Miami Sensor Network
Environmental Change Laboratory, FIU · FIU GIS Center
Overall status
Air quality
Connecting to sensor network…
Sea breeze, sun, rain, and humidity can shift pollution minute by minute.
How to use this section
This chart shows how particulate pollution, wind, and delayed NO₂ chemistry context changed over the selected time window.
Click the chart to add a note at that time · Notes sync with the list below
Similar timing can suggest a research question, but this chart alone does not prove a source or cause.
Long-term NO₂ and O₃ from the FIU Pandora spectrometer (Pandonia network). Charts below act like a compact data manual: time series, diurnal cycle by local hour, and vertical profile. The spectrometer points toward the sun for retrievals — that orientation is not the same as the wind rose under Drivers. Reference: Pandonia BLICK viewer (external).
The Sentinel-5P/TROPOMI instrument is a low-Earth-orbit UV-visible spectrometer with daily global atmospheric composition observations, including NO₂ column products. In these charts, TROPOMI labels are timing references, not local ground-sensor data.
TROPOMI provides NO₂ column products from orbit. Learn more from ESA/Copernicus Sentinel-5P NO₂ documentation.
TEMPO is NASA's geostationary air-quality instrument. It observes North America from a fixed orbital view and provides hourly daytime observations for products including NO₂. In these charts, TEMPO labels are satellite observation timing references.
TEMPO provides hourly daytime air-quality observations over Greater North America. Learn more from NASA Earthdata TEMPO documentation.
Pandora is a ground-based passive UV-visible spectrometer in the Pandonia Global Network. It provides delayed atmospheric trace-gas retrievals, including NO₂ tropospheric vertical column density and O₃ total vertical column density.
⏱ Delayed — not suitable for real-time assessmentThis chart shows daily Pandora NO₂ or O₃ column amounts over weeks to years.
Use this for longer-term chemistry context, not minute-by-minute current conditions.
Scroll to zoom · drag to pan · double-click to reset · select "All data" to see history since Dec 2023
Hour-of-day pattern for Pandora vertical column products (mean ± variability over the selected window).
Toggle Retrievals to hide scatter for a cleaner mean-only view. Surface and some species may have shorter history (PGN file end dates).
TROPOMI reference ≈ 13:30 EDT (17:30 UTC) · TEMPO daytime observation reference ≈ 12:00 EDT (16:00 UTC)
This chart shows how Pandora-derived NO₂ layer summaries vary by time of day and approximate altitude.
Use this curtain plot as atmospheric-structure context, not standalone proof of vertical transport or source height. If only one vertical slice is filled, the available retrievals are concentrated in that hour.
Color = mean NO₂ partial column for one retrieved layer during that hour of day (EDT). Blank areas = no available layer/hour summary. TROPOMI reference ≈ 13:30 EDT · TEMPO reference ≈ 12:00 EDT.
All parameters for one instrument over the selected time window. Tempest shows every channel from WeatherFlow.
About instrument channels
Each instrument reports different variables at different cadences: PurpleAir focuses on particulate matter, Tempest reports weather channels, and Pandora reports delayed atmospheric chemistry columns. Missing data can reflect reporting cadence, latency, source availability, or filtering.
Precipitation events with PM2.5 before and after — green delta means PM2.5 was lower after the rain window.
Click the chart to add a note at that time
Wet scavenging summary — for each detected rain event, was particulate matter lower in the hour after vs before?
Bar length = frequency from that direction. Color = average PM2.5 from those winds.
Daily mean PurpleAir PM2.5 at FIU MMC — each point is one calendar day, not an hour-of-day average.
For minute-by-minute PM2.5, use Time Series → Summary or Per Instrument → PurpleAir. Hour-of-day patterns are under Chemistry (Pandora diurnal).
For researchers
How to use these questions
These examples show analysis directions the dashboard can support. Treat them as starting points: each question still needs quality control, time-window selection, and supporting context before drawing scientific conclusions.