Udai Rathore | GeoAI & Remote Sensing

Spatial intelligence for India's water & climate risks 🛰️ Remote sensing · GIS · SAR · GeoAI | Daily #30DayMapChallenge maps | More on LinkedIn ↓

[Udai Rathore | GeoAI & Remote Sensing @udai225]

37% of this season's satellite fire detections in Punjab and Haryana came from 4.5% of the land, all in one belt from Amritsar to Firozpur.

Getis-Ord Gi* on a 5 km equal-area grid, FDR-corrected. NASA FIRMS VIIRS, 15 Sep–2 Nov 2026.

#30DayMapChallenge Day 4: Clusters

[Udai Rathore | GeoAI & Remote Sensing @udai225]

@bicycle_geek Letting people place bus stops and see slope-weighted coverage change live is a strong use of interaction. In access analysis, slope and walking time change the gap picture a lot. Which walking-speed assumption did you use?

[Udai Rathore | GeoAI & Remote Sensing @udai225]

@petrihelo Replacing distance with travel time shows how ferries and air links reshape the Nordic map. Switching scenarios makes the access gaps obvious. Do you weight waiting and transfer time in the ferry and air scenarios?

[Udai Rathore | GeoAI & Remote Sensing @udai225]

The Geo Daily | 13 Nov 🛰️ ESA: strong El Niño may put land carbon sinks at risk. MTG-I2 sends first images. CoStar to buy Vexcel for ~$730M. My take: local validation decides how useful forecasts are. #GeoDaily #EarthObservation #RemoteSensing

[Udai Rathore | GeoAI & Remote Sensing @udai225]

Stream order turns a tangle of lines into a story. Milan Janosov's animation draws 185,730 African river segments, largest to smallest, ranked by Strahler order. Which basin should be animated next? #Hydrology #GIS #30DayMapChallenge https://t.co/leIrhEhi7h

[Udai Rathore | GeoAI & Remote Sensing @udai225]

Resolution decides which questions you can ask. Milan Janosov's Amsterdam canal: 400 pixels at 5 cm, 2 at Sentinel-2's 10 m, under 1 at Landsat's 30 m. Pick the question first, then the pixel size. #RemoteSensing #GIS https://t.co/83P2y86T4N

[Udai Rathore | GeoAI & Remote Sensing @udai225]

@ShellyElbazZ Burn scars as inkblots is a clever way to show how irregular real fire footprints are. With every polygon mapped from Sentinel-2, each shape traces back to an actual event. Did you delineate them with a burn index like dNBR?

[Udai Rathore | GeoAI & Remote Sensing @udai225]

The Geo Daily | 12 Nov 🛰️ NISAR validated L0–L2 L-band data released 6 Nov. 🌡️ WMO at COP31: 2026 among 4 warmest years, record sea level, El Niño heading to record strength. Which NISAR product will you try first? #GeoDaily #EarthObservation #RemoteSensing

[Udai Rathore | GeoAI & Remote Sensing @udai225]

@FerTimJim A clever way to show how ecology is written into place names. Great use of INEGI data and a clear layout with the bar charts alongside the map.

[Udai Rathore | GeoAI & Remote Sensing @udai225]

@mepa1363 Great map, Abraham. Turning 165k time zone answers into a clear picture of who skipped the clock change is a smart use of prompting. Did you cross-check the API output against official sources? #30DayMapChallenge

[Udai Rathore | GeoAI & Remote Sensing @udai225]

The Geo Daily | 11 Nov 2026 🛰️
EARSC, ESA & IOGP workshop on EO for energy: 11–12 Jan 2027, ESA-ESRIN, Frascati (register by 20 Dec).
EO market: ~$7.68B (2026) → $14.55B (2034).
Sentinel-1D now continues C-band SAR from 1A.
#GeoDaily #EarthObservation #RemoteSensing

[Udai Rathore | GeoAI & Remote Sensing @udai225]

📰 The Geo Daily No. 2: ahead of #COP31 in Antalya, UCLA's analysis of Tanager-1 data flags 70 methane plumes over Türkiye's top 10 landfills; Istanbul's Seymen averaged ~7.4 t/h. Also: UK unveils Titania recon satellite. Carbon Mapper data is open. #Methane #RemoteSensing

[Udai Rathore | GeoAI & Remote Sensing @udai225]

Sources 📚
Floods: https://t.co/16bN2dw03u
USDA × NASA: https://t.co/CmZyFnq6QC
Pelican-12: https://t.co/fNcKxmVBL7
Imagery: Copernicus Sentinel-2 via Microsoft Planetary Computer

[Udai Rathore | GeoAI & Remote Sensing @udai225]

📰 The Geo Daily No. 1: the Four Corners floods from orbit. Sentinel-2, 5 Oct vs 3 Nov: the Animas River runs mud-red through Farmington, NM. Also: USDA × NASA team up on satellite crop forecasts; Planet's Pelican-12 sends 50 cm first light. #RemoteSensing #Sentinel2

[Udai Rathore | GeoAI & Remote Sensing @udai225]

@mepa1363 Keeping all 24 cities at one scale is what makes this work. Which city surprised you most on transit versus walking?

[Udai Rathore | GeoAI & Remote Sensing @udai225]

@liamreckziegel Pairing the fire edge with lidar terrain is a nice idea. Is the edge from your drone imagery or an agency perimeter? I'm mapping VIIRS fire clusters in Punjab and Haryana this week, so I'm curious how they compare.

[Udai Rathore | GeoAI & Remote Sensing @udai225]

@Guaiacumac Muy claro el contraste entre las colonias junto a los cerros y las del centro. ¿Usaron una sola escena Landsat o una mediana de temporada?

[Udai Rathore | GeoAI & Remote Sensing @udai225]

@ShellyElbazZ MNDWI is a good choice there, though salt crust on the shoreline can be tricky. Did you use one threshold across 1984, 2004 and 2025, or tune it per scene? The VHS look suits a 40-year comparison.

[Udai Rathore | GeoAI & Remote Sensing @udai225]

@brianjbusch Jordan basin still at 98% while the region fell from 47% to 23% is a striking contrast. Do you see that persistence in reservoir storage too, or mainly in the Drought Monitor?

[Udai Rathore | GeoAI & Remote Sensing @udai225]

@KaranD93 @OvertureMaps @openstreetmap Running HDBSCAN one cuisine at a time is a smart call, since a single run would let the big categories swallow the small ones. How did you handle restaurants that appear in both Overture and OSM?

[Udai Rathore | GeoAI & Remote Sensing @udai225]

@fajrunwm Lombok is a nice test of the method, since two thirds of that nest is the 2018 sequence. Did you try declustering aftershocks first (e.g. Gardner-Knopoff)? I'd be curious whether all 50 nests hold up or a few collapse into single sequences.