Sunday, January 29, 2017

Air Quality in Delhi (daily updates)

This is modeled source contributions to the ambient particulate pollution (PM2.5) in South Delhi district (one of the 14 districts in the National Capital Region (NCR) of Delhi). The calculations are conducted in forecast mode. More NCR district reports @ Delhi Air Quality.Info

  • DUST.ERO = wind blown dust from the dry and arid regions, dependent of hourly meteorological conditions
  • OPEN.FIRES = open biomass burning (both agricultural lands and forest areas), a seasonal affair linked to dry conditions and agricultural clearing patterns
  • TRANS.PASS = contribution of passenger vehicles (2Ws, 3Ws, 4Ws, Taxis, and Buses)
  • TRANS.FRGT = contribution of freight vehicles (heavy and light trucks, and non-road vehicles)
  • RD.DUST = contribution of re-suspended dust on the roads due to vehicle movement and construction activities
  • PP.DGS = contribution of power plants and in-situ diesel generator sets
  • INDUS = contribution of industrial activities
  • HOUS.HLD = contribution of domestic cooking, space heating, water heating, and lighting
  • WST.BURN = contribution of open waste burning
  • OTHERS = contribution of aviation, shipping, biogenics, etc..
A number of source apportionment studies were conducted in Delhi (more in Delhi than in any of the other Indian cities). We summarized the known particulate pollution source apportionment studies, as an open article, "what's polluting Delhi's air". A snapshot of the shares from one of the studies is presented below.

The modeled particulate pollution in the forecast mode for the next three days is presented, as time series for SO2 and Ozone for the district and as an animation for all PM2.5 covering the modeling domain. Similar animations and daily average concentration maps for all the criteria pollutants are available @ Delhi Air Quality Info

See what is happening at the regional scale, which is conducted as part of the all India air pollution forecasting program, hosted @ The animation below is from a WRF-CAMx simulation conducted @ 0.25x0.25 degree resolution (approximately, 25km x 25km).

The pollution patterns change every hour and every day, depending on the prevalent meteorological conditions - wind speed, wind direction, temperature, and precipitation. Under windy conditions, most of the emissions get dispersed to farther places; Under rainy conditions, most of the emissions get drained out. Want to see how the weather pattern is holding up for the next three days in Delhi. Check out @ Below is an animation of the anticipated wind speeds and wind directions from the WRF meteorological model - also used by IMD for their forecasts.

The monitoring data from the DPCC stations reported as an air quality index by AQICN is as follows for one of the stations


Pawan Gupta said...

Can you point to any validation effort of these forecast fields? Also, what is the confidence level and how far in future that confidence remains in these forecasts?

Sarath Guttikunda said...

Hindcast results and the comparison will be hosted at We forecast for 72 hours, which is 100% driven by met forecasts. Confidence level drops significantly after 48 hours (2 days). We run the system every 24 hours, although the GFS fields are available every 6 hours - this is a computational resource question than anything else.

Gs Virdi said...

From where can I get air pollution data of new delhi (Jan-Sep 2016 or 2015)? Monthly data would be good.

Sarath Guttikunda said...

You can download the data from

Open archive @