Web Maps, Dashboards and Story Maps

For most of history a map was a fixed object: one scale, one extent, one set of choices, printed. The web broke that. An interactive web map lets the reader pan, zoom across scales, toggle layers, click a feature to read its attributes, and filter to what they care about. The cartographer no longer decides the single view; they design a space of views and let the reader navigate it.

Under the hood, a web map is a stack much like your desktop project. At the bottom sits a basemap delivered as small pre-rendered image or vector tiles that load only for the area on screen - which is what makes a global map feel instant. On top sit your data layers, delivered as GeoJSON for small datasets or as vector tiles for large ones. A popup appears when the reader clicks a feature, showing its attribute row. Understand that stack and every web-mapping tool stops being mysterious.

You do not need a paid platform to publish a professional web map. Leaflet is a small, friendly JavaScript library for 2D maps - a few lines produce a pan-and-zoom map with markers and popups, and it is the fastest way to get a GeoJSON layer online. MapLibre GL JS is the open-source engine for smooth, WebGL vector-tile maps - the free, community-governed fork of Mapbox GL JS. OpenLayers is the heavyweight, strong on OGC services (WMS, WFS, WMTS). For very large data and analytical layers, deck.gl renders millions of features on the GPU, and kepler.gl wraps it in a no-code app you can drop a file into.

Mapbox GL JS itself is a paid service model - it needs a Mapbox account and token and its licence is proprietary from v2, so where you want fully open and self-hostable, MapLibre is the like-for-like replacement. On the data side, prefer open, standard delivery: OGC API - Tiles and the older WMTS for tiled maps, OGC API - Features (the modern successor to WFS) for queryable vector data, and GeoJSON for simple hand-offs.

Sometimes the reader does not need a story - they need to monitor. A dashboard puts several linked views of one dataset on a single screen: a map, a few charts, some headline numbers (KPIs), and filters that drive them all at once. Pick a ward from a dropdown and the map zooms, the bar chart re-sorts and the totals update together. It is the right tool when the data changes over time or the reader has questions you cannot predict.

On the Esri path, ArcGIS Dashboards builds this from a hosted web map with no code - you add map, list, chart and indicator elements and wire them to filters. On the open path there is no single equivalent, but you get there by combining pieces: a MapLibre or Leaflet map beside charts in a web page, or - fastest of all - kepler.gl, where you load a GeoJSON or CSV and get filters, a map and simple charts without writing anything. The design lesson from earlier still rules: even a live dashboard needs hierarchy, a sane colour scheme and honest classification.

A dashboard is for exploring; a story map is for telling. It pairs a scrolling narrative with maps, images and text, so that as the reader moves down the page the map changes to match the words - the city in 1990, the sprawl by 2010, the plan ahead. It is the format for a proposal, a public consultation, a before-and-after case for a design intervention.

ArcGIS StoryMaps is the best-known authoring tool - a template-driven builder that links narrative blocks to map states, again with no code. On the open side you build the same effect with a scrolling web page that swaps a MapLibre or Leaflet map's layers and extent as the reader passes each section (the technique often called scrollytelling). Reach for a story map when the sequence of maps carries the argument; reach for a dashboard when the reader should pick their own path. Both beat a stack of static slides for anything you want people to actually explore.

Going live adds duties a PDF never had. Performance: simplify geometry, serve large layers as vector tiles rather than one giant GeoJSON, and your map stays fast on a phone on mobile data - which is how most of India will open it. Attribution and licence: an OpenStreetMap basemap must carry the ODbL attribution, and if you publish anything derived from Survey of India Open Series Maps you are bound by their non-commercial terms - honour them or do not publish. Accessibility: colour-vision-safe palettes, readable text, and a fallback for readers who cannot use a mouse.

And note where your data lives. On ArcGIS Online your layers sit on Esri's cloud under your organisation's account; with the open stack you host the files yourself, which is more work but keeps you in full control and inside India's data-handling expectations under the National Geospatial Policy 2022. Neither is wrong - but decide it deliberately, because publishing a map is publishing the data behind it.