An analyzer rides a TRAX light-rail train across the valley and back, every day, and has done since December 2014. It is the only measurement covering the southern valley, and the only one that samples where the sources are rather than waiting for the wind to bring them past a tower.
Every measurement is expressed above the cleanest air the train saw that day, so the map shows local enhancement rather than the regional background. Switch the metric to separate places that are usually a little high from places that are occasionally very high.
The highest stretches of track, with the nearest known methane source. This is proximity, not attribution: it does not account for wind direction, and the train passing a landfill and reading high does not prove the landfill emitted it. It is a starting point for asking.
| # | Nearest known source | Distance | Typical | Worst 5% | Days elevated |
|---|---|---|---|---|---|
| Loading... | |||||
Each 2 km segment by year. A segment-year only appears when the train visited it on at least 20 days, so a route change does not read as a trend. Watch for a row that darkens from left to right.
Each point on the track, placed by how high it typically runs against how often it is elevated at all. A persistent source sits to the upper right. An intermittent one has a low typical enhancement and a long tail, so it sits low here but bright on the worst-days map above.
Pick a stretch of track and see its whole life: every visit the train ever made, what the spread of those visits looks like against the rest of the network, when in the day it runs high, and the days it ran highest. The hotspot rows above are clickable and land here.
One point per day the train served this segment. The line is a 90-day rolling median, so a run of high days separates from a single bad one.
How this segment's visits are distributed against every other segment's. A source that leaks steadily shifts the whole distribution right; one that vents occasionally leaves the bulk where it was and grows a tail.
Hour of day, in local standard time. A source that tracks the working day looks different from one that shows up whenever the boundary layer is shallow, which is overnight and early morning.
| Date | Day | Median | Peak sample | Measurements |
|---|---|---|---|---|
| Loading... | ||||
Hour-of-day median enhancement with the interquartile range shaded, by season.
On-track hours per month. The drop through 2022 is the change from 1 s to 10 s sampling.