Two measurements taken a few hundred metres apart can differ substantially. Air pollution is far less evenly mixed than most people assume, particularly close to its sources.
Traffic pollutants dilute quickly with distance
Nitrogen dioxide and fresh exhaust particles are produced at road level and disperse rapidly, so concentrations fall sharply within a short distance of a busy carriageway.
This is why a monitoring station beside a main road and one in a park nearby report different results without either being wrong.
It also means that where a station is placed determines what it measures, and comparisons between cities require knowing the siting convention used in each.
Street shape traps or clears the air
A narrow street lined with tall buildings forms a canyon in which air circulates rather than escaping, allowing emissions to accumulate through the day.
The circulation pattern inside a canyon can concentrate pollution on one side, so the two pavements of the same street differ measurably.
Open junctions and wide boulevards disperse far more effectively, which is why an apparently busier road can show lower concentrations than a quiet enclosed one.
Weather governs dispersion more than emissions do
Wind speed determines how quickly emitted material is carried away, and calm conditions allow concentrations to build even when traffic is unchanged.
Temperature inversions place a layer of warmer air above cooler air near the ground, preventing vertical mixing and holding pollution close to the surface for days.
Because of this, the worst episodes tend to occur in still, cold weather rather than at the times of heaviest traffic.
Particles behave differently from gases
Fine particles travel long distances and mix over wide areas, so background levels can rise across an entire region from sources far away, including agriculture and heating.
Coarser particles from brake wear, tyres and road dust remain closer to the source and are stirred up by passing vehicles rather than emitted from exhausts.
This distinction matters because measures targeting exhaust emissions address only part of the total, and the non-exhaust share does not decline as vehicles change.
Sensor type affects the number
Reference monitoring stations use expensive instruments with defined calibration procedures. Low-cost sensors are far more numerous but less accurate, particularly in humid conditions.
Dense networks of cheap sensors show spatial patterns well while reporting absolute values that may drift, so the two types answer different questions.
Anyone with a respiratory condition acting on these readings should discuss thresholds with a clinician, since individual sensitivity varies far more than the measurements do.