Identifying urban agglomeration boundaries

[Aditya Medury]

Hi,

I am trying make sense of how to translate the definitions of urban agglomerations/million-plus cities in the 2011 census. Census defines urban agglomeration as follows:

An urban agglomeration is a continuous urban spread constituting a town and its adjoining outgrowths (OGs), or two or more physically contiguous towns together with or without outgrowths of such towns. An Urban Agglomeration must consist of at least a statutory town and its total population (i.e. all the constituents put together) should not be less than 20,000 as per the 2001 Census. In varying local conditions, there were similar other combinations which have been treated as urban agglomerations satisfying the basic condition of contiguity. Examples: Greater Mumbai UA, Delhi UA, etc.

Source: https://censusindia.gov.in/2011-prov-results/paper2/data_files/India2/1.%20Data%20Highlight.pdf

Since they can be a combination of multiple towns/cities, outgrowths, municipal corporations, identifying their sociodemographic and built environment characteristics from other databases becomes a little challenging. For instance, the breakdown of all UAs with more than 1 lakh population is specified here:

https://censusindia.gov.in/2011-prov-results/paper2/data_files/India2/Table_4_PR_UAs_1Lakh_and_Above_Appendix.pdf

I can think of a few different ways in which I may be able to approximate some boundaries associated with these UAs:

• I could use the ADM3 shapefiles of sub-districts (https://earthworks.stanford.edu/catalog/stanford-rj389fh4679) and picking locations based on visual inspections/overlap of city points. I understand that municipal boundaries may or may not always match the administrative divisions (not very clear on this)
• I could also base it off of pincode layers (https://github.com/justinelliotmeyers/INDIA_PINCODES) which might be more fine-grained if I need additional flexibility in selecting small regions at the periphery. 
• I could go the population grid route and do some clustering to find regions that might approximate the UA populations provided in the 2011 census for some of the bigger cities. But this would get really crude with no unique boundary solutions.

Wondering if people have any thoughts on this.

Thanks!

Aditya

[Deepak Sharda]

Hello sir great thinking.

I suggest to do it with LULC data. We have color coded raster data of area (Red for urbanised or populated area). 

we can extract polygon from LULC  data for Boundaries.

regards

--
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/da0f619a-9d2a-4161-b3e8-d04ecbd334fbn%40googlegroups.com.

[Siddhant N]

Have you seen GHSL? https://ghsl.jrc.ec.europa.eu/data.php

They have a robust methodology for UA identification across the globe inc India. Their India UA data is not perfect but beats most other public datasets.

[Deepak Sharda]

thanks it helped a lot. Although it does not cover complete area. I am attaching Geo Package file here

 GHS_STAT_UCDB2015MT_GLOBE_R2019A.zip

--

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/31c8b0d0-0bbf-4f5b-ba12-5e934e7b7d84n%40googlegroups.com.

[amish sarpotdar]

Hi Aditya and others, 

I have often thought of this since my PhD deals with some of the boundary and territory issues. Some pointers from my own experience:

The GHSL is useful but will be hard to overlay with official boundaries and is a plumbing fix more than a sustainable solution for figuring out UA's as the resolution is quite flimsy at lower levels and for smaller UAs. 

I think the solution then is to have a tiered approach using Census Town Codes  that requires a bit of coding and lots of time and cleansing: 

1. Gather all the village or town vector boundaries which I think are already available

2. Check for OG vectors and missing territories. 

3. Then Add and combine the vector boundaries for each UA corresponding to the list in the Census according to town code numbers. 

There are a lot of limitations with this approach and it is hard to express in one single mail. However, the benefits of following the Census town code approach are massive:

1. Once you have the UA vector file as compiled through census, overlaying socio-economic and other indicators is a child's play as you just need to use Census Town codes

2. Time series is possible if done for 2001 and 2021. Note: Between 2001 and 2021 the boundaries have changed but not as massively as between 1991 and 2001. 

There are other approaches such as having a tiered composite region approach, but those are more for time-series of data. 

It is useful to acknowledge this is a long drawn process, but I would be happy to help and volunteer with any data/code/inputs when needed. 

Thanks. 

Kind Regards, 

Amish

--

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/da0f619a-9d2a-4161-b3e8-d04ecbd334fbn%40googlegroups.com.

[aditya medury]

Thank you, Siddhant and Deepak! I shall check this source out. 

Best,

Aditya 

To view this discussion on the web visit https://groups.google.com/d/msgid/datameet/CA%2B4SsQBqFeu4FsMD_kKS4GCe5rgO-RGX%2Bg%2B6F7ZDwmLbu-tmDg%40mail.gmail.com.