How the Nanda Devi Raj Jat map was made: data, methods and limits

By the Nanda Devi Raj Jat editors · Published 5 October 2026 · 4 min read

Every number on our interactive route map is either measured from the data, counted from the data, or labelled as unverified. This page explains how, so that you can judge how far to trust it. It also lists the places where the data is wrong or incomplete, because a map that hides its flaws is harder to use well than one that shows them.

Where the data comes from

The three GeoJSON files were supplied to this project. They carry the route lines and the point locations, but they do not record who surveyed them or when, and we have not independently checked them against the ground. Their details were read as supplied: the distances, day labels and place types are stored inside an HTML table in each feature’s description field, which we parsed out. A printed Raj Jat booklet was used for descriptions and legends, but not for any coordinate.

How distances are measured

Lengths are geodesic (great-circle) distances along each line on the WGS84 reference ellipsoid, calculated with the open-source Turf.js library, summed across all segments. The headline figure of 257.41 km is measured on the full-resolution lines, which hold 34,866 vertices. For speed in the browser, the lines drawn on the map are simplified (Douglas–Peucker, with a tolerance of 2.5 m), leaving 6,080 vertices; the simplified lines add up to 256.27 km, about 0.4% shorter. We publish both and always quote the unsimplified number as the headline.

Distances are straight-line along a digitised path. They do not include elevation gain, switchbacks the surveyor did not capture, or ceremonial circuits. The distance most often quoted for the walked path, about 280 km, is shown beside our figure on the map and marked unverified; the two are not reconciled.

How halts are tied to days

Each day’s route has an end point. A halt is attached to that day only if the end point lies within 2 km of it. Spelling differs between files (“Ida badhani” against “Idabanani”), so matching is by position and not by name. Halts with no route ending nearby are left without a day rather than guessed; 6 of the 17 halts in the current data end up that way.

How place categories are made

In the source files every one of the 46 locations is typed simply as a village. The categories you see on the map (village, pass, temple, lake, river, bugyal) are keyword-derived from place names, for example “Kund” for a lake or “Gali” for a pass. They are marked as derived on the map because they are our interpretation, not the surveyor’s.

What the data gets wrong

Days 1 and 2 share the same line. Day 2 is recorded over identical geometry to Day 1 (each 16.21 km), and both carry a rounded 15 km in the source attribute.
One real mismatch. “Bedni to PatarNachoniya” on Day 14 measures 4.87 km on the line against 3.62 km in the source, a difference of about 35%. We have not altered either.
Day 19 looks too long. It carries two line segments totalling 59.6 km. It is more likely that the source groups several return stages under one label than that anyone walks that far in a day.
No elevation. Every height value in the files is zero, so there is no surveyed height for any place.
Inconsistent spelling. One day label reads “Nauto”; it is a slip for Nauti and has been corrected in the display.

Over Days 3 to 19 the measured and source distances agree within about 0.6%. Across all 19 days the gap is about 1.5%, mostly caused by the rounded figures for Days 1 and 2.

Elevation

The files carry no heights, so the map shows none. Where a height appears on this site, as in the Bedni Bugyal and Roopkund articles, it is a commonly published figure and is labelled as such, not something we measured.

What the map deliberately does not show

The map does not show medical posts, police posts, disaster-response teams, shelters, water points, helipads, risk zones or live tracking. No source we hold supports them, and an invented facility on a pilgrimage map could do real harm. If the people who hold real records supply them, they can be added.

Corrections

If you know the route and see an error, please tell us through the contact page. We correct the page or the map, say what changed and keep the date of the change. See also our About page for our editorial approach.

Sources and notes

Map: Interactive route map. Text version: day-by-day table. Data download: main-route.geojson (display geometry).
Libraries: Leaflet, Turf.js, Leaflet.markercluster and Chart.js, all open source.