# Cyclone Murasu: satellite map of candidate flooded areas on the Andhra Pradesh coast **Event:** Severe cyclonic storm Murasu, landfall near Ongole early on 3 November 2026 **Where:** Bapatla, Guntur and Prakasam districts, Andhra Pradesh, India **Event date:** landfall 3 November 2026 **Imagery compared:** one Sentinel-1 radar pass and a Sentinel-2 optical composite, each against an image from before the event **Facts checked:** 2026-11-14 · **Analysis packaged:** 2026-11-13T18:34:48Z **Status:** Satellite-detected surface water: candidate flooded areas, not a confirmed flood extent. Impact figures are as reported; several are provisional. ## The story in brief Severe cyclonic storm Murasu crossed the Andhra Pradesh coast near Ongole in Prakasam district at about 1 am IST on Tuesday 3 November 2026, according to the India Meteorological Department (IMD). The UN and EU Global Disaster Alert and Coordination System (GDACS) gave the storm its highest, Red, alert. It arrived about a week after Cyclone Arnab, which had already hit the same coast. IMD warned of flash floods in Bapatla, Guntur, Krishna, Palnadu, Prakasam and SPSR Nellore districts. In the 24 hours to 8.30 am on 3 November, IMD recorded 22 cm of rain at Bapatla, and municipal gauges recorded about 313 mm at Ponnur and 298 mm in Guntur city. On 5 November officials told Chief Minister N. Chandrababu Naidu that eight people had died: five in Bapatla district, two in Guntur and one in Palnadu. Comparing satellite images from before and after the event, we found **1,739 candidate flooded areas covering about 264 km²** of low, flat ground in the study area. These are places where surface water appeared that was not there before. They are candidates, not a confirmed flood extent: the satellites passed days after the peak, so some water had already drained, and some new water may have other causes, such as irrigation or ponds being filled. ## What has been reported The figures below are as published by the sources shown. Where authorities had not confirmed a figure, it is marked as provisional. - **Deaths:** 8: five in Bapatla district, two in Guntur and one in Palnadu — official figures given to the Chief Minister's review. Source: [The News Minute](https://www.thenewsminute.com/andhra-pradesh/cyclone-murasu-8-dead-cm-naidu-estimates-damages-at-rs-1919-cr-in-andhra), 5 November 2026. - **Crops:** 143,122 hectares of agricultural crops damaged, affecting 186,063 farmers; 6,749 hectares of horticultural crops — preliminary; officials said a fuller assessment would follow once floodwater receded. Source: [The News Minute](https://www.thenewsminute.com/andhra-pradesh/cyclone-murasu-8-dead-cm-naidu-estimates-damages-at-rs-1919-cr-in-andhra), 5 November 2026. - **Homes, boats and roads:** 983 houses, 346 boats and 220 fishing nets damaged; 760 km of major district roads and state highways damaged — preliminary. Source: [The News Minute](https://www.thenewsminute.com/andhra-pradesh/cyclone-murasu-8-dead-cm-naidu-estimates-damages-at-rs-1919-cr-in-andhra), 5 November 2026. - **Estimated losses:** Rs 1,918.55 crore — preliminary. Source: [The News Minute](https://www.thenewsminute.com/andhra-pradesh/cyclone-murasu-8-dead-cm-naidu-estimates-damages-at-rs-1919-cr-in-andhra), 5 November 2026. - **Heaviest rainfall:** 22 cm at Bapatla in the 24 hours to 8.30 am on 3 November (IMD); 312.76 mm at Ponnur and 298.36 mm in Guntur city (municipal gauges) — as reported. Source: [Deccan Chronicle; The Times of India](https://www.deccanchronicle.com/southern-states/andhra-pradesh/murasu-causes-crop-damage-flooding-in-south-coastal-ap-2001480), 3 November 2026. ## What the satellites show - **Candidate flooded areas:** 1,739 areas (264 km²). - 362 areas (160 km²) were detected by two or more independent observations, each covering at least a quarter of the area: the strongest evidence. - 1,377 areas (104 km²) were detected by one observation, or by several that each covered only a small part of the area. - **Land cover under candidate water** (ESA WorldCover 2021): cropland 241 km², bare or sparse vegetation 9.70 km², herbaceous wetland 4.38 km², shrubland 3.88 km². - **Main roads:** about 79.8 km of OpenStreetMap main roads lie within 20 m of candidate water. This shows exposure, not damage or closure. - **Water already present before the event:** 191 km² on land (not counting the sea or permanent lakes and rivers), shown in navy on the map and excluded from the candidates. ### Observations - **Sentinel-1, 9 Nov (descending)** compared with 16 Oct: 154 km² of new surface water detected; valid data over 100% of the study area. - **Sentinel-2, 10 and 12 Nov** compared with 1, 6 and 26 Oct 2026: 253 km² of new surface water detected; valid data over 94% of the study area. ### Largest candidate areas | ID | Area (ha) | Confidence | Main land cover | Nearest mapped town or village | Main roads within 20 m | |---|---:|---|---|---|---| | MU0001 | 1,843.2 | higher | cropland | Swarna (3.4 km) | MDR036 Ravuru Road, MDR081, tertiary | | MU0002 | 1,622.1 | higher | cropland | Kunkalamarru (2.5 km) | NH167A | | MU0003 | 912.8 | higher | cropland | Bhallukanudupalem (1.5 km) | SH273 Ponnuru - Kakumanu Road | | MU0004 | 616.4 | higher | cropland | Murukondapadu (3.2 km) | — | | MU0005 | 597.9 | higher | cropland | Ipurupalem (2.1 km) | tertiary | | MU0006 | 589.4 | higher | cropland | Nanddiguntapalem (3.4 km) | tertiary | | MU0007 | 535.1 | higher | cropland | Chinlingayapalem (0.8 km) | MDR060, MDR061, tertiary | | MU0008 | 527.4 | higher | cropland | Sangam (3.1 km) | MDR0124 | | MU0009 | 446.8 | higher | cropland | Chirala (2.3 km) | Chirala - Karamchedu Main Road | | MU0010 | 361.9 | higher | cropland | Chirala (2.0 km) | Chirala - Karamchedu Main Road, NH167A, tertiary | ### Main roads closest to candidate water - Unnamed tertiary road: about 22.1 km within 20 m of candidate water - NH167A: about 11.2 km within 20 m of candidate water - SH273 Ponnuru - Kakumanu Road: about 3.9 km within 20 m of candidate water - MDR081: about 3.9 km within 20 m of candidate water - MDR061: about 3.8 km within 20 m of candidate water - SH271 Bapatla - Pedanandipadu Road: about 2.9 km within 20 m of candidate water - MDR0124: about 2.6 km within 20 m of candidate water - MDR060: about 2.2 km within 20 m of candidate water - SH301 Ongole - Kothapatnam Road: about 1.7 km within 20 m of candidate water - MDR0181: about 1.6 km within 20 m of candidate water ## Independent mapping India's National Remote Sensing Centre (NRSC/ISRO) published maps of water-accumulated areas in Andhra Pradesh from RISAT-1A radar on 5 November and from RADARSAT Constellation Mission radar on 6 November (see References). Those maps are not openly licensed, so they are not reproduced or compared pixel by pixel here, but they offer an independent official view of where water lay a few days before our images. ## How we did it 1. **Radar (Sentinel-1).** Radar sees through cloud. Calm open water reflects the radar beam away from the satellite, so it looks dark. We compared terrain-corrected VV backscatter after the event with an image from before it, smoothing over 3 × 3 pixels of 20 m to reduce speckle. A pixel was flagged as new water when it was -18 dB or darker after the event, was brighter than that before, and had darkened by at least 3 dB. 2. **Optical (Sentinel-2).** Where the sky was clear, we compared the modified normalised difference water index (MNDWI, green against short-wave infrared). A pixel was flagged when the index rose from 0 or below before the event to 0.1 or above after it. Cloud and cloud shadow were removed using the scene classification layer. 3. **Where water can lie.** Only flat ground was assessed: slopes under 5° and land no more than 15 m above the lowest ground within about 2–4 km (Copernicus DEM GLO-30). Permanent water, and water already visible in the before images, were excluded. Built-up land was excluded from the radar test, because buildings scatter radar strongly and hide water. 4. **Confidence and clean-up.** Each observation counts separately. Flagged pixels were grouped into patches, and patches smaller than 2 ha were discarded. A patch is higher confidence when two or more observations each flagged at least a quarter of it, and moderate otherwise. Each patch was turned into a polygon and attributed with land cover, nearby main roads and the nearest mapped town or village. ## Limits - **Candidates, not a confirmed flood extent.** Rain on ploughed fields, newly flooded rice paddies, filled fish or shrimp ponds and wet sand can all look like new water. Wind-roughened water can look bright to radar and be missed. - **Timing.** The satellites passed days after the heaviest rain. Water that had already drained is not shown, so the true extent at the peak was probably larger. - **Hidden water.** Flooding under tree cover or in built-up streets is largely invisible to this method. - **Resolution.** Pixels are 20 m across, so narrow channels and small patches are missed. Areas are planimetric. - **OpenStreetMap completeness.** Road and place data are incomplete in places, so proximity figures are minimums. ## References 1. GDACS (UN and European Commission), “Tropical cyclone MURASU-26: Red alert, 1–3 November 2026”, November 2026. 2. India Meteorological Department, “Severe cyclonic storm Murasu: bulletin issued at 11:20 IST on 3 November 2026”, 3 November 2026. 3. Hindustan Times, “Cyclone Murasu makes landfall in Andhra Pradesh, triggers heavy rain; thousands evacuated”, 3 November 2026. 4. The News Minute, “Cyclone Murasu: 8 dead, CM Naidu estimates damages at Rs 1,919 cr in Andhra”, 5 November 2026. 5. Deccan Chronicle, “Murasu causes crop damage, flooding in south coastal AP”, 3 November 2026. 6. The Times of India, “Cyclone 'Murasu' triggers heavy rainfall, inundation in coastal AP; Guntur records 298 mm, Ponnur 313 mm”, 3 November 2026. 7. NRSC/ISRO National Database for Emergency Management, “Andhra Pradesh cyclone: water accumulated areas from RISAT-1A, 5 November 2026 18:00”, 5 November 2026. 8. NRSC/ISRO National Database for Emergency Management, “Andhra Pradesh cyclone: water accumulated areas from RCM, 6 November 2026 06:00”, 6 November 2026. ## Data sources and licences - Copernicus Sentinel-1 radiometrically terrain-corrected backscatter (from GRD) — Copernicus Sentinel data terms (free, full and open); RTC processing by Microsoft Planetary Computer, CC BY 4.0. Attribution: “Contains modified Copernicus Sentinel data [2026]”. - Copernicus Sentinel-2 Level-2A surface reflectance — Copernicus Sentinel data terms (free, full and open). Attribution: “Contains modified Copernicus Sentinel data [2026]”. - Copernicus DEM GLO-30 — Copernicus DEM licence (free, worldwide, with attribution). Attribution: “Copernicus DEM GLO-30 © DLR e.V. 2010–2014 and © Airbus Defence and Space GmbH 2014–2018, provided under COPERNICUS by the European Union and ESA”. - ESA WorldCover 10 m 2021 v200 — CC BY 4.0. Attribution: “© ESA WorldCover project 2021 / Contains modified Copernicus Sentinel data (2021) processed by ESA WorldCover consortium”. - OpenStreetMap main roads, places and rivers (Overpass API) — Open Database Licence (ODbL) 1.0. Attribution: “© OpenStreetMap contributors, available under the Open Database Licence (ODbL)”. - EOX Maps WMS (OpenStreetMap locator) — OpenStreetMap data under ODbL; rendering by EOX and MapServer. Attribution: “Locator: data © OpenStreetMap contributors (ODbL); rendering © EOX and MapServer”. Contains modified Copernicus Sentinel data [2026]. OpenStreetMap data © OpenStreetMap contributors, available under the [Open Database Licence](https://opendatacommons.org/licenses/odbl/1-0/). The map and derived candidate polygons are released by TechnicalGIS under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). OpenStreetMap-derived attributes in the GeoJSON and CSV (road names, place names) remain under the ODbL. Prepared by TechnicalGIS Environmental News Desk. Package revision 1.