Data Methodology

IndiaRealTime republishes public and licensed data, not our own measurements. This page is the technical reference for exactly where every figure on the site comes from, how often it refreshes, and how any number we calculate (like AQI or a city/state ranking) is derived - the detail behind the summary on our About Us page.

Datasets and Refresh Schedule

Dataset Details: Source, Calculation, Coverage & Missing Data

For every dataset below: exactly where the raw data comes from, whether the number shown is that raw value as-is or something calculated from it, how much of India it covers, and - the part most methodology pages skip - what happens when a station, market or source simply doesn't report. Where a fallback shows a nearby city's data instead of this location's own (AQI, weather, mandi prices), that's a real, capped mechanism, never an invented number for a place with no data.

Air Quality (AQI)

Source

Central Pollution Control Board (CPCB) monitoring stations, via data.gov.in

API / Data Endpoint

data.gov.in resource 3b01bcb8-0b14-4abf-b6f2-c1bfd384ba69 ("Real-Time Air Quality Index - Various Locations") - one row per pollutant per station, not a pre-computed AQI

Update Frequency

Every 6 hours (4x daily: 6 AM, 12 PM, 6 PM, 12 AM IST)

Sourced or Calculated

Calculated, not sourced pre-computed. Each pollutant (PM2.5, PM10, NO₂, SO₂, CO, NH₃, Ozone) gets its own sub-index from CPCB's own official National AQI breakpoint tables, and the overall AQI is the WORST (highest) of those sub-indices, not an average - CPCB's own methodology. Full breakdown: How AQI Is Calculated.

Geographic Coverage

Whichever stations/cities data.gov.in's feed carries a current reading for on a given fetch; a city not in the API has no AQI page.

If a Station/Source Doesn't Report

Two separate safeguards, not one: (1) a computed AQI of exactly 500, or a jump of more than 200 from the previous fetch, is treated as a likely bad reading and quarantined - the last known-good value is shown instead, its own original timestamp kept rather than relabelled "now", or the city is dropped from that fetch entirely if no prior good value exists, never a fabricated number; (2) a city or district with no CPCB station of its own shows the nearest OTHER mapped city's live reading, capped at 150 km, with an on-page note ("No CPCB station directly for [city] yet - showing the nearest station with live data"). Past 150 km, an honest empty state is shown instead of a meaningless far-away match.

Historical Data

Every fetch is recorded into a persistent per-city trend table, powering the site's day-over-day and multi-day trend cards.

Weather

Source

WeatherAPI.com

API / Data Endpoint

WeatherAPI's forecast.json endpoint - current conditions, 3-day forecast and today's 24-hour hourly breakdown, one call per tracked city

Update Frequency

Once daily (a 15-minute cron tick batches ~70 cities per run; a full pass across the city list normally finishes within the 3-4 AM IST window)

Sourced or Calculated

Directly sourced, not recalculated: temperature, humidity, wind, UV and "feels like" are all WeatherAPI's own values, used as returned.

Geographic Coverage

Every city in the site's configured list (600+); a city not on that list has no weather page.

If a Station/Source Doesn't Report

If a city's fetch fails, the previous day's data stays in place rather than the page going blank for a transient API error. WeatherAPI's own resolved location (lat/lon/name) is checked so two configured cities that resolve to the same underlying station show as such, rather than silently duplicating one reading under two city names.

Historical Data

7-day temperature history is kept per city for the trend sparkline; no longer-range archive.

Mandi / Commodity Prices

Source

Government of India's Agmarknet mandi-reporting network, via data.gov.in

API / Data Endpoint

data.gov.in resource 9ef84268-d588-465a-a308-a864a43d0070

Update Frequency

Up to twice daily (rate-limited to 2 fetches/day, resetting at midnight - matches how often individual mandis actually report anyway)

Sourced or Calculated

Directly sourced: modal, minimum and maximum prices come straight from what each APMC market reports, grouped by market rather than averaged across a city's several mandis, since two mandis in the same city can legitimately report different prices for the same commodity on the same day.

Geographic Coverage

Whichever markets Agmarknet's own feed includes on a given day - coverage genuinely varies day to day, since not every mandi reports every day.

If a Station/Source Doesn't Report

A city or district with no direct mandi report of its own shows the nearest OTHER mapped city with data instead, capped at 150 km, the same mechanism and on-page note as AQI's station fallback above. Past that range, an honest "no mandi within range" state is shown rather than a meaningless far-away match.

Historical Data

None. Every fetch truncates and replaces the whole table, so only the latest snapshot per market is ever available - no price history or trend is kept for mandi prices.

Fuel (Petrol & Diesel)

Source

State-level prices scraped from Goodreturns.in's fuel-price tables; city-level prices scraped from MyPetrolPrice.in. Both are third-party aggregators of the rates the state-owned oil marketing companies (Indian Oil, Bharat Petroleum, Hindustan Petroleum) publish daily - this site does not fetch from IOCL/BPCL/HPCL directly.

API / Data Endpoint

HTML scrape (neither source offers a public API), with a manual-override fallback for state prices if a source page's structure ever changes

Update Frequency

Once daily

Sourced or Calculated

Directly sourced, with one documented exception: Andhra Pradesh's "state price" is calculated as a city average, since Goodreturns doesn't carry Andhra Pradesh in its state table at all.

Geographic Coverage

State-level covers every state/UT in Goodreturns' table (a handful, including Andhra Pradesh, Ladakh and Lakshadweep, simply aren't on that source); city-level covers 500+ cities via MyPetrolPrice, wherever that source lists one.

If a Station/Source Doesn't Report

A state missing from the source isn't shown with an estimated number - its page simply isn't published. If a scrape fails outright, city pages fall back to "Not available" rather than a stale-looking blank card.

Historical Data

State-level petrol/diesel prices are recorded into a persistent trend table on every fetch (day-over-day change, 7-day sparkline). City-level prices are not - only the latest fetched snapshot is kept, no history.

Gas (CNG / PNG / LPG / Autogas)

Source

A mix: CNG/PNG prices are scraped directly from official City Gas Distributor sites where one exists (MNGL for the Pune area, IGL for Delhi); everywhere else, and for LPG/Autogas everywhere (no official distributor source exists for either), prices come from the same aggregators as fuel (MyPetrolPrice, Goodreturns).

API / Data Endpoint

HTML scrape, no public API

Update Frequency

Once daily

Sourced or Calculated

Directly sourced; a distributor-sourced row is preferred over an aggregator-sourced row for the same city whenever both exist.

Geographic Coverage

Whichever city/type combination one of the checked sources actually publishes - not every city has every gas type.

If a Station/Source Doesn't Report

A city/type combination with no source at all shows "not available" rather than an estimated price.

Historical Data

Not tracked - latest fetch only.

Currency Exchange Rates

Source

Frankfurter exchange-rate API, built on European Central Bank reference rates

API / Data Endpoint

Frankfurter's public REST API

Update Frequency

Once daily, ~7 AM IST

Sourced or Calculated

Directly sourced - ECB reference rates as published, cached; a bounded, non-retried live fetch backs up a missed cron tick rather than showing a stale cached rate indefinitely.

Geographic Coverage

INR against every major world currency Frankfurter itself covers.

Historical Data

Not currently tracked as a persistent trend.

Gold, Silver, Platinum, Palladium

Source

GoldAPI.io

API / Data Endpoint

GoldAPI.io's market-rate endpoint

Update Frequency

Once daily, ~7 AM IST

Sourced or Calculated

Per-gram 999-fine (24K-equivalent) rate comes directly from the provider. Other purity tiers (22K, 18K, 958, 950, 925, 900, 500) are calculated proportionally from that base rate at each tier's real fineness ratio, except gold's 22K/18K, which the provider supplies directly.

Geographic Coverage

One national rate per metal, not city-specific - a jeweller's own local markup on top is never reflected here.

Historical Data

A daily rate is kept per metal, supporting a short trend view where shown.

Panchang, Muhurat & Vrat Calendar

Source

FreeAstrologyAPI.com

API / Data Endpoint

FreeAstrologyAPI.com's Panchang/Muhurat endpoints

Update Frequency

Once daily

Sourced or Calculated

Calculated from Vedic ephemeris data for the date and a single New Delhi reference point (Tithi, Nakshatra, Yoga, Karana, Muhurat windows). Sunrise, sunset and Rahu Kalam shown on a city page are shifted from that Delhi reference by roughly 4 minutes per degree of longitude, for the ~113 cities with known coordinates; other cities show the unshifted national time. Tithi and Nakshatra aren't location-dependent in any way that matters for a single calendar day, so they're shown as-is nationwide. Ekadashi/Pradosh Vrat dates follow the Smarta convention (whichever tithi is active at a fixed reference hour).

Geographic Coverage

National reference computation, location-shifted only for sunrise/sunset/Rahu Kalam as described above.

Historical Data

Not applicable - these aren't numeric metrics to trend.

Rashifal

Source

Prokerala Astrology API

API / Data Endpoint

Prokerala's API, via OAuth2 client_credentials

Update Frequency

Once daily

Sourced or Calculated

Directly sourced - the reading text itself comes from Prokerala, not written or edited by IndiaRealTime. Prokerala allows only 5 requests/minute, so all 12 signs are fetched a few at a time across repeated cron ticks rather than in one batch.

Geographic Coverage

All 12 zodiac signs.

If a Station/Source Doesn't Report

A sign whose fetch hasn't landed yet for today shows its most recent available reading rather than a blank page.

Historical Data

Not tracked - only the latest reading per sign is kept.

Cricket Scores

Source

CricketData.org (CricAPI)

API / Data Endpoint

CricketData.org's currentMatches endpoint

Update Frequency

Polled hourly

Sourced or Calculated

Directly sourced - scores and status as reported by the provider, not ball-by-ball.

Geographic Coverage

International and major domestic cricket; the 10 most recently updated matches shown on the index.

If a Station/Source Doesn't Report

A score can lag the real match by up to an hour between polls.

Historical Data

Not tracked as a trend - last 10 matches only.

Stock Market (Sensex, Nifty 50, NSE & BSE)

Source

Yahoo Finance's public market-data feed

API / Data Endpoint

Yahoo Finance's public quote endpoint

Update Frequency

Every few minutes during trading hours

Sourced or Calculated

Directly sourced.

Geographic Coverage

Sensex, Nifty 50, and individually tracked/watchlisted stocks.

Historical Data

Not currently stored server-side - only live/today figures are shown, since the site doesn't yet hold enough price history for a real week/month/year trend.

Mutual Fund NAVs

Source

AMFI (Association of Mutual Funds in India) for the daily scheme/NAV sync; mfapi.in for historical NAV used in returns and charts

API / Data Endpoint

AMFI's free NAVAll.txt for the daily sync; mfapi.in's free per-scheme API, called on demand

Update Frequency

Once daily, synced shortly after AMFI publishes each evening

Sourced or Calculated

Directly sourced - end-of-day NAV as published, never a live intraday value.

Geographic Coverage

India's ~10,000+ mutual fund schemes that AMFI publishes.

Historical Data

The daily NAV sync itself keeps no local history (overwritten each day); historical NAV for returns/charts is fetched on demand from mfapi.in and cached 24 hours, not permanently stored on this site.

FD Rates

Source

Each bank's own published rates, entered manually

API / Data Endpoint

No API - there is no reliable free per-bank FD-rate API, so rates are admin-maintained (manual entry or bulk CSV import) rather than synced.

Update Frequency

Updated whenever a bank's published rate is manually checked and re-entered

Sourced or Calculated

Directly sourced, entered as published; the maturity calculator itself needs no external data and works standalone.

Geographic Coverage

Whichever banks have been manually added.

Historical Data

Not tracked - current rate only.

Toll Rates

Source

NHAI's own published toll plaza rate schedules

API / Data Endpoint

Imported from NHAI's published rate schedules, not a live API

Update Frequency

Updated on each rate-schedule revision (BOT plazas typically revise annually, ~1 April, per NHAI's own formula)

Sourced or Calculated

Directly sourced - per-plaza, per-vehicle-class rates as published; nearest-plaza lookups use each plaza's published coordinates.

Geographic Coverage

1,000+ NHAI toll plazas across the states NHAI operates in.

If a Station/Source Doesn't Report

A plaza not in NHAI's published list simply isn't shown.

Historical Data

Not tracked - current rate schedule only.

Bank Holidays

Source

RBI's yearly Holiday Matrix (Negotiable Instruments Act declarations)

API / Data Endpoint

Imported once a year from RBI's published Holiday Matrix, not a live API

Update Frequency

Static per-year calendar, checked live

Sourced or Calculated

Mixed: the holiday dates themselves are sourced directly from RBI; "is a bank open today" is calculated live against that imported calendar plus a standing rule for Sundays and every 2nd/4th Saturday, rather than a separately cached yes/no flag.

Geographic Coverage

National holidays plus state-specific regional holidays.

Historical Data

Not applicable - a fixed calendar, not a changing metric.

PIN Codes

Source

India Post's own published records

API / Data Endpoint

Imported as a static reference dataset, not a live API

Update Frequency

Static reference dataset

Sourced or Calculated

Directly sourced.

Geographic Coverage

Every PIN code India Post's own records cover.

Historical Data

Not applicable.

Earthquake Alerts

Source

National Center for Seismology (NCS)

API / Data Endpoint

Scraped from NCS's own live event list at riseq.seismo.gov.in - no public API

Update Frequency

Checked periodically through the day

Sourced or Calculated

Directly sourced; origin time is confirmed already in IST as NCS itself publishes it, not converted or estimated.

Geographic Coverage

Recent events NCS's own list places in or near India.

Historical Data

Recent events only, not a long-term archive.

Weather Alerts

Source

India Meteorological Department (IMD)

API / Data Endpoint

IMD's official CAP 1.2 RSS feed (cap-sources.s3.amazonaws.com/in-imd-en/rss.xml), public domain, no API key

Update Frequency

Checked periodically through the day

Sourced or Calculated

Directly sourced; IMD's own CAP severity levels are mapped 1:1 to the Red/Orange/Yellow/Green colour scale IMD itself publishes.

Geographic Coverage

Active alerts by affected state/region.

Historical Data

Active alerts only, shown until IMD's own stated expiry, not an archive.

How AQI Is Calculated (Worst-Pollutant Method)

India's Air Quality Index combines six pollutant readings, PM2.5, PM10, NO₂, SO₂, CO and O₃, from CPCB-monitored stations into a single 0-500 score. Each pollutant gets its own sub-index on the same 0-500 scale; the overall AQI is the worst (highest) of those sub-indices, not an average of all six. This is CPCB's own official National AQI methodology, and it means a single badly-elevated pollutant, commonly PM2.5, can and does drive the whole number even when every other pollutant is in a much better band. City and state AQI rankings shown across the site (e.g. "cleanest of N tracked cities") are computed fresh from the same live station data on every page load, from whichever cities currently have a recent reading, rather than a separately maintained ranking table that could drift out of sync with the number displayed next to it.

Who Runs IndiaRealTime

IndiaRealTime is independently and privately run. We are not affiliated with, endorsed by, or acting on behalf of the Government of India, the IMD, the CPCB, or any other government or licensing body; we aggregate and republish their public or licensed data in a faster, more accessible form. See our About Us page for more on how the Site is funded and operated.

Accuracy and Corrections

We take reasonable care in aggregating and displaying this data, but underlying sources can themselves be delayed, revised, or wrong, and every figure here should be treated as informational rather than authoritative for high-stakes agricultural, health, travel, or financial decisions - always verify against the original official source when accuracy matters. Spotted a figure that looks wrong? Email contact@indiarealtime.com with the exact page URL and city, and we'll look into it. See our Terms and Conditions for the full accuracy disclaimer.