The research, conducted by Abdulrhman M. Almoadi of King Abdulaziz City for Science and Technology (KACST) in Riyadh, Saudi Arabia, introduces the Enhanced Normalized Difference Water Index, or ENDWI. It builds on one of the most widely used tools in remote sensing, the Normalized Difference Water Index proposed by S. K. McFeeters in 1996. The classic NDWI exploits a fundamental property of water: it reflects green light strongly but absorbs near-infrared radiation. By combining the green and near-infrared bands in a normalized ratio, NDWI produces high values over water and low values over most land. The trouble is that built-up surfaces, such as concrete roofs and paved roads, also reflect more green than near-infrared, landing in the same spectral neighborhood as water and triggering false positives across urban landscapes.

ENDWI attacks this confusion with a single additional operation: it divides the NDWI value by the green band reflectance. The logic rests on a subtle difference between how water and urban surfaces behave. Both may reflect green light similarly, but water’s strong near-infrared absorption drives its NDWI value far higher than that of rooftops or asphalt, which reflect energy in both bands. When NDWI is divided by the green band, water pixels retain proportionally larger values while non-water urban surfaces are strongly attenuated. The green band effectively acts as a suppressor of urban noise, amplifying the contrast that near-infrared absorption creates. Because the method relies only on the green and near-infrared bands, the same bands used by NDWI, it works with virtually any modern satellite or drone platform, avoiding the shortwave infrared bands that some competing indices require and that are less commonly available on low-cost sensors.

But the elegant formula came with an unexpected mathematical headache. Dividing by the green band compresses ENDWI’s raw values into an extraordinarily narrow range near zero, spanning roughly from minus 0.000412 to plus 0.000150, more than three orders of magnitude tighter than typical spectral indices. That near-zero clustering breaks Otsu’s method, the standard automatic thresholding algorithm that separates water from land by finding the optimal split in a bimodal histogram. To solve this, Almoadi invented a second novel technique called Zero-Preserving Split Normalization, or Z-Split. The method independently rescales the positive and negative portions of the ENDWI distribution to the intervals from zero to one and minus one to zero, while strictly preserving zero as the natural boundary between water and non-water pixels. The result is a cleanly bimodal histogram that Otsu’s method can split reliably. Importantly, Z-Split is not part of the index itself but a preprocessing step, and the author suggests it could benefit any bipolar spectral index suffering from similar value compression.

To test the new approach, the study turned to a dramatic real-world event. On 23 November 2018, heavy rainfall in the upstream catchment of Wadi Al-Lith triggered the partial breach of an earthen retaining dam on Saudi Arabia’s Red Sea coast. A surge of floodwater reached the urban areas of Al-Lith Governorate within roughly four hours, submerging roads and low-lying neighborhoods, with water levels exceeding 1.7 meters in some sectors after further rain on 25 November. Al-Lith’s mix of residential buildings, paved roads, bare soil, and scattered vegetation makes it a textbook example of the spectral confusion that plagues optical flood detection, since water, shadows, and built-up surfaces can appear nearly identical in multispectral imagery.

The validation design was unusually rigorous. Almoadi used Sentinel-2 imagery acquired at 10-meter resolution on 28 November 2018 as the analytical dataset, and validated results against WorldView-4 imagery pan-sharpened to 0.31-meter resolution, acquired within two days of the event. From manually digitized flooded and non-flooded polygons, buffered inward by 10 meters to avoid mixed edge pixels, the study generated 1,262 ground-truth points: 559 flooded and 703 non-flooded. That validation density, concentrated in a compact area of roughly 18 square kilometers, lends considerable statistical weight to the reported accuracies.

The results unfolded in three stages. First, a receiver operating characteristic analysis of eight raw indices, including ENDWI and seven established competitors such as MNDWI, AWEIsh, AWEInsh, WI, LSWI, and SWI, showed the shadow-aware AWEIsh leading with an area under the curve of 0.908, followed by NDWI at 0.671 and ENDWI at 0.636. Second, after Otsu thresholding, ENDWI stole the show among individual indices: it achieved the highest precision at 79.41 percent and the lowest false alarm rate at 10.95 percent, roughly half the false alarm rate of AWEIsh and about one-third that of NDWI. Third, and most strikingly, the study fused the two complementary indices using a pixel-wise maximum operation, taking whichever value, ENDWI or AWEIsh, indicated water more strongly at each pixel. AWEIsh contributed broad separability and shadow handling, while ENDWI suppressed false positives from dark impervious surfaces. The fused map, thresholded with Otsu’s method, achieved 82.65 percent overall accuracy, 94.50 percent precision, an F1-score of 76.73 percent, and a Kappa coefficient of 0.637, with only 21 false positives out of 703 non-flooded validation points, a false alarm rate of 2.99 percent.

Those numbers translate into a 15-fold reduction in false alarms compared with the classic NDWI pipeline, achieved at essentially zero additional computational cost. No deep learning models, no labeled training datasets, no extra sensors, and no multi-sensor fusion are required, just one division operation and a maximum rule that can run in seconds in standard GIS software. That efficiency matters enormously during rapid disaster response, when deep learning approaches, however powerful, demand labeled data and computational resources that are often scarce in the critical first hours after a flood. The method’s reliance on free Sentinel-2 data also makes it attractive for resource-constrained agencies in arid and semi-arid cities, where flash floods are becoming more frequent and severe as climate change intensifies extreme rainfall. Floods affected 2.3 billion people between 1995 and 2015, and between 1980 and 2009 they killed more than half a million people worldwide, making them the deadliest natural disaster type in that period.

The study is candid about its limits. It rests on a single post-event image pair from one flood event, so generalization across seasons, turbidity levels, and vegetation conditions remains unproven, though a preliminary visual check against an independent flood event reported elsewhere suggested ENDWI again outperformed NDWI. Optical methods also require cloud-free skies, and Sentinel-2’s 10-meter resolution can miss narrow urban water features. Future work, the author suggests, could explore adaptive fusion weights, additional bands such as red-edge for vegetation masking, time-series validation across multiple events, and integration with radar data for all-weather capability. Still, the core lesson is a refreshing one for a field increasingly dominated by heavyweight machine learning: sometimes a modest, interpretable change to a 30-year-old formula, applied to free data, can deliver a dramatic practical win. The code, ground-truth points, and the Z-Split Python package have been released openly on GitHub and PyPI, inviting flood mappers everywhere to put the new index to the test.

Subject of Research: An enhanced spectral water index and hybrid fusion method for reducing false alarms in satellite-based urban flood detection

Article Title: Reducing false alarms in urban flood detection: an enhanced NDWI (ENDWI) with Hybrid Max Fusion on Sentinel-2 Data

Article References: Almoadi, A. M. (2026). Reducing false alarms in urban flood detection: an enhanced NDWI (ENDWI) with Hybrid Max Fusion on Sentinel-2 Data. Earth Observation, 1(1), 43-57. https://doi.org/10.5194/eo-1-43-2026

Image Credits: AI Generated

DOI: 10.5194/eo-1-43-2026

Keywords: urban flooding, flood detection, remote sensing, Sentinel-2, NDWI, water index, Otsu thresholding, false alarms, satellite imagery, disaster response, Saudi Arabia, spectral indices

The research, drawn from two prospective clinical trials conducted at specialized Dutch epilepsy centers, followed 65 children aged 3 to 16, roughly half of them girls, who each experienced at least one nocturnal major motor seizure per week. The children wore the NightWatch system, a wearable armband that triggers alerts based on movement and heart rate, for a median of 47 nights each, either at home or in residential care units. Continuous video monitoring served as the gold standard against which every alarm was judged. An alarm was counted as true if a major motor seizure, encompassing generalized tonic–clonic seizures, focal-to-bilateral tonic–clonic events, and prolonged tonic or clonic seizures, occurred within a six-minute window around the trigger. Every other alarm, including those caused by minor seizures, was labeled false. In total, the researchers analyzed 2,876 alarms, of which 742, or 25.8 percent, were genuine seizures.

To test whether true and false alarms followed different temporal rhythms, the team turned to a statistical framework known as generalized estimating equations, or GEE. Each alarm was treated as a single observation, with alarm type as the binary outcome and the clock hour of occurrence as the predictor. Because some children generated far more alarms than others, the model used an exchangeable working correlation structure keyed to each child’s identity, preventing the heaviest alarm producers from dominating the results. The researchers also normalized hourly alarm counts by the cumulative device-hours within each clock-hour bin, so that variations in bedtimes and wake-up times across the cohort could not distort the picture. A permutation test compared the skewness of the two alarm distributions, asking whether either true or false alarms showed greater temporal asymmetry across the fourteen-hour monitoring window running from 19:00 to 09:00.

The answer was unambiguous. The temporal occurrence of true and false alarms differed significantly, with the GEE model yielding a p-value below 0.001. True alarms were spread across the nocturnal period with moderate peaks around 21:00 to 22:00 and again around 05:00 to 06:00, roughly tracking the natural ebb and flow of seizure likelihood during sleep. False alarms told a very different story. They were heavily skewed toward the morning transition, and the permutation test confirmed a significant divergence in skewness between the two distributions, with the false alarm skew measuring 4.12 standard deviations from the null distribution. Fully 30.2 percent of all false alarms fell within the 06:00 to 09:00 interval. Under a uniform distribution, any three-hour block within the fourteen-hour window would be expected to contain only 21.4 percent of events, meaning the morning cluster represents a relative enrichment of more than 40 percent above baseline.

The cumulative picture is perhaps the most striking way to visualize the divergence. While true alarms accumulated steadily through the night, false alarms remained sparse during the early and middle hours before rising sharply after 06:00, as if the device’s error rate woke up with the children. The hourly ratio of false to true alarms made the pattern clinically concrete. In the early evening, between 20:00 and 22:00, the normalized ratios sat well below the nightly mean, at 0.67 and 0.52 respectively, indicating hours in which the device was comparatively trustworthy. Between 06:00 and 09:00, those ratios climbed to 1.66, 1.70, and finally 2.56, with confidence intervals that never dipped below 1.2. In other words, the last three hours of the monitoring window carried a false alarm burden more than two and a half times the nightly average in the final hour.

Why would a seizure monitor fail most often just before dawn? The researchers point to wakefulness and morning arousal as the likely culprits. As children transition from sleep toward waking, they stretch, shift position, and experience surges of autonomic activity, precisely the signals the armband is designed to interpret as danger. Notably, because the NightWatch automatically suppresses alerts when the user leaves a recumbent position, the morning surge reflects in-bed detection of increased motor or autonomic activity rather than simple removal of the device. The early evening pattern offers a mirror image: between 19:00 and 22:00, false alarms were relatively scarce and true alarms relatively abundant, possibly because arousals and awakenings that trigger false detections are less likely at that hour, or because seizures themselves are more prone to occur as children settle into sleep.

The implications reach well beyond one device. Current detection algorithms generally operate under a static sensitivity model, assuming that physiological conditions remain uniform throughout the night. This study demonstrates that the assumption is fundamentally flawed. A one-size-fits-all threshold performs well during deep nocturnal sleep, when the noise associated with wakefulness is minimal, but stumbles badly during the waking transition. The authors argue that their findings make the case for dynamic detection models, in which thresholds shift with time of night or with the wearer’s inferred sleep state. Such adaptive algorithms could tighten sensitivity during the high-risk early morning window while relaxing it during the quiet hours when false detections proliferate, potentially easing the alarm fatigue that drives caregivers to ignore alerts or abandon devices altogether.

Yet the path to smarter thresholds is fraught with clinical risk. True seizures also populate the early morning hours, and any modification that dampens sensitivity during the 06:00 to 09:00 window could cause the device to miss a genuine, potentially life-threatening event. The authors are careful to stress that time- or state-dependent thresholds will require prospective validation before they can be deployed. They also acknowledge limitations in their own analysis. The GEE framework estimates population-averaged effects, which means it identified cohort-wide temporal hotspots but may have obscured individual patterns that vary from child to child, and the cohort size was not designed to characterize person-specific periodicities. The team relied on video recordings without electroencephalography, so some alarms labeled false may actually have stemmed from minor seizures rather than innocent movements, introducing physiological heterogeneity that complicates any effort to eliminate false detections entirely.

Even with those caveats, the clinical relevance of the timing finding stands. Most caregivers report that their primary goal is to be alerted to major, dangerous seizures that demand immediate intervention, and the burden of false alarms falls unevenly across the night in ways that matter for real families. Because many false alarms occur near typical awakening times, they may prove less disruptive than those piercing deep sleep, when caregiver vigilance is at its lowest and the stakes of a missed seizure are highest. Conversely, the concentration of false alarms in the morning may compound fatigue precisely when parents are already groggy and less able to discriminate between a device error and an emergency. By mapping when monitoring devices fail, this study transforms alarm fatigue from an vague complaint into a measurable, time-stamped phenomenon, and hands algorithm designers a concrete target: teach the machine to know the difference between a child waking up and a child in danger.

Subject of Research: Temporal patterns of true versus false alarms in wearable pediatric nocturnal seizure detection

Article Title: Temporal Divergence of True and False Alarms in Pediatric Nocturnal Seizure Monitoring

Article References: Shahbakhti, M., Bosch, A. T., van der Palen, J., Leijten, F. S., van Dijk, J. P., Lazeron, R. H. C., & Thijs, R. D. (2026). Temporal Divergence of True and False Alarms in Pediatric Nocturnal Seizure Monitoring. Annals of Clinical and Translational Neurology, Article acn3.70530. https://doi.org/10.1002/acn3.70530

Image Credits: AI Generated

DOI: 10.1002/acn3.70530

Keywords: epilepsy, seizure detection devices, nocturnal seizures, false alarms, alarm fatigue, wearable technology, pediatric neurology, tonic-clonic seizures, sudden unexpected death in epilepsy, detection algorithms, sleep monitoring, adaptive thresholds