasets rely on, which misses narrow rivers and small water bodies. Without this layer, the flood

What would settle it

Compare the dataset's permanent-water masks against independent manual delineations of rivers and reservoirs for a random sample of non-European test events; if systematic disagreement exceeds the flood-IoU gap between GEOID-Flood and the next-best dataset (0.590 vs 0.568), the three-class labels and the transfer conclusions built on them would be undermined.

Watch

Extended reading notes

Core claim

The central discovery is a dataset construction and evaluation result: separating permanent water from transient flood water by deriving a permanent-water layer from annual geospatial embeddings, rather than using coarser static water products, yields training labels that transfer. Concretely, a U-Net with a fixed encoder, trained and fine-tuned on GEOID-Flood, reaches 0.590 frozen / 0.601 fine-tuned IoU on flooded water for 2026 events never seen in training, versus 0.568 / 0.544 for the next-best existing dataset under a common reprocessing and evaluation protocol. The flood-label gain is not simply inherited from the binary water gain, since the held-out permanent-water labels share the d

Load-bearing premise

The permanent-water mask is generated by a model trained on 788 tiles from 2019 and assumed to generalize to every flood area and year; for events before 2017, the nearest available embedding year is used on the assumption that permanent water is stable over multi-year periods.

Editorial extensions

If this is right

  • If the transfer result holds, flood-mapping models trained on this benchmark will be a stronger starting point for rapid emergency-response mapping than current public options.
  • The finding that pre-event optical context plus post-event SAR, fused and fine-tuned, best resolves transient flooding suggests operational pipelines should prioritize adding a clear pre-event optical composite even when post-event optical imagery is cloudy.
  • The modest gap between foundation models and conventional encoders implies that, for water-body segmentation, architecture and training protocol matter more than the choice of pretrained backbone.
  • The dataset's event-level splits and temporally disjoint held-out set provide a protocol that future flood benchmarks can adopt to avoid spatial leakage.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the permanent-water prior is the source of the transfer gain, the same pipeline could be applied to other transient-delineation tasks (landslides, burned area, snowmelt) where a stable pre-event class must be separated from a transient event class.
  • The paper's claim that elevation adds no measurable gain is tested only with raw DEM; a hydrologically conditioned terrain index (such as height above nearest drainage) might still help, since the paper leaves that input unexploited.
  • Because the held-out permanent-water labels share the benchmark's derivation, the cleanest test of the transfer claim is the flooded-water metric; future work could re-score the held-out set with independently hand-corrected permanent water to disentangle label generation from learning.
  • The 140-of-219 European skew suggests the cross-region generalization claim is strongest within Europe; deliberately over-sampling non-European events in the training split would provide a sharper test.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. GEOID-Flood introduces a large-scale flood segmentation benchmark derived from Copernicus Emergency Management Service (CEMS) activations: 219 events across 65 countries, more than 14,000 1024x1024 tiles with co-registered pre/post Sentinel-1 (GRD and RTC), pre-event Sentinel-2, DEM, and three-class labels (background, permanent water, flooded water). The authors describe a reproducible five-stage construction pipeline, event-level splits with a temporally disjoint 2026 held-out set, and a permanent-water layer generated from AlphaEarth Foundation (AEF) embeddings via a lightweight AEF-MLP model. They benchmark foundation models against conventional encoders under single-image, paired, and fusion scenarios, and evaluate cross-dataset generalization to unseen 2026 events. The paper reports three findings: foundation models offer a consistent but modest advantage; optical-SAR fusion with finetuning best resolves transient flooding; and models trained on GEOID-Flood transfer better than models trained on existing flood datasets. The claims are grounded in extensive tables and a shared protocol, and limitations are acknowledged in the conclusion.

Significance. If the results hold up, GEOID-Flood would be a substantial community resource: it is the largest flood benchmark by event count and area, jointly provides bi-temporal SAR and optical imagery at 10 m, uses event-level splits to avoid leakage, and ships code and dataset. The paper also contributes a reproducible benchmark protocol and a careful cross-dataset comparison. However, the headline comparative claims currently rest on single-run metrics and on a permanent-water layer that is not validated on the benchmark itself; moreover, the cross-dataset comparison shares label construction between training and held-out sets. The significance is therefore conditional on additional validation, but the resource itself is valuable and the limitations are mostly addressable within the manuscript's scope.

major comments (4)
  1. [Tables 2, 3, 5] All metrics in Tables 2, 3, and 5 are single-run; Table 4 is the only table with error bars. Conclusions such as "foundation models offer a consistent but modest advantage" and "optical–SAR fusion ... best resolves transient flooding" are based on differences as small as 0.011–0.021 IoU (e.g., Table 2: TerraMind-L FT IoU_bin 0.884 vs. Swin-T 0.873; Table 3: early fusion FT S2→S1 IoU_flood 0.521 vs. mid fusion FT S1+S2→S1 0.513; Table 5: GEOID-Flood frozen IoU_flood 0.590 vs. Kuro Siwo 0.568). Without multiple seeds or confidence intervals these orderings may be within training noise. Please report mean±std over at least 3 seeds for the primary metrics in the main tables, or justify why a fixed seed suffices.
  2. [Sec. B.1–B.3, Table 6] The permanent-water layer is the linchpin of the three-class labels (Sec. 3.2) and the Table 5 transfer claim. AEF-MLP (41K params) is trained on 788 ESW tiles from 2019 and applied globally to all 219 events/65 countries, but the paper reports no evaluation of this model on GEOID-Flood tiles; Table 6 gives only ESW test F1=0.963. If AEF embeddings fail on an AoI (urban canals, arid ephemeral water, turbid/frozen water), the permanent mask changes the flooded-vs-permanent boundary in every downstream metric. Please validate AEF-MLP on a stratified sample of GEOID-Flood tiles (with per-continent/event breakdown), and/or run a sensitivity analysis of Table 3/5 metrics to perturbations of the permanent mask.
  3. [Sec. B.1/B.4 and Sec. 3.3] The held-out set (Sec. 3.3) consists of February–March 2026 events, but B.1 states AEF embeddings are available for 2017–2025. B.4 says the embedding for the year of the flood event is used, and discusses only the pre-2017 proxy. The paper never states which embedding year produced permanent water for the 2026 held-out tiles. If 2025 embeddings were used, the "permanent water is stable over multi-year periods" assumption is doing unquantified work for the headline transfer numbers. Please specify the exact embedding year(s) for the held-out set and provide evidence for multi-year stability (e.g., agreement of AEF-MLP masks across consecutive years on the same AoIs).
  4. [Sec. 3.2 and Sec. 5.5, Table 5] The paper states in Sec. 5.5 that the binary-water lead is "partly expected" because held-out permanent-water labels share GEOID-Flood's derivation, and that flood metrics are "independently derived from CEMS." This is not fully accurate: label composition (Sec. 3.2) merges CEMS flood polygons with the AEF permanent-water layer, so a permanent-water pixel inside a CEMS flood polygon is labeled permanent, not flooded. Thus flood labels are also entangled with the AEF prior. Models trained on GEOID-Flood learn this prior; models trained on other datasets do not. Please re-run Table 5 with controls: e.g., exclude pixels within a buffer of AEF permanent water, replace held-out permanent water with an external product (JRC-GSW), or report flood IoU only on pixels away from permanent-water boundaries.
minor comments (5)
  1. [Sec. 3.2] The pipeline says labels are "manually inspected and corrected where necessary," but no annotator details, number of annotators, or inter-annotator agreement are given. Please add a quality-control subsection with statistics on how many tiles/AoIs required correction.
  2. [Table 1] GEOID-Flood's temporal coverage is listed as 2016–2026 while AEF coverage is 2017–2025. Clarify how 2016 and 2026 events are handled in the permanent-water layer, since this is directly relevant to the embedding-year concern in B.4.
  3. [Fig. 2] Given Europe dominates (140 of 219 events), provide per-split, per-continent tile counts to substantiate the claim that non-European regions are "well represented across splits."
  4. [Sec. A.3 / Sec. 5.5] For the cross-dataset comparison, specify whether the same early-stopping criterion and validation protocol are used for the external datasets. Dataset sizes and label distributions differ substantially, which can affect the finetuning comparison.
  5. [Sec. 5.5] WorldFloods v2 is optical-derived and matched to the temporally closest Sentinel-1 scene; state how many pairs were dropped due to the two-day cutoff and whether this changes the composition of the training set.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: GEOID-Flood is an empirical benchmark; the acknowledged shared derivation of permanent-water labels is a stated limitation, not a construction-level reduction.

full rationale

The paper's central claims are empirical benchmark results, not a derivation chain. GEOID-Flood's labels combine manually inspected CEMS flood polygons with a permanent-water layer produced by AEF-MLP (Sec. B.2-B.4). AEF-MLP is trained on the external Earth Surface Water (ESW) dataset and evaluated on ESW test tiles (Table 6), so its use in label construction is an externally sourced component, not a parameter fitted to the benchmark's own predictions. The cross-dataset transfer claim (RQ4, Table 5) is tested against four external datasets (Kuro Siwo, MMFlood, WorldFloods v2, Sen1Floods11) on a temporally disjoint 2026 held-out set; the comparison does not reduce to any equation of the paper. The only self-referential element is that the held-out permanent-water labels share GEOID-Flood's derivation. The authors explicitly flag this: 'Since the held-out permanent water labels share GEOID-Flood's derivation, the binary water lead is partly expected; the flood metrics, independently derived from CEMS, confirm that scale and diversity yield genuine transfer gains' (Sec. 5.5), and the conclusion anchors the cross-dataset claim on binary water delineation. MMFlood, co-authored by two of the present authors, appears only as an external baseline, not as evidence for a claimed derivation. The 2026 embedding-year gap (AEF coverage stated as 2017-2025, Sec. B.1, vs. held-out events in February-March 2026, Sec. 3.3) is an unquantified limitation but not a circularity. No fitted parameter is renamed as a prediction, no result is forced by self-citation, and no equation equates an output to an input by construction.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The central claims rest on two fitted components: the AEF-MLP permanent-water decoder and its binarization thresholds, plus the reliability of CEMS delineations and AlphaEarth embeddings. These are disclosed but not independently validated on GEOID-Flood. No new physical or postulated entities are introduced; GEOID-Flood is a compiled dataset, not a new force, particle, or conserved quantity.

free parameters (3)
  • AEF-MLP permanent-water decoder weights = 41,281 parameters trained on 788 ESW tiles; ESW test F1 = 0.963
    This model produces the permanent-water class for every GEOID-Flood label; any error in it propagates directly into the three-class segmentation labels and into the transfer claims. Sec. B.3.
  • Permanent-water hysteresis thresholds = 0.5 seed probability, 0.3 extension probability
    Hand-chosen thresholds used to binarize AEF-MLP output into a permanent-water mask; they affect narrow-river recall and the permanent/flood boundary. Sec. B.4.
  • Sentinel-1 GRD normalization statistics = muVV=-12.6 dB, sigmaVV=5.2; muVH=-20.3 dB, sigmaVH=5.9
    Fitted to the GEOID-Flood training set and used to normalize inputs for most encoders. Because the benchmark compares many encoders, normalization choice could influence the modest ranking differences. Sec. A.3.
assumptions (6)
  • domain assumption CEMS Rapid Mapping flood delineations are reliable enough to serve as ground truth after manual inspection/correction.
    All flood labels are rasterized from CEMS polygons (Sec. 3.1); if CEMS systematically mislabels flood extent, every benchmark number inherits that error.
  • domain assumption AlphaEarth annual embeddings represent permanent water stably across regions and years, so a model trained on 2019 ESW scenes transfers globally.
    The AEF-MLP model is trained on 788 ESW tiles, mostly 2019 Sentinel-2 scenes, and applied to AoIs worldwide (Sec. B.1-B.4). Global transferability is asserted, not measured on GEOID-Flood itself.
  • domain assumption Permanent water bodies are stable over multi-year periods, so using the earliest available AEF year for pre-2017 events introduces negligible error.
    Stated in Sec. B.4; if false, permanent-water labels for floods before 2017 could be wrong, affecting the flooded-water class.
  • domain assumption Event-level partitioning with adjacent/overlapping AoIs constrained to the same split prevents spatial leakage.
    Sec. 3.3 describes this split design. If spatially contiguous tiles from the same flood event still share correlated errors, test scores may be optimistic.
  • domain assumption Quality-filtered tiles are missing at random and do not bias the benchmark toward easy cases.
    Sec. 3.2 discards boxes with missing/partial modalities, excessive cloud, or imagery inconsistent with labels; no analysis of the 278 dropped event-AoI pairs is provided, so flash-flood or cloudy-event cases may be under-represented.
  • domain assumption Reprocessing external datasets through a common Sentinel-1 RTC pipeline makes their labels comparable to GEOID-Flood's.
    In Sec. 5.5, external datasets are re-downloaded and reprocessed, and WorldFloods v2 is paired with the closest Sentinel-1 scene within two days. The transfer comparison assumes remaining differences are due to training-set quality rather than label or modality mismatch.

how reviews work

0 comments
Cite this review

Pith. "Pith review of GEOID-Flood: A Large-Scale Multi-Modal Benchmark Dataset for Flood Segmentation." pith.science (2026). https://pith.science/paper/2QA26NU6

@misc{pith2026260802315,
  author       = {Pith},
  title        = {Pith review of: GEOID-Flood: A Large-Scale Multi-Modal Benchmark Dataset for Flood Segmentation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2QA26NU6}},
  note         = {Machine review of arXiv:2608.02315}
}
read the original abstract

Geospatial foundation models aim to learn representations that transfer across regions and sensors, yet evaluating them on specific tasks requires large, high-quality, multi-modal benchmarks that measure how well such models extract value from data. Concerning flood mapping, existing datasets rarely combine bi-temporal SAR and co-registered optical imagery at scale, leaving the value of foundation models for this downstream task largely untested. We introduce GEOID-Flood, a large-scale multi-modal flood segmentation benchmark, derived from Copernicus Emergency Management Service activations, spanning 219 events across 65 countries over ten years. The dataset provides more than 14,000 tiles with co-registered pre- and post-event Sentinel-1, in GRD and RTC format, pre-event Sentinel-2 composite, and DEM, including manually validated labels that separate background from permanent water and flooded water. Using this benchmark, we evaluate foundation models against conventional encoders across single-image, multi-temporal, and multi-modal protocols. We report three main findings: foundation models offer a consistent but modest advantage; optical-SAR fusion with finetuning best resolves transient flooding; and models trained on GEOID-Flood transfer to unseen events better than those trained on existing datasets. Dataset and code available at https://github.com/links-ads/geoid-flood.

Figures

Figures reproduced from arXiv: 2608.02315 by the authors.

Figure 1
Figure 1. Representative tiles from three GEOID-Flood events. From left to right: pre- and post-event Sentinel-1 GRD and RTC (VV/VH format), pre-event Sentinel-2 (RGB), DEM, and label. Flooded water is shown in cyan, permanent water in blue, invalid pixels in gray. cloud cover, at night, and across wide geographic extents. Synthetic Aperture Radar (SAR) meets these requirements, but open water is intrinsically ambigu￾ous in S… view at source ↗
Figure 2
Figure 2. Global distribution of GEOID-Flood AoIs, colored by split assignment (train/validation/test). The inset enlarges Europe, where touching AoIs share a split to prevent boundary leakage. 3.3 Dataset statistics GEOID-Flood covers 219 flood events spanning January 2016 to March 2026, of which the most recent form a temporally disjoint held-out set reserved for cross￾dataset experiments. The dataset spans 65 countries acr… view at source ↗
Figure 3
Figure 3. Visual comparison of GEOID-Flood against other popular datasets on a shared area. First row: available labels, second row: corresponding post-event modality of each dataset. construction, from our own splits and all the other datasets mentioned in Sec. 2. The held-out set follows the same construction pipeline and modality structure as the main dataset, and is used exclusively for the cross-dataset experiments in Se… view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Qualitative comparison on three GEOID-Flood test events (rows) (Sec. 4.2), using five different encoders, finetuned under scenario (i) [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: Qualitative comparison on three GEOID-Flood test events (rows) (Sec. 4.2) across the scenarios of Tab. 3: the single-image baseline (i, TerraMind-B on S1), the paired double-pass model (ii), and early- and mid-fusion (iii) [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: Qualitative permanent water delineation on an ESW test tile. Rows correspond to the four embedding models (TESSERA and AEF, each with a linear probe and an MLP decoder); columns show the Sentinel-2 RGB composite, a PCA projection of the annual embedding, the ground-tru…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

40 extracted references · 3 linked inside Pith

  1. [1]

    Bandara, W.G.C., Patel, V.M.: A transformer-based Siamese network for change detection.In:IGARSS2022-2022IEEEInternationalGeoscienceandRemoteSens- ing Symposium. pp. 207–210. IEEE (2022)

    2022

  2. [2]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    Bastani, F., Wolters, P., Gupta, R., Ferdinando, J., Kembhavi, A.: SatlasPretrain: A large-scale dataset for remote sensing image understanding. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 16772–16782 (2023)

    2023

  3. [3]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recog- nition workshops

    Bonafilia, D., Tellman, B., Anderson, T., Issenberg, E.: Sen1Floods11: A georefer- enced dataset to train and test deep learning flood algorithms for Sentinel-1. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recog- nition workshops. pp. 210–211 (2020)

    2020

  4. [4]

    A global multi-temporal satellite dataset for rapid flood mapping

    Bountos, N.I., Sdraka, M., Zavras, A., Karasante, I., Karavias, A., Herekakis, T., Thanasou, A., Michail, D., Papoutsis, I.: Kuro Siwo: 33 billion m2 under the water. A global multi-temporal satellite dataset for rapid flood mapping. In: Advances in Neural Information Processing Systems (NeurIPS), Datasets and Benchmarks Track (2024)

    2024

  5. [5]

    arXiv preprint arXiv:2507.22291 (2025),https://arxiv.org/abs/2507.22291

    Brown, C.F., Kazmierski, M.R., Pasquarella, V.J., Rucklidge, W.J., Samsikova, M., Zhang, C., Shelhamer, E., Lahera, E., Wiles, O., Ilyushchenko, S., Gorelick, N., Zhang, L.L., Alj, S., Schechter, E., Askay, S., Guinan, O., Moore, R., Boukouvalas, A., Kohli, P.: AlphaEarth foundations: An embedding field model for accurate and efficient global mapping from...

    arXiv 2025

  6. [6]

    In: Proceedings of the European conference on computer vision (ECCV)

    Chen, L.C., Zhu, Y., Papandreou, G., Schroff, F., Adam, H.: Encoder-decoder with atrous separable convolution for semantic image segmentation. In: Proceedings of the European conference on computer vision (ECCV). pp. 801–818 (2018)

    2018

  7. [7]

    Directorate Space, Security and Migration, European Commission Joint Research Centre (EC JRC) (2012–2025),https://mapping

    Copernicus Emergency Management Service: Copernicus emergency management service on-demand mapping. Directorate Space, Security and Migration, European Commission Joint Research Centre (EC JRC) (2012–2025),https://mapping. emergency.copernicus.eu/,©European Union, 2012–2025

    2012

  8. [8]

    CRED, UNDRR: The human cost of disasters: An overview of the last 20 years (2000–2019). Tech. rep., Centre for Research on the Epidemiology of Disasters (CRED) and UN Office for Disaster Risk Reduction (UNDRR), Geneva (2020) 16 G. Chiriaco et al

    2000

Show all 40 references
  1. [9]

    Doerksen, K., Kerner, H.: Earthshift: a benchmark for measuring robustness to real-world distribution shifts in earth observation (2026),https://arxiv.org/ abs/2605.29330

    2026 arXiv

  2. [10]

    IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing15, 2341–2356 (2022)

    Drakonakis, G.I., Tsagkatakis, G., Fotiadou, K., Tsakalides, P.: OmbriaNet— supervised flood mapping via convolutional neural networks using multitemporal Sentinel-1 and Sentinel-2 data fusion. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing15,...

    2022

  3. [11]

    OpenTopography (2021).https://doi.org/10.5270/ESA-c5d3d65

    European Space Agency, Sinergise: Copernicus global digital elevation model. OpenTopography (2021).https://doi.org/10.5270/ESA-c5d3d65

    2021 doi

  4. [12]

    In: Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition (CVPR) (2026),https: //arxiv.org/abs/2506.20380

    Feng, Z., Atzberger, C., Jaffer, S., Knezevic, J., Sormunen, S., Young, R., Li- saius, M.C., Immitzer, M., Jackson, T., Ball, J., Coomes, D.A., Madhavapeddy, A., Blake, A., Keshav, S.: TESSERA: Temporal embeddings of surface spectra for Earth representation and analysis. In: P...

    2026 arXiv

  5. [13]

    In: IEEE International Geoscience and Remote Sensing Symposium (IGARSS) (2025)

    Gomes, C., Blumenstiel, B., Almeida, J.L.d.S., de Oliveira, P.H., Fraccaro, P., Es- cofet, F.M., Szwarcman, D., Simumba, N., Kienzler, R., Zadrozny, B.: TerraTorch: The geospatial foundation models toolkit. In: IEEE International Geoscience and Remote Sensing Symposium (IGARSS) (2025)

    2025

  6. [14]

    He,K.,Zhang,X.,Ren,S.,Sun,J.:Deepresiduallearningforimagerecognition.In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. pp. 770–778 (2016)

    2016

  7. [15]

    International Journal of Applied Earth Observation and Geoinformation117, 103197 (2023)

    He, X., Zhang, S., Xue, B., Zhao, T., Wu, T.: Cross-modal change detection flood extraction based on convolutional neural network. International Journal of Applied Earth Observation and Geoinformation117, 103197 (2023)

    2023

  8. [16]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Herzog, H., Bastani, F., Zhang, Y., Tseng, G., Redmon, J., Sablon, H., Park, R., Morrison, J., Buraczynski, A., Farley, K., et al.: Olmoearth: Stable latent image modeling for multimodal earth observation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Patte...

    2026

  9. [17]

    Interagency Implementation and Advanced Concepts Task Force: ETCI 2021 com- petition on flood detection. Tech. rep., NASA (2021)

    2021

  10. [18]

    Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change

    IPCC: Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press (2021)

    2021

  11. [19]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    Jakubik, J., Yang, F., Blumenstiel, B., Scheurer, E., Sedona, R., Maurogiovanni, S., Bosmans, J., Dionelis, N., Marsocci, V., Kopp, N., et al.: TerraMind: Large-scale generative multimodality for Earth observation. In: Proceedings of the IEEE/CVF International Conference on Co...

    2025

  12. [20]

    Sensors18(9), 2915 (2018)

    Kang, W., Xiang, Y., Wang, F., Wan, L., You, H.: Flood detection in Gaofen-3 SAR images via fully convolutional networks. Sensors18(9), 2915 (2018)

    2018

  13. [21]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., Guo, B.: Swin Trans- former: Hierarchical vision transformer using shifted windows. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 10012–10022 (2021)

    2021

  14. [22]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Liu, Z., Mao, H., Wu, C.Y., Feichtenhofer, C., Darrell, T., Xie, S.: A ConvNet for the 2020s. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 11976–11986 (2022)

    2022

  15. [23]

    In: International Conference on Learning Representations (2019)

    Loshchilov, I., Hutter, F.: Decoupled weight decay regularization. In: International Conference on Learning Representations (2019)

    2019

  16. [24]

    Zenodo (2021).https://doi.org/10.5281/ zenodo.5205674 GEOID-Flood: Multi-Modal Flood Segmentation Benchmark 17

    Luo, X.: Earth surface water dataset. Zenodo (2021).https://doi.org/10.5281/ zenodo.5205674 GEOID-Flood: Multi-Modal Flood Segmentation Benchmark 17

    2021

  17. [25]

    Natural Hazards and Earth System Sciences9(2), 303–314 (2009)

    Martinis, S., Twele, A., Voigt, S.: Towards operational near real-time flood de- tection using a split-based automatic thresholding procedure on high resolution TerraSAR-X data. Natural Hazards and Earth System Sciences9(2), 303–314 (2009)

    2009

  18. [26]

    Scientific Reports11, 7249 (2021)

    Mateo-García, G., Veitch-Michaelis, J., Smith, L., Oprea, S.V., Schumann, G., Gal, Y., Baydin, A.G., Backes, D.: Towards global flood mapping onboard low cost satellites with machine learning. Scientific Reports11, 7249 (2021)

    2021

  19. [27]

    IEEE Access10, 96774–96787 (2022)

    Montello, F., Arnaudo, E., Rossi, C.: MMFlood: A multimodal dataset for flood delineation from satellite imagery. IEEE Access10, 96774–96787 (2022)

    2022

  20. [28]

    Big Earth Data9(3), 412–438 (2025)

    Notarangelo, N., Wirion, C., van Winsen, F.: STURM-Flood: a curated dataset for deep learning-based flood extent mapping leveraging Sentinel-1 and Sentinel-2 imagery. Big Earth Data9(3), 412–438 (2025)

    2025

  21. [29]

    Nature540(7633), 418–422 (2016)

    Pekel, J.F., Cottam, A., Gorelick, N., Belward, A.S.: High-resolution mapping of global surface water and its long-term changes. Nature540(7633), 418–422 (2016)

    2016

  22. [30]

    Scientific Reports13(1), 20316 (2023)

    Portalés-Julià, E., Mateo-García, G., Purcell, C., Gómez-Chova, L.: Global flood extent segmentation in optical satellite images. Scientific Reports13(1), 20316 (2023)

    2023

  23. [31]

    IEEE Access9, 89644–89654 (2021)

    Rahnemoonfar, M., Chowdhury, T., Sarkar, A., Varshney, D., Yari, M., Murphy, R.R.: FloodNet: A high resolution aerial imagery dataset for post flood scene un- derstanding. IEEE Access9, 89644–89654 (2021)

    2021

  24. [32]

    arXiv preprint arXiv:2603.02386 (2026),https://arxiv.org/abs/2603

    Robinson, C., Lehmann, N., Stewart, A.J., Ekim, B., Fang, H., Corley, I.A., Cordeiro, M.: Advancing Earth observation through machine learning: A TorchGeo tutorial. arXiv preprint arXiv:2603.02386 (2026),https://arxiv.org/abs/2603. 02386

    2026

  25. [33]

    In: International Conference on Medical image computing and computer-assisted intervention

    Ronneberger, O., Fischer, P., Brox, T.: U-Net: Convolutional networks for biomed- ical image segmentation. In: International Conference on Medical image computing and computer-assisted intervention. pp. 234–241. Springer (2015)

    2015

  26. [34]

    ISPRS Journal of Photogrammetry and Remote Sensing212, 440–453 (2024)

    Saleh, T., Weng, X., Holail, S., Hao, C., Xia, G.S.: DAM-Net: Global flood de- tection from SAR imagery using differential attention metric-based vision trans- formers. ISPRS Journal of Photogrammetry and Remote Sensing212, 440–453 (2024)

    2024

  27. [35]

    Satellite data access and processing platform,https:// www.sentinel-hub.com

    Sinergise: Sentinel Hub. Satellite data access and processing platform,https:// www.sentinel-hub.com

  28. [36]

    International Journal of Remote Sensing37(13), 2990–3004 (2016)

    Twele, A., Cao, W., Plank, S., Martinis, S.: Sentinel-1-based flood mapping: a fully automated processing chain. International Journal of Remote Sensing37(13), 2990–3004 (2016)

    2016

  29. [37]

    IEEE Geoscience and Re- mote Sensing Magazine11(3), 98–106 (2023)

    Wang, Y., Braham, N.A.A., Xiong, Z., Liu, C., Albrecht, C.M., Zhu, X.X.: SSL4EO-S12: A large-scale multimodal, multitemporal dataset for self-supervised learning in Earth observation [Software and Data Sets]. IEEE Geoscience and Re- mote Sensing Magazine11(3), 98–106 (2023)

    2023

  30. [38]

    Remote Sensing of Environment322, 114694 (2025)

    Wright, N., Duncan, J.M., Callow, J.N., Thompson, S.E., George, R.J.: Training sensor-agnostic deep learning models for remote sensing: Achieving state-of-the-art cloud and cloud shadow identification with OmniCloudMask. Remote Sensing of Environment322, 114694 (2025)

    2025

  31. [39]

    arXiv preprint arXiv:2403.15356 (2024)

    Xiong, Z., Wang, Y., Zhang, F., Stewart, A.J., Hanna, J., Borth, D., Papoutsis, I., Le Saux, B., Camps-Valls, G., Zhu, X.X.: Neural plasticity-inspired multimodal foundation model for Earth observation. arXiv preprint arXiv:2403.15356 (2024)

    2024

  32. [40]

    International Journal of Applied Earth Observation and Geoinformation116, 103132 (2023) 18 G

    Zhao, B., Sui, H., Liu, J.: Siam-DWENet: Flood inundation detection for SAR im- agery using a cross-task transfer Siamese network. International Journal of Applied Earth Observation and Geoinformation116, 103132 (2023) 18 G. Chiriaco et al. A Models and Optimization This appen...

    2023

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.