Census, Land Records & Field Survey

The Census of India is the decadal count of the population, and it is the richest free source of who-lives-where data in the country: population, age, sex, literacy, household amenities and more, published down to the village and ward. In a GIS you join these numbers to boundary polygons - state, district, tehsil/taluk, town, village, ward - and suddenly you can map density, dependency, deprivation and demand.

The catch is that the Census publishes tables; the matching digital boundaries are a separate problem. Official administrative boundaries exist, and community efforts have also released census-linked boundary sets - the widely used Datameet community boundaries, for example, are shared under a Creative Commons Attribution licence (CC BY 2.5 India). Whichever you use, the discipline is the same: join the demographic table to the boundary layer on a shared code, and always name your boundary vintage, because ward lines change between censuses.

The moment you map a rate - people per hectare, literacy percentage - by census area, you inherit a subtle statistical trap. The same underlying reality looks different depending on how you drew the boundaries and at what level you aggregated. Merge two wards and an apparent hotspot can vanish; split them differently and one appears. This is the Modifiable Areal Unit Problem (MAUP), and it is not a curiosity - it can flip which ward looks most deprived, and therefore where investment goes.

The practical defences are honest ones. Report at the finest level your data supports, not a coarse one that hides variation. State the unit you used. Be wary of comparing rates across areas of wildly different size or population. And never let a striking choropleth map make a decision that the aggregation alone could have manufactured. Census data is indispensable; it is also easy to abuse, and a good analyst says so out loud.

Where the Census tells you about people, the cadastre tells you about parcels - the legal map of who owns which piece of land, tied to the Record of Rights (RoR). India's programme to digitise and geo-reference this is the Digital India Land Records Modernization Programme (DILRMP), and Karnataka's pioneering Bhoomi system is the best-known state example. A modern goal is the ULPIN - a unique land-parcel identification number, a sort of Aadhaar for a plot - that ties a cadastral polygon to its ownership record.

Here is the honest India caveat. Cadastral access is decentralised and guarded: there is no single national open-download portal for parcel geometry. Records are held state by state, usually accessed one parcel at a time through a state revenue portal, often for a fee, and bulk geometry is generally not open. So for a plan you can often see that parcels exist and read an individual RoR, but assembling a clean, complete cadastral layer for a whole area is frequently the hardest, most manual data task in Indian GIS.

Rural inhabited land - the abadi area of a village - was historically un-surveyed at the parcel level, leaving millions of households without a formal property document. The SVAMITVA scheme (Ministry of Panchayati Raj) is closing that gap by flying drones over village abadi areas to map every rural property and issue property cards to owners.

SVAMITVA matters to this course for two reasons. First, it is creating a genuinely new cadastral layer for rural India where none existed. Second, it is the clearest official proof that drone survey has become mainstream, authoritative geospatial practice in India - not a novelty. The property cards go to owners, and the resulting spatial data feeds the wider land-records ecosystem, though access to the geometry, like the rest of the cadastre, is administered rather than openly downloadable.

Sometimes no dataset exists at the grain you need, and you must collect it - field survey. At the simplest that is a smartphone or handheld GPS capturing points (a tree, a bore-well, a broken drain) with attributes typed in on site; for accuracy it is a differential GPS or total station; for area coverage it is a drone flying a grid to produce an orthophoto and a fine surface model. Field data is the only data whose accuracy and currency you fully control - and the only data you are responsible for licensing correctly when you share it.

That brings in India's data-policy layer. Public government datasets are released under the National Data Sharing and Accessibility Policy (NDSAP 2012) through the OGD platform data.gov.in, under the Government Open Data License - India (GODL-India). Your own acquisition of geospatial data and its sharing sit under the National Geospatial Policy 2022 and the 2021 DST Guidelines that liberalised geospatial data for Indian entities. The takeaways are practical: know whether a dataset is GODL-India (reusable with attribution) or restricted; and know that flying a drone and sharing the map you make from it happens within the geospatial guidelines, so read them before you publish.