Air Quality, Weather & Research Dashboard

Current Conditions

Delhi, Delhi, India · 28.614, 77.209

Observed

Loading…

Source: Sample dataset (demo)Resolution: Station-level, hourly

Weather Summary

Current surface observations

Loading…

Source: Sample dataset (demo)Condition:

Pollutant Levels

Hover a card for units and health relevance

Loading…

Source: Sample dataset (demo)

Sources & Methodology

Data provenance, limitations, and how to wire in live sources

Datasets used in this dashboard
  • City / country conditions & time seriesLive: current AQI and pollutant concentrations from WAQI (World Air Quality Index, nearest reporting station); current weather, forecasts, and pollutant history/forecast from OpenWeather. Composite AQI for history/forecast is derived from measured PM2.5 via the standard EPA breakpoint formula.
  • AQI health categories & breakpointsU.S. EPA Air Quality Index technical guidance (2024) — Good through Hazardous, PM2.5 breakpoint table.
  • Adult obesity prevalence (choropleth / cartogram)Illustrative values modeled on CDC Adult Obesity Prevalence Maps (BRFSS, self-reported BMI ≥ 30) reporting patterns.
  • Smoking prevalence & cigarettes/dayIllustrative values modeled on CDC smoking surveillance reporting patterns (BRFSS / NHIS-style indicators).
  • State PM2.5 / AQI annual layersIllustrative values modeled on EPA / AirNow annual summary reporting patterns.
  • U.S. state boundariesUS Census TIGER/Line boundaries via the us-atlas TopoJSON distribution (Natural Earth / Census-derived, 1:10m), served from /api/geography/states.
Methodology notes
  • Observed vs. predicted values are always labeled separately; forecast points carry a widening uncertainty band with lead time (currently modeled for AQI).
  • The bivariate choropleth classifies each variable into terciles independently — a 3×3 grid, not a joint quantile.
  • Environmental and public-health layers shown together are exploratory correlations, never causal claims.
  • City/country conditions, forecasts, and the bivariate choropleth are live provider data (WAQI, OpenWeather), served through a normalized internal REST API (see Network tab, /api/**). U.S. state-level public-health layers (obesity, smoking, correlation research) remain illustrative sample data modeled on CDC/EPA reporting patterns, not live feeds.
  • Missing data is never silently interpolated across incompatible geographies or time windows; gaps are left visible in the underlying series.