Timeline & Temporal Playback
The timeline bar at the bottom of the viewer controls time-based data playback. It gives every time-aware dataset its own labeled track, so you can see multiple datasets at once and navigate their temporal extents independently.
Timeline bar
The timeline bar sits at the bottom of the viewer. Click the collapse button on the right to hide it and reclaim screen space; click again to restore it.
| Control | Description |
|---|---|
| Play / Pause | Starts or stops time animation. |
| Reset to start | Jumps the playhead back to the beginning of the focused time range. |
| Loop toggle | When enabled, playback restarts from the beginning when it reaches the end. |
| Speed dropdown | Choose how fast time advances from a compact dropdown. Presets: 1×, 10×, 1 min/s, 10 min/s, 1 h/s, 1 d/s, 1 mo/s, 1 yr/s. |
| LIVE toggle | Appears when one or more live datasets are loaded or wind is enabled. Pause or resume automatic refresh of all live data. See Live updates. |
| Current timestamp | Shows the exact date and time at the playhead position. |
| Collapse button | Minimises the timeline bar to save screen space. |
The date picker is hidden for span-based layers where setting an exact start date does not apply. It remains visible for point-in-time and building dataset layers.
Public share links include a simplified timeline with playback, loop, speed, and scrub controls. Dataset tracks are hidden in the public viewer to keep the shared map focused, but the playhead still controls time-aware layers.
Stacked dataset tracks
Each time-aware dataset gets its own labeled lane stacked vertically below the main scrubber. Lane colours are assigned by activation order so you can quickly tell datasets apart at a glance.
The track for each dataset shows the span of time covered by its data. Hover over a track to see the layer name and the temporal extent in a tooltip.
Out-of-range markers
If a dataset has data outside the currently visible portion of the timeline, a small marker arrow appears at the left or right edge of its track. This lets you know data exists there even when the current view is zoomed in on a different period.
Snap to extent on activation
When you activate a time-aware dataset — either by toggling it on in the Data panel or clicking its track in the timeline — the timeline automatically jumps to that dataset's temporal extent. This saves you from manually scrolling the timeline to find where data starts.
Playhead and window band
A single playhead marks the current simulation time shared across all datasets. Drag the playhead left or right to scrub through time.
Around the playhead, a band halo shows the currently active time window — the span of time that span-based layers use for their temporal filter. The halo is visual only; resize the window using the window width presets below the tracks.
Window width presets: 1 h / 1 d / 1 w / 1 m / 1 y — click a preset to rescale the visible timeline to that width, centred on the playhead. The start and end labels update immediately. Near the latest data, the window clamps so it does not run past the present.
The old draggable Time Window Slider has been replaced by the band halo around the playhead. The window width presets work the same way as before.
Building dataset tracks
Building datasets use single-instant semantics: the attribute data they carry is not time-varying in the same way as CZML or filtered GeoJSON layers. The timeline shows their track as a single marker under the playhead rather than a span bar.
Building dataset colouring is not affected by the window filter that controls which span-based features are shown. The buildings remain visible regardless of the active window width.
Temporal filtering for vector layers
Vector layers (GeoJSON, CSV, Shapefile, and similar) can be filtered by a date/time attribute field. Features outside the active time window are hidden from the map.
What "outside the window" means depends on how the layer's dates are configured — see What a date means below.
Configuring a temporal field
- Open the layer in the Data panel and expand its settings.
- Open Advanced → Time (temporal data).
- Select which attribute field contains the time or date data. Auto-detected temporal fields appear at the top with a confidence percentage and an Auto-detected badge.
- Optionally select an End date field. With one, each feature spans
[start, end]and is shown whenever that span overlaps the window — a feature with a blank end value stays visible from its start onwards. - Choose the date format: ISO 8601, Unix seconds, Unix milliseconds, or Compact (YYYYMMDD).
- Pick what a single date means — hover either button for a one-line description, or see What a date means below. This row is hidden when an End date field is set, because the span already states both ends.
- Below the controls, the panel shows the detected date range and how many of the layer's dated features the current window is showing ("312 of 340 dated features in the current time window"). If that reads 0, the layer is not broken — the window is simply somewhere its data isn't. Features whose date is blank or unreadable are not counted in either number.
- Click Clear to disable temporal filtering for the layer. Clearing keeps the auto-detected field list, so you can pick a different field afterwards without re-importing.
Auto-detection skips record-keeping dates — fields that record when the row was last edited or exported, not when anything happened, such as oppdateringsdato or dataUttaksdato in Geonorge/SOSI downloads. They still appear in the dropdown: pick one by hand and it is applied, with a warning that the timeline will show features appearing as the data was maintained rather than as things happened. Analytics charts apply the same rule to their default time axis, unless the layer's temporal field was picked by hand.
What a date means
When a layer has a start date but no End date field, the Shows as switch decides how that single date is read:
- From its date on (the default) — each feature appears on its date and stays visible as the timeline moves forward. This is right for things that come into being and remain: a hazard zone mapped on a survey date, a building completed in a given year. A dataset whose dates cluster on a handful of maintenance stamps would otherwise appear one cluster at a time, with nothing visible in between.
- On its date only — each feature is visible only while the time window covers its date. This is right for repeated readings of the same thing: a stream-gauge measurement, a solar-wind sample, a storm position. Reading these as "from its date on" would keep every past sample on screen and pile them onto a single coordinate instead of animating them.
Live catalog sources that publish repeated observations — USGS stream gauges, NOAA solar wind, NHC active storms, NASA EONET events, USGS volcano status — ship set to On its date only. Event feeds such as earthquakes and GDACS alerts use the default, since each record is a distinct event that happened.
If a layer looks emptier than expected, check the feature count under the field selector first. "0 of 340 dated features" means the time window is in the wrong place, not that the import failed — press Fit all on the timeline, or scrub toward the data.
The timeline is most useful with datasets that have a date or timestamp field — for example, construction phases, historical events, or sensor readings. Enable temporal filtering on the layer first, then use the playhead to step through time.
Temporal series playback
Point layers (GeoJSON or CSV) that carry multiple time-stamped samples per entity — such as sensor readings from a network of stations — can be played back as an animated Temporal Series instead of using the time-window filtering described above. In Series mode, the map shows each entity's most recent sample at the current clock time, and Point Marker, 3D Bar Chart, Heatmap, and 3D Model styles all animate as the timeline plays. Enabling a temporal series automatically sets the timeline's clock range to the data's full time span. See Temporal Series for how to set it up.
Live updates
The timeline bar shows a LIVE toggle whenever one or more live datasets are loaded or wind visualization is enabled. Live datasets are layers added from the Live tab in the data-search panel — for example, earthquake feeds, NASA imagery, flight positions, or FX rate overlays. Many live layers carry no temporal field; they refresh their content rather than advancing a timeline.
When the LIVE toggle is active, its icon pulses and all live data (including wind) refreshes automatically in the background. Click the toggle to pause refreshing; click again to resume. The toggle enables itself automatically when you add your first live dataset.
When a live CZML stream is active, a pulsing LIVE badge also appears on the timeline bar and the playhead follows real-time data.
Wind is one of several live sources. Enabling wind visualization alone is enough to show the LIVE toggle. Adding any live dataset from the Live catalog has the same effect.
Animate through time
Click Play to advance the playhead automatically. The speed dropdown controls how fast time moves:
| Preset | Meaning |
|---|---|
| 1× | Real time — one second of simulation per second of real time |
| 10× | Ten seconds of simulation per second of real time |
| 1 min/s | One minute of simulation time per second of real time |
| 10 min/s | Ten minutes of simulation time per second of real time |
| 1 h/s | One hour of simulation time per second of real time |
| 1 d/s | One day of simulation time per second of real time |
| 1 mo/s | One month of simulation time per second of real time |
| 1 yr/s | One year of simulation time per second of real time |
Use Loop to repeat playback continuously — useful for presentations or unattended displays.
When a scenario snapshot is saved while playback is active, the scenario remembers the playback state and selected speed. Public share links restore those settings, so recipients see the same time animation you prepared.
Temporal filtering requires the layer to have an attribute field with date/time values. Layers without temporal data are not affected by timeline playback.