Live Spatial Data Connector
Paste a public data endpoint — an earthquake feed, an open-data GeoJSON, a CSV with latitude/longitude columns — and this connector renders it on the map. Refresh re-reads the source, so the map is as live as the feed itself.
By Aryan Khan, Founder & Editor · Updated October 2026
Data is fetched directly by the browser when you connect or refresh; CORS must allow access.
How to use Live Spatial Data Connector
No uploads, no server of ours — your browser talks to the feed directly:
Fetch a compatible GeoJSON Feature or FeatureCollection, or a CSV with latitude and longitude headers, directly in the browser. Refresh is manual; the endpoint receives the request and its CORS rules apply.
- 1Paste a GeoJSON or CSV endpoint URL (or pick a built-in example).
- 2Press “Connect”.
- 3Press “Refresh data” any time to re-read the source.
What people use Live Spatial Data Connector for
1. Earthquake and weather watching
Public science feeds (USGS earthquakes, river gauges, storm tracks) were published to be mapped. This connector is the two-minute route from feed address to situational picture.
2. Open-data exploration
City and national open-data portals offer GeoJSON endpoints for everything from trees to planning applications. Preview one here before building anything around it. Static files open in the GeoJSON Viewer instead.
3. Checking a feed before you build on it
Developers: paste the endpoint and see what actually comes back — geometry types, coverage, freshness — before writing a line of integration code.
4. Pairing with your own field data
Collect points with the Map Data Collection Form Builder, publish or host the export, and watch it beside the public feeds it relates to.
Worked examples
The USGS 'all earthquakes, past hour' GeoJSON feed regenerates about every minute. Connect it on a quiet morning and the map still shows a dozen small quakes — the Earth is never actually still, and a live feed proves it.
A 'live' city dashboard endpoint turns out to be a file someone exported in March. The connector shows the data faithfully — the timestamps inside it confess the rest. Liveness is a property of the source, not the map.
What the connector accepts
Two honest formats, fetched straight from the source by your browser.
| Source | Requirement | Typical publisher |
|---|---|---|
| GeoJSON endpoint | A URL returning a FeatureCollection | USGS, open-data portals, ArcGIS services |
| CSV endpoint | Columns recognisable as latitude and longitude | Data dashboards, GitHub-hosted datasets |
| Refresh | Re-fetches the same URL | You control when — the map is as fresh as your last refresh |
| Not accepted | Feeds behind logins or keys | Those belong in the publisher's own app |
Honest limits: the fetch happens in your browser, so the endpoint must allow cross-origin access (CORS) — many public feeds do, some do not, and the tool cannot (and should not) sneak around a publisher who said no.
How to interpret the result
The most common errors come from using a correct result to answer the wrong geographic question. Keep these points in mind for Live Spatial Data Connector:
- Choose colour for category, not decoration: two or three high-contrast hues read instantly; a rainbow reads as noise. Labels carry identity, colour carries grouping, and the viewport should crop tightly to the subject.
- Export twice: an image for the slide and a data format (GeoJSON/CSV) for the record. The image communicates; the data survives editing, and the pair keeps your map honest and reusable.
Limits & appropriate use
Custom maps inherit two honest limits: the basemap's and the message's. The basemap is OpenStreetMap's current picture with its ODbL attribution ask, and its gaps — new roads, thin rural detail — travel into every export that includes tiles. The message limit is design's classic one: a map can carry one argument well, and beyond a few dozen symbols the right instrument changes to data-driven mapping with clustering and categories. Both limits are stated where they matter, because the alternative is a beautiful figure that misleads its own author.
Within those bounds the workflow is complete and durable: pins, labels, colours, tight framing, pixel-ratio PNG for the slide, GeoJSON/CSV for the record, URL-state for the living version. Escalation is gentle by design — choropleths and territory shading belong to GIS with real boundaries (the blank-map library and GeoJSON exports feed it), cartographic print production wants vector pipelines, and brand-strict design systems want tokens this tool deliberately doesn't fake.
- Data-dense maps → CSV-to-map with clustering and category colour.
- Boundary shading/choropleth → GIS with census or OSM boundaries.
- Print cartography → vector exports into a design pipeline.
Data & methodology
Direct browser fetch with GeoJSON FeatureCollection rendering, or CSV parsing keyed on latitude/longitude columns.
For source details, licences and model limitations, see Data Sources and Methodology.
Key terms worth knowing
- Cluster marker
- A grouped dot representing many nearby points at low zoom.
- Share-state URL
- Encoding the whole map configuration in the link itself.
Editorial guidance: MapBench explains the method and limitations so the tool can be used as a transparent reference, not as a black-box answer. For legal boundaries, safety-critical routing, surveying, engineering, regulated decisions or official statistics, verify the result with the responsible authority or a qualified professional.
Category: Map Creation. Results are intended for everyday analysis and planning unless a page explicitly states otherwise.
Method, assumptions & limits
- This tool’s method
- Direct browser fetch with GeoJSON FeatureCollection rendering, or CSV parsing keyed on latitude/longitude columns.
- Before relying on the answer
- Inspect the labels, map extent and export options in the workspace. Test the exported or shared version rather than assuming it matches your current screen.
- What the result cannot establish
- Map styling does not verify a location or grant rights to third-party data. Keep attribution with any reused basemap or dataset.
Coverage: Worldwide. Search, basemap and external data requests may use third-party services; browser-based does not mean every operation is offline. Read the source and licence notes, calculation methodology and privacy policy. If a result looks inconsistent, report a correction with the inputs and mode used.
Frequently asked questions
Clear answers about the calculation, accuracy and appropriate use.
Why did my endpoint fail?
Usually CORS: the data owner must allow browser access, and no viewer can bypass that honestly. Public open-data feeds (like the USGS example built in) generally allow it.
How “live” is it?
The map re-fetches whenever you press Refresh; freshness depends on the publisher's own update cycle.
Can I share the map or output?
Many MapBench tools support shareable URLs or exports. Check the tool controls and the source-data licence before republishing externally.
Which export format should I choose?
Use PNG for presentation, CSV for tabular data, and GeoJSON/KML/GPX when geographic geometry needs to remain editable.
Will a shared link preserve my settings?
Tools that store state in the URL can reopen that configuration. Sensitive or private data should not be placed in public URLs.
Does MapBench store the feed or its data?
No. Your browser requests the endpoint directly and renders what returns; nothing is proxied through or kept on a MapBench server. Close the tab and the connection is gone.
What kinds of feeds work best?
Small-to-medium GeoJSON FeatureCollections and tidy CSVs with clear latitude/longitude columns. A national dataset with a million features will make any browser sweat — use a bounded or summarised endpoint where the publisher offers one.
Can I connect a private or key-protected API?
The browser makes a direct fetch and provides no dedicated credential controls. Avoid placing secrets in URLs; the endpoint and browser history may expose them.
How do I know the data is current?
Read the feed's own timestamps — good feeds carry generated-at or per-feature times. The connector renders faithfully; judging freshness is reading, and the map makes stale data easy to spot.
Why does the map show nothing for my CSV?
Usually the coordinate columns: headers the parser cannot recognise as latitude/longitude, coordinates in the wrong order, or a delimiter surprise. Open the endpoint in a tab and look at the first rows — the diagnosis is almost always visible.
Can I style the points by a data field?
No. The connector has fixed styles; the area dashboard also does not provide field-based styling.
Does it work with KML feeds?
KML files you have belong in the KML Viewer. For a hosted KML endpoint, many publishers also expose the same data as GeoJSON — that is the version to connect here.
Related tools and resources
Open static files with the GeoJSON Viewer or KML Viewer, and create your own feed-worthy data with the Map Data Collection Form Builder.
More free tools
Sources, methodology & editorial trust
- ✓ Browser-first processing where supported
- ✓ External sources are documented
- ✓ No account required for public tools
- ✓ Review data sources
Calculations are documented with their assumptions, limitations and source context rather than presented as unexplained answers.
Reviewed and maintained by Aryan Khan, Founder & Editor of MapBench.
Part of our map maker hub
Live Spatial Data Connector sits in the map maker cluster, alongside 26 other tools that answer neighbouring questions with different methods or datasets.
You may also need
When this answer is only part of the job, these clusters cover the next step.
Read the guide behind this tool
Free Printable Blank Maps: How to Choose Outline, Labeled, or City Variants — Classroom worksheets, sales territories, and presentation slides need different map styles. How to pick blank vs labeled vs coloured — and SVG vs PNG.