MausamNow

About MausamNow

Free weather intelligence for India. Compares 5 forecast models, analyzes live IMD radar, tracks storms via satellite, and monitors air quality from ground stations.

No ads No tracking Free to use
⚠️ Under active development. Forecasts are experimental and combine multiple data sources with varying accuracy. Always verify critical weather decisions with official IMD warnings.
What does MausamNow do?

"Will it rain in the next 2 hours?" — We combine five sources every 10 minutes: IMD radar (what's happening now), INSAT-3DS satellite (clouds building up), and 119 ensemble model runs across GFS, ECMWF and ICON. When radar, satellite and models agree, the answer is reliable. When they disagree, you'll see that too.

"What does the week look like?" — We compare forecasts from 5 different weather models. When 3 or more agree, confidence is high. When they disagree, we show you.

How confidence is calculated

Every forecast on MausamNow shows a confidence level. Here's what each means:

High 3+ data sources agree (radar + satellite + models). Highly reliable.
Medium 2 sources agree, or models partially diverge. Reasonably reliable.
Low Models disagree or limited data available. Treat as uncertain.

For daily forecasts: Day 0–1 is typically high confidence (many data sources), Day 2–3 medium, and Day 4+ low. Indian models (MausamGram, MOSDAC) boost confidence for their coverage area.

No forecast is 100% certain. Confidence shows how much agreement exists between independent sources. High confidence can still be wrong, and low confidence doesn't mean it's wrong — just less certain.

Data Sources

Radar & Satellite

SourceWhat it provides
IMD Doppler RadarLive rain detection from 38 stations, storm tracking, intensity mapping
INSAT-3DS Satellite (ISRO)India's primary weather satellite — cloud state, IR imagery
INSAT-3DR Satellite (ISRO)India's second geostationary satellite — complementary coverage
Himawari-9 (Japan)Best for eastern India — cloud development, early storm detection
Meteosat-9 IODC (EUMETSAT)Best viewing angle for western India coast

IMD Official Data

IMD Official APINowcasts, district warnings (17 hazard types), 7-day city forecasts, 1,185 AWS/ARG stations (temperature, humidity, wind, pressure, observed rainfall)
IMD MausamGramIndia-specific multi-model ensemble forecast
IMD GFS T1534India's own GFS model — precipitation, CAPE, thunderstorm prediction, 10-day range

ISRO / MOSDAC

MOSDAC / ISRONWP rain forecast, satellite-derived products (cloud mask, rainfall estimation)

Ground Stations

WeatherUnion (Zomato)Ground station rain measurement — actual rainfall at your location
CPCB Ground StationsAir quality (PM2.5, PM10, NO₂, AQI) from real monitoring stations

Global Forecast Models

GFS (NOAA, USA)Global forecast model, 16-day range
ECMWF IFS (Europe)Most accurate global model, 15-day range
GraphCast (Google DeepMind)AI weather model, 10-day range

Global forecasts via Open-Meteo.com (CC BY 4.0). All other data from freely available government and public APIs.

How accurate is it?

Radar (now): Reliable for rain detection and general intensity. We read published radar images, not raw reflectivity, so exact mm/hr values are approximate.

Forecasts (next 24h): Generally reliable when 2+ models agree. Indian models (MausamGram, MOSDAC) tend to be better for short-range.

Forecasts (2–5 days): Useful for planning. Confidence decreases with time.

Forecasts (5–16 days): General trends only. Not reliable for specific plans.

Thunderstorm prediction: Conditions-based, not guaranteed. High risk means conditions strongly favor storms.

Air quality: CPCB ground station readings are real measurements. More accurate than model estimates.

Radar coverage: IMD's 38 stations don't cover all of India. Some rural areas may be outside range.

Data freshness: Radar 10–40 min old. Satellite updates every 10 min. Models run every 1–12 hours.

About the data

Multi-model comparison: MausamNow fetches forecasts from 6 independent weather models (GFS, ECMWF, GraphCast, IMD GFS, MausamGram, MOSDAC) and shows where they agree and disagree. No single model is always right — comparing them gives a better picture.

Rain windows: When the expanded daily card shows "4–7 PM: Likely, 4/6 models agree", it means 4 out of 6 models predict rain during that window. The more models agree, the more confident the forecast.

Tip badges: ☂️ = rain likely, ⚡ = storm risk (CAPE > 1500 J/kg), 💨 = strong gusts (50+ km/h), 🥵 = extreme heat (40°C+). These are computed from hourly model data for each day.

Expert mode: Shows raw output from each model side-by-side. Useful for weather enthusiasts and for understanding why models disagree on a particular day.

Built by Abhijit Vaidya

Software engineer based in Pune. Built MausamNow to make publicly available weather data accessible to everyone.

Questions? Get in touch

Not affiliated with IMD, ISRO, or any government body. Independent, non-commercial project. For informational use only — always check official IMD warnings for severe weather.

Data attribution: Radar imagery, weather warnings, nowcasts, city forecasts, NWP meteograms, and 1,185 AWS/ARG station observations from India Meteorological Department (IMD) via api.imd.gov.in. Forecast data from IMD MausamGram. NWP rain forecast and satellite-derived products from MOSDAC / ISRO. Satellite imagery from ISRO (INSAT-3DS, INSAT-3DR), JMA Himawari-9 via SSEC RealEarth, and EUMETSAT Meteosat-9 IODC.

Global forecasts (GFS, ECMWF, GraphCast), ensemble probabilities, and current conditions by Open-Meteo.com (CC BY 4.0). Air quality from Central Pollution Control Board via data.gov.in (OGDL). Ground station rainfall from WeatherUnion (Zomato).

Indian government data used under the Open Government Data License (OGDL) which permits free use, sharing, and adaptation with attribution.