Hapi: A Multivariable Land-Surface Transformer for Medium-Range Hydrological Forecasting at Continental Scale

Hong Zhang; John K. Hutchison; Rao Kotamarthi; Jeremy Feinstein; Haiwen Guan; Romit Maulik; Ross M. Alexander; Vijay P. Ramalingam

cs.AISep 19, 2026

Hapi: A Multivariable Land-Surface Transformer for Medium-Range Hydrological Forecasting at Continental Scale

Authors: Hong Zhang, John K. Hutchison, Rao Kotamarthi, Jeremy Feinstein, Haiwen Guan, Romit Maulik, Ross M. Alexander, Vijay P. Ramalingam, +2 more

Organizations: Argonne National Laboratory, Lemont, IL 60439, USA

Abstract

Accurate flood forecasts several days in advance are essential for flood control, water-resource management, and emergency response. A central challenge is to produce high-resolution forecasts across continental domains where hydrological behavior varies widely from place to place. We developed Hapi, a U-Net Swin Transformer that uses fine three-dimensional patches and hierarchical shifted-window attention to forecast river discharge, surface runoff, snow water equivalent, and soil wetness index across the contiguous United States. The model produces medium-range forecasts (24--72~h) at $0.05^{\circ}$ resolution with adaptive task weighting and required only 0.11 seconds for a four-variable 72-h CONUS forecast on one A100 GPU. In a held-out 2024 potential-skill evaluation with ERA5-Land inputs prescribed over the forecast horizon, Hapi achieved the highest F1-score for floods in 20 of 21 comparisons across seven GloFAS return periods and three forecast leads. Independent validation against observed daily discharge at 3{,}881 U.S. Geological Survey gauges showed that Hapi achieved the highest median Nash--Sutcliffe efficiency at every lead, supported by regional-cluster bootstrap intervals. In a matched 24-h comparison of loss formulations, adaptive task balancing produced the lowest discharge errors and the highest F1-score for floods.

Figures & tables

Figure 1: Continental illustration of a Hapi discharge forecast. a , Discharge at initialization on 20 June 2024, represented by the field carried forward in the persistence forecast. b , Hapi’s 72-h discharge forecast valid on 23 June. c , GloFAS reanalysis for the same valid date. d , Correctly predicted exceedances (hits), missed exceedances, and false alarms at each cell’s local 2-year return level; cells with no forecast or reference exceedance are shown in grey. The valid date was selected independently of Hapi as the 2024 test date with the largest number of reference 2-year-threshold exceedances over CONUS. The discharge maps share a logarithmic colour scale.

Role	Source	Variables and units
Prognostic	GloFAS reanalysis	River discharge, m 3 s -1 ; runoff water equivalent, mm; snow water equivalent, mm; soil wetness index, dimensionless
Forcing	ERA5-Land	Total runoff, ro , m; precipitation, tp , m; evaporation, e , m; soil water in layer 1, swvl1 , m 3 m -3 ; solar radiation, ssrd , J m -2 ; 2 m air temperature, t2m , K
Static	DEM; GloFAS v4.0	Surface elevation, m; upstream drainage area, m 2

Table 1: The 12 model inputs, grouped by their role in the autoregressive rollout. ERA5-Land fields come from the Copernicus Climate Change Service Climate Data Store ( Copernicus Climate Change Service, 2019 ) ; upstream drainage area comes from the GloFAS v4.0 auxiliary data ( ECMWF, 2023 ) . DEM denotes digital elevation model.

Figure 2: Hapi architecture. The encoder on the left downsamples through $N$ hierarchical Swin stages, and the symmetric decoder on the right mirrors the first $N-1$ stages after the shared bottleneck. Each Swin block uses 2D rotary position embeddings and gated attention, both described below, while skip connections transfer fine-grained features from matching encoder stages to the decoder. During autoregressive rollout, prognostic channels are predicted and fed back, forcing channels are supplied externally, and static channels remain fixed, as described in Section 3.2 .

		Return period, years
Model	Lead	1.5	2	5	10	20	50	100
persistence	24 h	0.836	0.799	0.709	0.627	0.562	0.496	0.470
GloFAS	24 h	0.773	0.718	0.584	0.473	0.396	0.306	0.264
RiverMamba	24 h	0.728	0.660	0.479	0.347	0.259	0.187	0.145
Hapi	24 h	0.885	0.863	0.786	0.700	0.612	0.524	0.468
persistence	48 h	0.726	0.671	0.544	0.434	0.353	0.275	0.233

Table 2: F1-score for floods on the CONUS land mask in the 2024 test window at 24-, 48-, and 72-h leads and GloFAS return periods from 1.5 to 100 yr. Higher is better; bold denotes the best score for each lead and threshold. Results at the more uncertain 200- and 500-yr thresholds appear in Appendix M .

Figure 3: F1-score difference and false-alarm ratio for floods on the CONUS land mask in the 2024 test window. Panel a shows Hapi’s F1 minus the highest F1 among the three baselines at each lead and return-period threshold. Line color denotes forecast lead; marker shape identifies the strongest baseline at each point. Panels b and c show the false-alarm ratio,

\frac{\mathrm{FP}}{\mathrm{TP}+\mathrm{FP}}

, at the 1.5-yr and 100-yr thresholds, respectively. Color, line style, and marker encode the model, with Hapi shown by thicker lines.

	24 h lead				48 h lead				72 h lead
Model	MAE	RMSE	NSE	KGE	MAE	RMSE	NSE	KGE	MAE	RMSE	NSE	KGE
persistence	2.43	41.64	0.996	0.998	4.55	78.38	0.985	0.992	6.30	109.68	0.970	0.985
GloFAS	2.79	43.29	0.995	0.997	2.94	44.59	0.995	0.997	3.15	46.90	0.995	0.997
RiverMamba	2.11	35.67	0.997	0.997	3.05	54.53	0.993	0.993	3.80	70.46	0.988	0.986
Hapi	1.57	32.30	0.997	0.990	2.94	63.19	0.990	0.980	4.08	89.72	0.980	0.976

Table 3: Discharge forecast skill on the CONUS land mask in the 2024 test window, pooled over space and time. MAE and RMSE are in m 3 s -1 ; NSE and KGE are dimensionless. Bold denotes the best score per column. Per-cell skill is reported in Table 4 .

Lead	Model	Median NSE	Median KGE	Median $\beta$	Hapi wins vs.
					% cells	median $\Delta$ NSE
24 h	Hapi	0.902	0.891	0.973	–	–
	persistence	0.823	0.911	1.000	89.4%	+0.054
	GloFAS	0.698	0.804	0.993	93.0%	+0.142
48 h	Hapi	0.823	0.832	0.967	–	–
	persistence	0.560	0.779	1.000	91.5%	+0.217

Table 4: Per-cell continuous skill on the CONUS land mask in the 2024 test window. NSE, KGE, and the KGE bias ratio

\beta

are computed independently at every grid cell, and their medians over cells are reported, complementing the values pooled over space and time in Table 3 . The last two columns give the percentage of cells at which Hapi’s NSE exceeds that of the comparison model and the median per-cell difference. Scores use 374,297 cells, representing 90.2% of the mask, where all three sources are defined at the relevant lead; the remainder have near-zero target variance, which makes NSE undefined. Bold marks the best NSE and KGE at each lead.

Lead	Model	NSE	KGE	RMSE	PBIAS	$f_{\mathrm{NSE}>0}$	$f_{\mathrm{NSE}>0.5}$
		median	median	m 3 /s	%	%	%
24 h	Hapi	0.160	0.310	19.4	+1.2	59.7	23.2
	GloFAS	0.076	0.273	20.6	+4.8	54.7	20.9
	persistence	0.070	0.271	20.9	+5.3	54.8	18.6
48 h	Hapi	0.165	0.312	19.3	+0.5	59.7	23.4
	GloFAS	0.064	0.271	20.8	+4.7	54.4	20.0

Table 5: Independent validation against USGS daily discharge in the 2024 test window. Of 7,881 candidate gauges, 3,881 meet the matching and coverage criteria in Appendix J . NSE, KGE, RMSE, and percent bias, PBIAS, are medians across gauges. The final columns give the percentages with NSE above 0 and 0.5. Bold marks the best summary in each column.

Figure 4: Discharge at 72-h lead against USGS observations during the 2024 test window. Rows show Missouri, South Atlantic–Gulf, and Pacific Northwest water-resource regions; columns show drainage areas of 500–5,000, 5,000–50,000, and at least 50,000 km 2 . Within each bin, the selected gauge has Hapi NSE nearest the bin median among gauges with at least 300 paired days and finite NSE. The Alabama River bin contains only one eligible gauge. Panel titles give Hapi NSE against observations; selection details are in Appendix J .

					F1 by GloFAS return period, yr
Variant	MAE	RMSE	NSE	KGE	1.5	2	5	10	20	50	100
Full Hapi	1.57	32.30	0.997	0.990	0.885	0.863	0.786	0.700	0.612	0.524	0.468
Discharge-only	2.07	41.14	0.996	0.992	0.869	0.842	0.761	0.672	0.593	0.522	0.481

Table 6: Comparison of four-variable and discharge-only models at 24-h lead over CONUS in the 2024 test window. Training differences are described in Section 4.4 . Full Hapi matches Table 3 . MAE and RMSE are in m 3 /s; other metrics are dimensionless. Bold marks the higher skill in each column.

					F1 by GloFAS return period, yr
Variant	MAE	RMSE	NSE	KGE	1.5	2	5	10	20	50	100
Learned task scales	1.59	30.94	0.998	0.990	0.877	0.856	0.779	0.690	0.601	0.513	0.460
Uniform task weights	1.97	38.71	0.996	0.985	0.830	0.802	0.710	0.607	0.505	0.410	0.345
Fixed task scales	1.64	31.77	0.997	0.987	0.856	0.835	0.748	0.647	0.544	0.445	0.382

Table 7: Loss-weighting ablation at 24-h lead over CONUS in the 2024 test window. The same four-variable architecture and data were used for single-step training on 16 GPUs, with checkpoints selected by minimum validation loss at epochs 952–953. Each loss-weighting variant is represented by one training run under these conditions. MAE and RMSE are in m 3 /s; other metrics are dimensionless. These pretraining checkpoints precede the fine-tuning used in Table 3 .

Variable	Lead	Model	RMSE	MAE	NSE	KGE
Runoff water equivalent, 418,792 cells
	24 h	Hapi	1.8580	0.2850	0.3281	$0.3474$
		persistence	$2.4500$	$0.4363$	$-0.1686$	0.4156
	48 h	Hapi	1.8900	0.2999	0.3107	0.3365
		persistence	$2.6790$	$0.5245$	$-0.3862$	$0.3039$
	72 h	Hapi	1.9070	0.3105	0.2980	0.3327

Table 8: Skill of the non-discharge prognostic variables against reanalysis over the 2024 test window. Hapi and analysis-time persistence use identical finite cell–day pairs on each variable’s channel mask. Runoff and snow RMSE and MAE are in mm of water equivalent; soil wetness, NSE, and KGE are dimensionless. Bold marks the better score for each variable and lead. The screened snow rows exclude three persistent accumulation cells using the rule in Appendix K .

Subset	Lead	Model	1.5-yr	2-yr	5-yr	10-yr	20-yr	50-yr	100-yr
AIFAS	24 h	Hapi	0.895	0.870	0.785	0.690	0.585	0.461	0.366
		RM	0.755	0.681	0.505	0.366	0.282	0.212	0.159
	48 h	Hapi	0.846	0.807	0.687	0.551	0.420	0.259	0.174
		RM	0.722	0.641	0.459	0.321	0.236	0.158	0.101
	72 h	Hapi	0.816	0.767	0.633	0.481	0.344	0.182	0.116
		RM	0.679	0.604	0.420	0.289	0.212	0.131	0.084

Table 9: Per-subset F1-score for floods at all seven primary return periods over CONUS in the 2024 test window. Hapi and RiverMamba, abbreviated RM, use their nominal forecast leads. RiverMamba’s latest discharge input is one day older. Bold marks the higher value using unrounded scores.

		Log-noise amplitude $\sigma$
Lead	Metric	0.00	0.20	0.50	1.00
48 h	F1 @ 1.5-yr	0.833	0.827	0.800	0.709
	F1 @ 5-yr	0.691	0.677	0.602	0.357
	F1 @ 20-yr	0.460	0.442	0.334	0.114
	F1 @ 100-yr	0.269	0.254	0.168	0.039
	NSE	0.990	0.990	0.990	0.987

Table 10: Sensitivity to precipitation perturbations at rollout steps 2–3 over CONUS in the 2024 test window. The zero-preserving multiplier

e^{\sigma\varepsilon}

uses a standardized Gaussian field smoothed with a spatial kernel of width 3 grid cells and a first-order autoregressive temporal coefficient of 0.8. The log-noise amplitude

\sigma

is distinct from the smoothing width. The multiplier is not mean-centered; its nominal mean is

e^{\sigma^{2}/2}

. The 24-h forcing is unchanged, and

\sigma=0

is the unperturbed baseline.

Figure 5: Hurricane Helene discharge forecasts valid on 27 September 2024 over the southern Appalachians, spanning 32.5–38.5 ∘ N and 87.0–78.5 ∘ W, on a common logarithmic scale. The top row shows Hapi and the bottom row GloFAS forecasts, each beside the GloFAS reanalysis reference. The 24-, 48-, and 72-h forecasts were initialized on 26, 25, and 24 September, respectively. The high-flow median ratio is the forecast median divided by the GloFAS reanalysis median over the same 204 cells in the highest 1% of reference discharge; a value of 100% would indicate equal medians. The reference median is 3,547 m 3 s -1 .

System	Spatial units	Params	Leads	Time	Memory	Normalized time
		M			GB	$\mu$ s/unit/lead
Hapi	589,824 grid cells	43.96	3	109.5 ms	1.69	0.062
RiverMamba	6,221,926 river points	4.38	7	13.75 s	13.33	0.316

Table 11: Average GPU-resident inference time over 10 repetitions at a batch size of 1 using 32-bit floating-point precision on one NVIDIA A100-SXM4 GPU, with 80 GB capacity for Hapi and 40 GB for RiverMamba. Normalized time is execution time divided by the number of spatial units and output leads. Memory is peak allocated memory. Data loading, transfer, CONUS extraction, metrics, and file writing are excluded.

Appendix figures & tables17 assets

Supplementary material from the paper’s appendix.

Appendix

Variable	Norm	Skew	Zero %	Max/Med	Rationale
River discharge	log1p	38.9	35%	$7.4\times 10^{5}$	Extreme skew, 6 orders of magnitude
Runoff water equiv.	log1p	10.5	33%	$4.1\times 10^{3}$	Heavy right skew
Snow water equivalent	log1p	118.5	78%	$8.8\times 10^{2}$ †	Extreme skew and seasonal absence
Soil wetness index	identity	$-$ 0.0	0%	2.2	Already bounded in $[0,1]$
Runoff	log1p	11.7	4%	$2.1\times 10^{3}$	Heavy right skew
Total precipitation	log1p	5.8	8%	$3.1\times 10^{3}$	Heavy right skew

Appendix

Table 12: Per-variable normalization choices and descriptive distributional statistics. The model’s normalization parameters were estimated from the 2005–2022 training period only. The descriptive statistics were computed over the full 2005–2024 record (every tenth day) on the

0.05^{\circ}

CONUS grid. The statistics use the 419,682 cells where 2 m temperature is defined. Flood forecasting uses the 414,795-cell discharge mask described in Appendix B . Skewness is the standard biased estimator. Zero % counts exact zeros, and Max/Med is the maximum divided by the absolute median. † Snow depth has a zero median, so its Max/Med ratio uses the median of nonzero values.

Figure 6: Raw distributions of the 12 sampled input variables over the 419,682-cell 2 m-temperature mask. This illustration uses a shorter window than the descriptive statistics in Table 12 . Each panel shows skewness, zero fraction, and the maximum divided by the absolute median. Med 0 denotes the median of nonzero values when the overall median is zero. Variables normalized with log1p use a logarithmic

x

-axis to show the tail structure. Red and orange lines mark the median and mean. Large separation indicates high skewness.

Hyperparameter	Values
Encoder depth schedule	$(1,1,4)$ , $(1,2,4)$ , $(1,2,8)$ , $(2,2,4)$ , $(2,2,8)$ , $(2,3,8)$ , $(3,3,8)$
Embedding dimension	48, 96, 192, 240
Patch size $p$	2, 4
Window size $w$	4, 8

Appendix

Table 13: Hyperparameter search space. Depth schedules list the encoder stages only; the decoder mirrors the encoder around the shared bottleneck, which is counted once.

Component	Tokens	Width	MACs, billion
Backbone, 2 blocks	147,456	192	148.6
Backbone, 4 blocks	36,864	384	289.9
Backbone, 4 blocks	9,216	768	286.3
Patch embedding	147,456	192	2.7
Merge, unmerge, and skip projections	—	—	65.2
Output expansion and projection	—	—	22.2

Appendix

Table 14: Estimated Hapi arithmetic for one 24-h forward pass. The three backbone rows include both encoder and decoder blocks at each resolution.

	Return period (years)
Lead	1.5	2	5	10	20	50	100	200	500
96 h (4 d)	0.580	0.542	0.352	0.224	0.146	0.078	0.049	0.030	0.017
120 h (5 d)	0.566	0.518	0.328	0.200	0.124	0.063	0.036	0.022	0.013
144 h (6 d)	0.571	0.490	0.283	0.160	0.090	0.041	0.024	0.014	0.009
168 h (7 d)	0.477	0.458	0.268	0.151	0.084	0.037	0.023	0.016	0.010

Appendix

Table 15: RiverMamba F1-score for floods at 4–7-day leads on the CONUS land mask, 2024 test window. Hapi is not evaluated at these leads. CONUS-pooled from the per-subset TP/FP/FN of the same evaluation used throughout the paper.

Model	Lead	200-yr F1	500-yr F1
persistence	24 h	0.459	0.431
GloFAS	24 h	0.234	0.196
RiverMamba	24 h	0.114	0.086
Hapi	24 h	0.429	0.367
persistence	48 h	0.199	0.149
GloFAS	48 h	0.154	0.117

Appendix

Table 16: CONUS-aggregated F1-score for floods at the 200- and 500-yr return periods. These levels extrapolate the GloFAS Gumbel fit to

4.5\times

and

11\times

its 44-year record. The rankings are descriptive and are not accompanied by significance tests.

		24 h			48 h			72 h
Subset	Model	MAE	NSE	KGE	MAE	NSE	KGE	MAE	NSE	KGE
AIFAS	Hapi	4.14	1.00	0.99	7.84	0.99	0.98	10.91	0.98	0.98
AIFAS	RM	5.39	1.00	1.00	7.89	0.99	0.99	9.91	0.99	0.99
flood-active	Hapi	1.56	0.99	0.97	2.77	0.97	0.95	3.67	0.96	0.93
flood-active	RM	2.90	0.98	0.99	4.10	0.95	0.97	4.83	0.94	0.97
Small rivers	Hapi	0.10	0.82	0.82	0.13	0.78	0.81	0.15	0.76	0.80

Appendix

Table 17: Per-subset continuous discharge skill, pooled over paired cell–day values within each subset, at nominal forecast leads over CONUS in the 2024 test window. MAE is in m 3 s -1 . NSE and KGE are dimensionless. Bold marks lower MAE or higher NSE and KGE for each subset and lead, using values before rounding to two decimal places.

Figure 7: Complementary empirical distributions of per-cell NSE for four forecast sources across five spatial subsets (columns) and nominal 24-, 48-, and 72-h leads (rows). Each curve gives the fraction of cells attaining at least the indicated NSE and uses every finite cell value. The horizontal axis is displayed from

-1

to 1. Values below

-1

remain in the empirical sample and contribute to the fraction below the plotted range.

Figure 8: Complementary empirical distributions of per-cell F1 at the 2-yr return-period threshold for four forecast sources across five spatial subsets (columns) and nominal 24-, 48-, and 72-h leads (rows). Each curve gives the fraction of event-containing cells attaining at least the indicated F1 and uses every finite cell value. Hapi and RiverMamba are compared at their nominal reported leads.

Figure 9: Per-cell NSE over the 2024 test window for each model and lead, CONUS land mask. Cells where the score is undefined (near-zero target variance) are left blank. The pooled values in Table 3 are dominated by high-discharge cells and do not reflect the typical cell shown here.

Figure 10: Per-cell NSE difference, Hapi minus GloFAS, by lead. Positive (blue) favors Hapi. The remaining GloFAS-favored cells form coherent regional clusters. Possible explanations of these regional differences are discussed in Section 5 .

Return period (yr)	1.5	2	5	10	20	50	100
Exceeding cells per day	7,883	3,453	689	270	126	59	36
Lag-1 autocorrelation	0.938	0.908	0.847	0.747	0.652	0.614	0.594
$\tau_{\mathrm{int}}$ (d)	42.3	25.7	10.0	4.8	3.5	3.0	2.7

Appendix

Table 18: Temporal dependence of the observed daily exceedance count on the CONUS mask, 2024 test window, by return-period threshold. Computed at 24-h lead. The series is the common verification target, so the other leads agree to within four cells per day.

\tau_{\mathrm{int}}

is in days.

Block length	7 d	15 d	21 d	30 d	45 d	60 d
F1, Hapi – persistence	14	14	14	14	12	13
F1, Hapi – GloFAS	17	17	17	18	17	21
F1, Hapi – RiverMamba	21	21	21	21	21	21
SEDI, Hapi – persistence	12	12	12	12	12	12
SEDI, Hapi – GloFAS	17	16	16	17	16	19
SEDI, Hapi – RiverMamba	7	7	6	6	6	6

Appendix

Table 19: Cells surviving Holm correction (of 21 per family) as the bootstrap block length varies,

R=10{,}000

throughout. The 30-day column corresponds to the block length used in Figure 11 . Reproduced by bootstrap_block_sensitivity.py , which also emits Table 18 .

Figure 11: Paired moving-block-bootstrap differences in flood forecasting skill over the 2024 CONUS test window. Values are Hapi minus the indicated comparator at each nominal forecast lead and return-period threshold. Positive values favor Hapi. Panel a shows F1 differences and panel b SEDI differences. Bold values and dots indicate $p<0.05$ after Holm correction across the 21 lead–return-period cells within each comparison and metric. RiverMamba’s latest discharge input is one day older than Hapi’s at the same nominal lead. The corresponding 2.5th–97.5th percentile intervals are retained in the source-data file.

Figure 12: Symmetric extremal dependence index (SEDI) for floods on the CONUS land mask over the 2024 test window. SEDI is invariant to the event base rate and ranges from $-1$ to 1, with 0 indicating no skill. Rows group persistence, GloFAS, and Hapi at 24-, 48-, and 72-h leads. Columns give the GloFAS flood thresholds for different return periods. Exact SEDI values are printed in each cell. Bold values and dots mark the best model within each lead–return-period comparison.

Precision
		Return period (years)
Model	Lead	1.5	2	5	10	20	50	100
persistence	24 h	0.836	0.800	0.709	0.627	0.562	0.496	0.470
persistence	48 h	0.726	0.672	0.544	0.434	0.353	0.275	0.233
persistence	72 h	0.653	0.590	0.447	0.334	0.256	0.177	0.140
GloFAS	24 h	0.778	0.723	0.590	0.478	0.405	0.314	0.267

Appendix

Table 20: Precision, recall, and critical success index (CSI) for floods at GloFAS return periods, CONUS land mask, 2024 test window. Metric definitions are given in Appendix H . Recall exceeding precision indicates overforecasting of floods. The reverse indicates underforecasting. Bold marks the highest value for each lead and return period.

Figure 13: Performance diagram for flood forecasting over the CONUS land mask in the 2024 test window. The horizontal axis shows success ratio (precision), and the vertical axis shows probability of detection (recall). Performance improves toward the upper-right corner, and the shaded contours show CSI. Dashed rays show frequency bias, with values above and below one indicating overforecasting and underforecasting, respectively. Marker size decreases from the 1.5-yr to the 100-yr return-period threshold.

Explore similar work

A Deep U-Net Framework for Flood Hazard Mapping Using Hydraulic Simulations of the Wupper Catchment
Apr 22, 2026Christian Lammers, Fernando Arévalo, Leonie Märker-Neuhaus +3Accurate Flood Prediction Flood Detection
Institute for Computing and Information Sciences Radboud University Nijmegen, Netherlands · Wupper Association (Wupperverband) Wuppertal, Germany
Data-efficient flood depth prediction through domain-aware coreset selection and tabular foundation models
Jun 3, 2026Lipai Huang, Adithi Srinath, Manas Singh +2Accurate Flood Prediction Tabular Foundation Models
Urban Resilience.AI Lab, Zachry Department of Civil and Environmental Engineering, Texas A&M University, College Station, TX, USA · Department of Computer Science and Engineering, Texas A&M University, College Station, TX, USA · Resilitix Intelligence LLC, Houston, TX, USA +1
Advanced Flood Prediction with Physics-Guided Deep Learning: Combining UNet, FNO, and SAR/Optical Imagery
Jun 2, 2026Tewodros Syum Gebre, Jagrati Talreja, Leila Hashemi-BeniAccurate Flood Prediction Flood Detection
Built Environment Department, College of Science and Technology, North Carolina A&T State University, Greensboro, NC 27405, USA · United Nations University Institute for Water, Environment and Health, Richmond Hill, ON, Canada

Apr 22, 2026, cs.LGJ/K move · Enter open · S save

A Deep U-Net Framework for Flood Hazard Mapping Using Hydraulic Simulations of the Wupper Catchment

Christian Lammers, Fernando Arévalo, Leonie Märker-Neuhaus, Daniel Heinenberg+2

Institute for Computing and Information Sciences Radboud University Nijmegen, Netherlands · Wupper Association (Wupperverband) Wuppertal, Germany

Accurate Flood Prediction Flood Detection Surrogate Models U-Net Deep Learning Water+1

Open paper AskPDF arXiv

The increasing frequency and severity of global flood events highlights the need for the development of rapid and reliable flood prediction tools. This process traditionally relies on computationally expensive hydraulic simulations. This research presents a prediction tool by developing a deep-learning based surrogate model to accurately and efficiently predict the maximum water level across a grid. This was achieved by conducting a series of experiments to optimize a U-Net architecture, patch generation, and data handling for approximating a hydraulic model. This research demonstrates that a deep learning surrogate model can serve as a computationally efficient alternative to traditional hydraulic simulations. The framework was tested using hydraulic simulations of the Wupper catchment in the North-Rhein Westphalia region (Germany), obtaining comparable results.

Abstract

Explore similar work

A Deep U-Net Framework for Flood Hazard Mapping Using Hydraulic Simulations of the Wupper Catchment

Data-efficient flood depth prediction through domain-aware coreset selection and tabular foundation models

Advanced Flood Prediction with Physics-Guided Deep Learning: Combining UNet, FNO, and SAR/Optical Imagery