2001 village and town boundaries

[Mark Montgomery]

Let me introduce myself to the group in this way: I am an Economics professor at Stony Brook University in New York, with a long-time interest in Indian urbanization. I am also keen to see as much as possible of the spatial and socioeconomic detail on urbanization placed in the public domain. Toward that end, colleagues and I have been knitting together the 2001 and 2011 primary census abstracts (PCAs) that the Indian census authorities have made available on the census website and incorporating published data from the District Census Handbooks, all of these at the level of individual settlements with coverage of wards for the PCAs. Our aim is to create an integrated and publicly-accessible database based only on publicly-available sources. As you would know very well, the spatial side of the task is more challenging for 2001 than 2011.

At the moment, I seek your guidance on the remarkable DataMeet collection of polygons for villages, census towns, and statutory urban centers, to which a number of you have contributed months or even years of effort. I have linked your spatial records to the PCA identifiers (including subdistrict and district) and in the process have come across some issues (mainly concerning the vintages of the maps that were used, and various oddities regarding identifiers) that some of you may know about. My own spatial work uses R, but I am happy to share these results with the group in other spatial formats (for instance, as geojson or geopackage files). The next steps I have in mind are to compare the DataMeet polygons with the often-mentioned Meiyappan et al. (2018) polygons that have been publicly available at the Socioeconomic Data Applications Center (SEDAC) site since 2018, and with a lesser-known but evidently high-quality collection of 2001 point coordinates for villages and some hamlets assembled by a University of Tokyo history professor and available on his website.

I'm attaching a short pdf that explains these three public-domain sources (with links to the SEDAC and Univ. of Tokyo sources, and with a critical review of aspects of those spatial datasets), and which in particular lays out some of the issues I've encountered with the DataMeet collection. (I've yet to get to grips with the Karnataka data for 1991, and with the Rajasthan data that I believe are for 2011 or later.) I would be really grateful for criticism and suggestions!

[Sharad Lele]

Dear Mark,

Happy to know about your interest. I am also quite interested in these issues, having worked on 1991, 2001 and 2011 census datasets and their spatial representation (at least for Karnataka and some other states). There are many issues, both with the census datasets themselves and with the spatial boundary datasets released by Meiyyappan et al. I may not be able to lay out everything immediately, because of being in the throes of some deadlines, but hope to go through your writeup and respond a bit later--maybe mid-June, if that is okay with you.

Best,

Sharad

[Digvijay Bendrikar Shinde]

Hello Prof Mark,

Thank you for the resource.

Have you seen the India GIS portal? http://www.censusgis.org/india/ it has the census data of 2001 and 2011 integrated with (up to) Village level shapefiles. you can make basic spatial viz using this. But files can not be downloaded.

Also, CSE department, IIT Bombay has put Maharashtra state's Census '11 data integrated village level shapefiles here https://www.cse.iitb.ac.in/~pocra/MahaCensus_shapefile_data1.2/Boundary.html

Hope this helps in you.

Regards,

Digvijay

PhD Scholar

CTARA, IIT Bombay

--
Datameet is a community of Data Science enthusiasts in India. Know more about us by visiting http://datameet.org
---
You received this message because you are subscribed to the Google Groups "datameet" group.
To unsubscribe from this group and stop receiving emails from it, send an email to datameet+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/datameet/515058c6-e520-405c-8611-e7cfc94a4e91%40googlegroups.com.

[Cara Foss Arellano]

Dear Mark,

Super interesting project! May I schedule a brief call with you sometime this week? 

-Cara Foss Arellano

--
Datameet is a community of Data Science enthusiasts in India. Know more about us by visiting http://datameet.org
---
You received this message because you are subscribed to the Google Groups "datameet" group.
To unsubscribe from this group and stop receiving emails from it, send an email to datameet+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/datameet/faee6884-40af-448b-9b0e-5b1663613cc3%40googlegroups.com.

--

Cara Foss Arellano

cara.a...@berkeley.edu

Interdisciplinary Studies 

Fair Validation for grassroots NGOs 

Mobile/WhatsApp: +01(408)393-5646

in...@daanmatch.com

[Mark Montgomery]

Many thanks to Digvijay and Sharad, and I hope to continue sharing ideas and data with you.  I'm excited to hear about the census gis site and the availability of integrated files for Maharashtra; I did not know about either resource.

One issue that I am concerned about is whether it will be possible to create boundary files for 2001 villages and (especially) statutory towns that respect subdistrict borders. The aim is to link the spatial boundaries as closely as possible with the PCAs, which provide the detailed settlement-within-subdistrict counts. I doubt that this can be done perfectly with the spatial data we have, but it might prove adequate to overlay the 2001 subdistrict boundaries (one version of which was supplied on Justin Elliot Meyers' Github, announced to this group some time back) on top of the settlement boundaries, so as to approximate the portions of a settlement that lie within each subdistrict. I hope to experiment with this idea later this week to see if it is over-ambitious.

Best regards, Mark

To unsubscribe from this group and stop receiving emails from it, send an email to data...@googlegroups.com.

[Sharad Lele (शरच्चंद्र लेले)]

Dear Digvijay:

Thanks for this information. I did not know about the censusgis.org portal. It seems to have boundaries for areas where I have otherwise drawn a blank. BUT it is really hard to use this portal for anything other than to 'see' approximately what is going on. Pity. Do you know who RIDDHI is? It appears to have been authorised by Census of India to put this up.

You may also notice that the IITB maps are different from the Census maps. They appear to come from the MRSAC maps that are present on the MRSAC portal. (Not sure why CSE-IITB folks have not indicated the source). These maps/boundaries are in my opinion much more accurate (both in terms of their shape and their geo-positioning) as compared to Census maps, but that leaves certain questions about missing villages unanswered... The missing polygons in these maps appear to be either forest polygons or town polygons, and am curious why they have gone missing. Is there someone from CSE-IITB team that we can loop in on this?

Sharad

You received this message because you are subscribed to a topic in the Google Groups "datameet" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/datameet/1OXkpdB-2iU/unsubscribe.
To unsubscribe from this group and all its topics, send an email to datameet+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/datameet/CA%2BsGXOKhw1OmFwgRiZC5AAoWezaOTB1b70xLu_TZYgP%3DGdvkYg%40mail.gmail.com.

[Digvijay Bendrikar Shinde]

Connecting Prof Jitendra Shah from GISE Lab of CSE IIT Bombay. He may like to discuss this with you.

About RIDDHI foundation, it's an organisation based in Kolkata. They work in GIS technologies to support social projects.

Regards

Digvijay