Talk:List of metropolitan areas by population/Archive 2

Archive 1

Archive 2

Archive 3

Archive 4

Archive 5

Urban area population of Osaka and Nagoya

Population of Osaka urban area is wrong with 17,514,959, its Nagoya is 8,764,837, and there is not it. Please do not change it without permission.

List of metropolitan areas in Japan by population

Population of FukuokaーKitakyusyu urban area (Japan) is 4,990,000 people (2006).

7th Osaka-Kobe-Kyoto metropolitan areas 17,514,959

25th Chukyu Metropolitan Area 9,480,000

59th FukuokaーKitakyusyu urban area 4,990,000

I think that we should make the other pages agree with each other. The value in List of metropolitan areas in Japan by population should agree with the value in this document for Keihanshin (Osaka).

Please comment

Paullb 02:39, 1 February 2007 (UTC)

Source Dispute

My opinion that we should restore City Population figures. Unfortunately, UN came unreliable source for urban area population. That review does not representing real figures. Front page for this article of wikipedia says "these figures include suburban areas immediately surrounding a major city and sometimes multiple major cities which may be close enough together to function essentially as one area.", what become not true with UN figures. UN counted some cities with metropolitan areas, some with agglomeration and some with just city propers that is absurd. Examples is Moscow (if Polaron is right on definition, however it wasn't 10,7 in 2003 within city limits, but much less) and London. CityPopulation figures is more close to reality. Please do not updating the article, before we come to consensus. I'm awating for replys. Elk Salmon 19:34MSK, 19 October 2005

I have no problem with putting city population.de figures as long as there are definitions (and land areas) given for each metropolitan area. The citypopulation site does not indicate the boundaries of each area while the UN compilation does. I agree that the UN uses different types of definitions for different countries so they are not directly comparable. Polaron 15:57, 19 October 2005 (UTC)

Making a population survey of cities that is 100% consistent would be amazingly hard, and downright impossible for some countries (not all countries take stock of the same things in their Census). The UN Survey is both public domain and explains the sources well. Please also see the debate below. DirectorStratton 15:11, 21 October 2005 (UTC)

I found this article [1] while looking for information on some of these urban areas. It looks like a paper on metropolitan area definitions and why the numbers vary so much across different lists. It's very informative and those interested in making this list more reliable should read it. One of the authors, Richard Forstall, attempted to make a standardised definition of metro areas and has compiled his own list (only the Top 20 are in the paper). I've been looking for more information on his definitions but can't seem to find them online. Polaron 17:59, 21 October 2005 (UTC)

IS THIS JUST A SELF-RANKED FIGURES OUT OF RAW DATA FROM THE 2003 UN STUDY?

THE STUDY WASNT INTENDED TO DEFINED OR ASSIGNED A STANDARDISED METROPOLITAN AREAS IN THE WORLD. THE STUDY IS SIMPLY EXPLAINING THE TRENDS AND PROSPECTIVE DEVELOPMENT OF URBAN AREA IN THE WORLD.

NEVERTHELESS THAT IS WHY MANY GROUPINGS OR THE "AGGLOMERATION" OF THE U.N. ARE QUITE DEBATABLE ON THE DEFENING OF WHAT ARE THE METROPOLITAN AREAS IN THE WORLD OF TODAY. BEC THEY CANNOT SIMPLY STANDARDIZE SUCH THINGS, IT CAN ONLY BE DONE IN A LOCAL LEVEL.

ONE EXAMPLE IS THE SAN FRANCISCO-SAN JOSE AREA WAS SPLIT INTO TWO IN THE STUDY.

SAN JOSE AND SAN FRANCISCO WERE 2 SEPERATE METRO AREAS BACK THEN BUT THROUGH THE YEARS THE TWO URBANISED AREAS HAVE ALREADY MERGED PHYSICALLY AND CREATED A WHOLE NEW BIGGER AGGLOMERATION.

SINCE THE STUDY IS FOR ON HISTORICAL COMPARISON OF METRO AREA IT WOULD BE HARD TO COMPARE 2 SEPERATE METROS BACK THEN (LIKE SF & SJ) AND COMPARE IT TO THIER RECENT COMBINED STATISTICAL DATA OF TODAY. THAT IS WHY THE U.N. KEPT THEM SEPERATE ON THE STUDY.

THIS CASE IS ALSO TRUE TO MANY MORE METRO AREAS AROUND THE WORLD SOME EXAMPLE ALSO INCLUDE D.C., BOSTON, DALLAS, SEOUL, MOSCOW, KUALA LUMPUR, TAIPEI, ETC. ETC. WHICH HAS EXPERIENCED Conurbation TRHOUGH THE YEARS.

THAT IS WHY WE SHOULD USE A MUCH COMPREHENSIVE LIST THAT ACTUALLY FOLLOWS THE STANDARDS OF EACH LOCAL NATION ON WHICH A METRO AREA BELONGS TO.

WE SHOULD USE THE "Citypopulation.de: The Principal Agglomerations of the World" WHICH LOOK MORE PROFFESIONALLY MADE AND MOST OF ALL UPDATED FOR 2005. ITS VERY RELIABLE WEBSITE USING DIFFERENT SOURCES FROM EACH INDIVIDUAL LOCAL NATION. Edits by 71.107.251.137 around 07:50, October 15, 2005

I don't respond to anons writing in all caps. DirectorStratton 13:36, 15 October 2005 (UTC)

The "Citypopulation.de:" source is a Public Domain so it can be use as long you cite the source. Again I want you to know that the 2003 UN figures has a lot of problems in what we are representing here. first it is already out dated, secondly I already adressed this that many of the metropolitan areas are fragmented bec. it was only use for hitorical comparison. Third you should not rank them all by yourself beccause it can cause multiple errors.

2005 "Citypopulation.de:" source, Im not saying it is perfectly accurate but its certainly reflects the closest of defining each metropolitan areas around the world today. It was proffesionally made and they used reliable sources from each individual country to create the rankings.

P.S. your reply to my first comment seems like childish. Just because Im "annonymous", that Im not important enough to be responded to. You seem like belittling people here I hope the editors would notice this. Edits by 71.107.251.137

Maybe you're not aware, but writing in all caps is typically used only by vandals, trolls, "drive-by insulters" and other miscreants. Being an anon further increases that likelihood. Why not register an account? As for citypopulation.de, it is not public domain. Every page says "Copyright Thomas Brinkoff"; although the Agglomerations page does allow distribution with citation, we can only use works if they are placed in the public domain or are licensed under the GFDL (eg the article must say that it is placed in the public domain). Also, I haven't found any information as to the methodology used in the rankings anywhere on the website, which makes it hard to determine neutrality. Although there are problems with the UN report as you mention, it satisfies some important conditions:

• Is (or attempts to be) objective and neutral in classification
• Comes from a well known and regarded agency
• Is public domain

Yes, generating a self-ranking can produce errors, but that's why anyone can edit Wikipedia. If someone feels that the page has an error, they can correct it. That isn't an argument for not incorporating a set of data. DirectorStratton 20:53, 15 October 2005 (UTC)

---

First of all the UN study was a 2003 Data and not 2005 as your own article claims. it was published in 2004 or 2005 but the figures are all from 2003 figures. beyond this are just propectives.

But that was not really the point here

The thing is the problem is not on the UN study itself but its on your own self imposed methodology.

I already brought up this--- many Metropolitan Areas was misrepresented in the study. they didnt actually intetionally misrepresent them bec. The study wasnt INTENDED TO DEFINED OR ASSIGNED A STANDARDISED METROPOLITAN AREAS IN THE WORLD. THE STUDY IS SIMPLY EXPLAINING THE TRENDS AND PROSPECTIVE DEVELOPMENT/GROWTH OF URBAN AREA IN THE WORLD.

Let me give you an example again Seoul should have been a lot higher on the rank so as San Francisco so as Tapei and Washinston DC. because these metro has already merged with another metro. I already explained this so I wont do it again.

You said you "haven't found any information as to the methodology used in the rankings anywhere on the website, which makes it hard to determine neutrality."

On the overview section of the website its sates clearly

"If you require explanations concerning the data, the abbreviations or the usage of the interactive maps, please click on Info. There exists also a page showing answers to frequently asked questions. Under Refs, you will find links to other web sites and the references."

consider this:

The website Uses "MULTIPLE RELIABLE SOURCES" from Each Individual nation and used a proffesional methodology by proffesional people. So it is very objective, neutral and stadardise.

vs.

You only using outdated raw data from a single source that you self ranked.

Let me give you an anaolgy. If we are buying a Television what would you rather buy a "complete and assembled one" or "raw materials of TV (bits and parts)" that you have to assemble from scratch?

You are doing the start from the scartch thing, you are really not sure what you are doing making it prone to error. Because of this expect a massive amount of editing, complaints and it a lot of dispute. Or would we rather have use a "complete product" to avoid many objection.

Since the Agglomerations page does allow distribution with citation why not just use it?

I uderstand you put some work on this but be reasonable. You even admitted that you have not found the Morroco or North Korea figures. so where did you find them eventually? from other sources? If it is I guess the "netrality" and "objectivity" of your work is really questinable

I STRONGLY suggest that you do not continue what you have been doing and just use the more updated complete figures, because I know many people will just challenge you. Edits by 71.107.251.137

Many arguments you have made (UN survey as estimate, not a definition) are apparent filibusters, as the arguments apply equal to cp.de or any other survey.

I find the cp methodolgy rather vague with only one paragraph of explanation:

Official censuses and estimations are mostly the basis of the population figures presented on this web site; the definition of agglomerations is sometimes based on unofficial or own estimates. Please note that the data of such statistics are all of varying, and some of suspect accuracy. There are several reasons: the varying relevance and accuracy of the base data, the poor comparability of the definitions of agglomerations, errors in the projections and so on.

The UN report is very specific, starting on page 123 of the report, with over 40 pages of references and information on the difficulties and flaws with specific data in the report.

Again, cp.de is not public domain so it can't be used. Do you run this site? If so, I can help you make it compatible with US public domain or the GFDL.

Also, I could not find the flag templates for Morocco and North Korea; the statistics are per the UN report. DirectorStratton 23:40, 16 October 2005 (UTC)

The references for the cited UN report indicate that different types of areas are being reported for the different countries. In the Top 10, for example, Tokyo, Mexico City and Sao Paolo (and possibly Buenos Aires) are official metropolitan areas defined by the countries themselves. Jakarta, New York City, and the three big cities of India are all defined as the urbanised area (typically smaller than the metro area). Shanghai, on the other hand is defined as the municipality (province) of Shanghai which includes a large area of primarily rural counties under the administration of the Shanghai local government. The UN relies on official statistics that each country compiles, which are not always comparable across different countries. What if we put the areas that these populations refer to so that people get an idea of how big a given area is? This makes this list a little more useful for comparisons across different countries. (68.85.39.28)

• Great idea. I don't have time to do it. Be bold. DirectorStratton 00:17, 19 October 2005 (UTC)

On a more thorough reading, the UN report cited is basically tracking population growth in urban areas so if this list is used, the article title should probably be changed to "list of urban agglomerations". Metropolitan areas typically include suburban and sometimes rural areas that are economically tied to the core city. (68.85.39.28)

• • • the fact that the UN holds to the old rule for its own historical reasons shows that this data is not progressive enough to change with the times and a changing world. This list is OBSOLETE. 75.3.234.103 13:33, 2 May 2006 (UTC)rick

Any population figure must be accompanied by 3 things: a date, a surface area, and a definition of the boundaries. Also, figures must be verifiable from primary sources. So, how would you choose the criteria for a locality to be included or excluded in a specific metropolitan area? Now what the UN WUP report does is exclude rural area from its figures. If it has to choose between two available figures, it uses the smaller figure. If you were to include rural areas, which would you include and which would you exclude and why? Would your criteria be applicable worldwide? Would you be able to defend your system against people wanting to make their cities figures bigger? Polaron 13:49, 2 May 2006 (UTC)

What happens next?

Are people willing to go back to the citypopulation.de figures or not? Are we going to change the title to urban areas? At the moment the article is riddled with anomalies so we have to agree on some course of action! Jameswilson 02:30, 28 February 2006 (UTC)

There will always be anomalies as no ranking of metropolitan areas can be completely neutral. In any case, perhaps there should be some kind of voting as to which source to use. We should definitely stick to one source for the entire list though to prevent random people from changing figures of individual cities. We should also list the extent of each area as is done now should the source be changed. In addition to citypopulation.de, world-gazetteer.com is another possible source. Polaron 03:09, 28 February 2006 (UTC)

I agrre about the impossibility of getting perfection here. I would prefer citypopulation as the figures are taken from actual national census figures. world-gazetteer used to be very good but its not updated so much and, more importantly, extrapolates the current population from the changes during the previous intercensal period, which is a defect in my view. Using citypopulation would mean omitting the area which I dont mind at all (seems to me population density is a separate topic) but that would be a problem for you? In an ideal world the UN would be the ideal source for this sort of thing, but they admit their definitions vary between cities which rather defeats the point of this list. If we did take the UN figures I think we would be unable (reasonably) to reject edits based on national censuses as they came along. Jameswilson 03:38, 28 February 2006 (UTC)

I have no problem with citypopulation as long as it can be made clear what definition the figure refers to. We don't have to put land areas but I think we do need to indicate the definition of the area associated to each figure. There must be people who are familiar with various cities enough to know the area definitions so we can put them in. In any case, my opinion is we should still put this up for voting but I'm not quite sure how to do this. And it might be good to hear from a few other people as well. Polaron 13:42, 28 February 2006 (UTC)

I dont know how to do a vote either! maybe recruit a few people from the Urban area page which is more active. Jameswilson 02:41, 1 March 2006 (UTC)

---

The heading for this page should be Urban Areas, not Metropolitan Areas. These figures are for the city proper and the immediate built-up suburbs. The exception is Seoul - the figure is for the greater metropolitan area. If we're going to use figures like the one used for Seoul, then please be consistent and use sources that reflect greater metropolitan areas, e.g. CityPopulation.de, PopulationData.net, etc.

For example, estimates of Mexico City's metro area range from 22.8 million (CityPopulation.de) to 27.6 million (Harm de Blij in his Geography: Realms, Regions, & Concepts textbook), while New York's metro area estimates range from 21.9 million (CityPopulation.de & U.S. Census) to 24 million (PopulationData.net). Similiar examples of much higher metro populations can also be given for just about every other city on this list. [[User:BillZav] 11;45, 22 March 2006 (EST)

What about moving this list to "List of urban agglomerations by population" and then creating a new list of metropolitan areas using citypopulation.de figures, which seems to be the alternative that most people who post here prefer? If anybody wants to go ahead and put in the citypopulation figures, please do so. If you know the definitions associated with each figure, put that in as well. Polaron 14:37, 23 March 2006 (UTC)

• • • since when has the UN been good for anything (except maybe UNESCO). Citypopulation gets my vote. 75.3.234.103 13:14, 2 May 2006 (UTC)rick

So which source should this list use?

I have posted this debate on WP:RFC/HIST so that other people can comment. Polaron 20:16, 23 March 2006 (UTC)

• Option 1: UN list
• Option 2: Citypopulation.de list
• Option 3: World Gazetteer list
• Option 4: Use multiple sources for the list or use some other source.

I like the idea of Options 1 & 2, if just for consistency. Not everyone will be satisfied no matter which list you use. Another alternative might be to list the largest metropolitan areas by range, e.g. Tokyo 30-40 million, Mexico City 18-26 million, etc. [BillZav 12:06, 27 March 2006 (EST)

• • • with just a glance, option 2 seems to workd the best. 75.3.234.103 13:10, 2 May 2006 (UTC)rick

 I prefer option 1 at least for the sake of consistency. Option 2 relies on Consolidated Statistical Area figures for the U.S.  CSAs are not to be 
 understood as metro areas as in many cases they combine metro areas.  Part of the problem rests in the fact that prior to having CSAs, the OMB (Office      
 of Management and Budget) used the term Combined Metropolitan Statistical Areas (CMSA) rather than CSA. In any case option 2, like option 3, makes 
 American metro areas ridiculously large in the spatial sense while adding little in terms of population. chazman 02:02, 13 February 2007 (UTC)

London

Greater London is the term used to describe the city-limits of London.

The population of London within the city limits is around 7,500,000. This figure is the one used for the table, and is grossly incorrect.

On the London page, the metro population is specified at between 12,000,000 and 14,000,000.

Many websites support this figure:

demographia.com - 13,945,000.

emporis - london page - 11,850,000

http://www.answers.com/topic/london-commuter-belt - "As of the 2001 census the London Metropolitan Area had a population of 13,945,000"

Suggested edit - As this article details metropolitan areas, surely it would be more accurate to post the 14m figure, and the 8m figure in the article concerning cities, as this number only relates to those living inside the boroughs of Greater London, i.e. within the city boundaries. Polaron is right in saying that the figure for metropolitan areas is grossly incorrect - James

Greater London Authority Figures

The Greater London Authority, the authority headed by the Mayor of London, said that the metropolitan population of London was 18,400,000.

demographia.com - showing GLA's figures

experts.com - showing GLA's figures

The current list is based on the UN World Urbanization Prospects report. This report uses census figures (and not figures from regional planning agencies) as the basis for its projections. The report uses the census figure which is closest to the urban agglomeration definition. The report tries to avoid adding up multiple figures if possible. For some reason the UN has not included the 2001 UK census in its estimates resulting in a greatly undercounted (by ~10%) London urban area figure. I will add a footnote indicating more commonly tabulated metropolitan area figures. Polaron 04:06, 15 February 2006 (UTC)

I don't think we should be basing our statistics off the UN. I totally disagree with this list. First off, it's common knowledge that London is the largest city in the European Union. So how did Paris get a higher ranking? This Wikipedia website actually sais it. And Miami, and Philadelphia were put on this list even though they aren't the 4th and 5th largest metropolitan areas in America. Does that make since? Those spots should go to Atlanta, GA, and Houston, TX. And finally, howcome Sydney, NSW is so low on that list. I know it's about the same size as Chicago.

The UN list uses data from census offices exclusively -- although they have their own projection methods. In the case of London it uses the urban area as defined by the national statistics office, which happens to be smaller than the census defined urban area of Paris. In any case, there is a note for London indicating other commonly used figures. For US cities, it uses urbanized area data which may have slightly different orderings than metropolitan statistical areas or combined statistical areas. (BTW, Atlanta and Houston are #59 and #60). Can you provide official census data indicating that Sydney has a population of ~9 million? If so, we can put that in a footnote. Polaron 23:54, 6 April 2006 (UTC)

London urban area is less populated than Paris urban area. Paris urban area is around 10 million people and London urban area is around 8 million people. As for metropolitan areas, Paris is at 11.6 million people and there's no official figure for London, which should be around 12 million people. Finally, the Greater London authority has defined a metropolitan region, larger than the Netherlands, in which London is only part of. It has nothing to do with the metropolitan statistical areas as conventionnally described. Metropolitan 14:17 10 April 2006 (UTC)

The Londoners (?) seem to be dismayed their city is ranked below Paris... I would not agree that it´s common knowledge that London is the largest city in the European union (perhaps by people in the UK). If we regard the Greater London region as the true metropolitan area, it would in turn give Amsterdam a population of 16.3 million people. An urban area ends where the rural area begins, i.e. before you reach London´s airports.--JFK 10:48, 13 April 2006 (UTC)

- Yes - but this is a list of METROPOLITAN AREAS defined as the urban area plus the surrounding area which serves as a commuter belt. Whilst the Paris Urban Area has a higher population than London's, beyond the urban area London has a large number of satellite towns with populations of 100,000 plus such as Southend, Reading, Oxford, Milton Keynes, Crawley, Brighton etc which together with London mean that the Metropolitan Area of London has a higher population than that of Paris, which is surrounded by relatively sparsely populated agricultural areas with few large towns. The reason for this is largely the Green Belt legislation that created artificial "green" areas separating London from its satellite towns.

Exile 19:18, 16 October 2006 (UTC)

Oxford? Brighton? My arse! --86.137.247.21 18:41, 18 October 2006 (UTC)

Paris metropolitan area

The figure given in that list is the one of Paris urban area (INSEE unité urbaine), Paris metropolitan area (INSEE aire urbaine) as defined officially represents 11,174,743 inhabitants on 14,518 km² as of the census of 1999. For more informations, please check Paris wikipedia page. If you are looking for an official link, you can find here a list of French aire urbaine on the INSEE website, the official body organizing census in France. Metropolitan 05:36, 13 March 2006 (CET)

The urban area is the definition that the UN uses for its list. Citypopulation also uses the urban area while World Gazetteer uses the metropolitan area. I'll just add a footnote for Paris until the source issue is resolved. Polaron 02:44, 14 March 2006 (UTC)

Cairo

There are a number of contradictions between this page and the respective pages of the cities on wikipedia. For example, Cairo lists itself as having 15.2 million inhabitants. This page lists it closer to 11.

See here Polaron 14:04, 26 January 2006 (UTC)

Internal Consistency

Shouldn't this list be remotely close in numbers to the list of Most Populous Cities by PopulationList_of_cities_by_population?

On the list of city populations Shanghai has a population of 15 million, yet on the list of metro areas Shanghai has 12 million. So the Shanghai area (including the city of Shanghai) has less people than the city of Shanghai? Someone needs to fix this.

Chinese "cities" are weird aren't they? :-) It all depends on how you define what a city is. The municipality of Shanghai has a population of 18 million but includes substantial rural areas. This means that the "city" is actually larger than both its urban area and its functional metropolitan area. The discrepancy is because those two lists use different sources. The 15 million figure that World Gazetteer uses includes all districts classified as urban (16 out of 19). The 12 million figure the UN uses includes only the 9 small, core districts + 4 adjacent districts. Polaron 01:36, 24 March 2006 (UTC)

Well then it seems there needs to be some kind of consistent standard used on wikipedia. Or, perhaps it would be best to list several of the sources side-by-side. Regardless, this mix and match crap doesn't hack it for making wikipedia appear as a respectable source of information.

If you know of a list that is consistent across all countries, please go ahead and change the list then. Both World Gazetteer and Citypopulation do this "mix and matching" as well. Polaron 23:36, 17 April 2006 (UTC)

NO, NOT AT ALL, particularly. Cities are easily and well defined within given borders. metro areas are not. Some cities CAN ALSO be their metro area, some can't. Prime example: Los Angeles, city 3.7 million; metro (depending on the list) 12 to 18 million.

NO, they needn't and shouldn't be the same. 75.3.234.103 12:56, 2 May 2006 (UTC) rick

San Francisco

Q:"San Francisco is more often considered to be part of a combined metropolitan area with the city of San Jose. The total population of the combined area is 7,154,350 (2003 estimate)."

If San Jose is more often included, why should it be excluded here? Thus I think it should be included and San Francisco data should be fixed. Oumnique 13:57, 29 March 2006 (UTC)

The UN just follows urbanized area definitions of the US Census Bureau for US cities. It is because the Census Bureau itself separates the two urbanized areas why they are separated in the list. Polaron 14:38, 29 March 2006 (UTC)

San Francisco is a small city compared to a few other cities in the vicinity. When local people refer to the metropolitan area, we call it the "Bay Area" where "Bay" is short for "San Francisco Bay" which is a body of water, not a city. The Bay Area includes San Jose. I am not sure if the "Bay Area" metropolitan area and "San Francisco" metropolitan area are referring to the same thing, unless "San Francisco" here is short for "San Francisco Bay" also. But I doubt it. Kowloonese 19:35, 29 March 2006 (UTC)

In the article, each area is defined as urbanized area vs. metropolitan. I am not sure how you draw the line between the two. Do you go by whether there is an organisized governing body that ties all the cities together? Or do you go by whether the cities are physically joining together. If you go by the former, check out http://www.abag.ca.gov/abag/local_gov/city/city.html which is a body of government that combined multiple counties and cities into a metropolitan area known as the "Bay Area". If you go by the latter, then simply check out satellite pictures at http://maps.google.com and scroll from San Francisco to San Jose. All the cities between San Francisco and San Jose area are physically connected, you cannot find any unurbanized gaps between them. If you drive along the El Camino Real between San Jose and San Francisco, you can't tell you are passing through multiple cities if you don't pay attention to the road signs. In either case, San Jose and San Francisco should be grouped as one metropolitan area. Besides, all local people, including all the media, refer to the "Bay Area" as one metropolitan area instead of a collection of multiple US Census Bureau defined urbanized areas. The Bay Area is basically a ring of tightly fused urbanized areas with a bay sitting in the middle.

If the title of this article were "List of urbanized areas by population", then going by US Census Bureau's definition of urbanized area is fine. But when the title says "Metropolitan Area", then you probably need to give a more consistent, global definition of what metropolitan means. The US Census Bureau's definition of urbanized area is irrelevant when comparing metropolitan areas around the globe. Kowloonese 20:11, 29 March 2006 (UTC)

Please have a look at this list and see if that is more to your liking. I am interested in hearing from other people whether they prefer that list or the current UN list. If there is clear preference for the other list, we can probably put the current list at List of urban agglomerations by population and put the other list at the current article. Polaron 20:29, 29 March 2006 (UTC)

The list you propose seems to mix up different statistical concepts. In some cases are used combined statistical areas, in others metropolitan areas, and in others urban areas. It means this list will most likely be severly disputed, especially that some figures are based on no official data at all. Comparing internation city population datas is already a tricky issue, but if they aren't even based on similar statistical tool, it becomes totally irrelevant. Metropolitan 02:04, 2 April 2006 (UTC)

The thing is, all lists of "metropolitan areas" do this "mixing" of definitions. Do you know of a list that has a uniform definition throughout? Even officially defined metropolitan areas are not consistent worldwide. One country's metropolitan area may be another's urban area. One advantage of the UN list is they only use figures from national statistics offices of each country. The disadvantage is that if statistics office does not tabulate a figure for some "metropolitan area", it uses what is available, which is usually an administrative area of some kind. Also, the UN tends to use figures that exclude rural areas resulting in population counts lower than what many people are used to seeing for metropolitan areas. Polaron 02:13, 2 April 2006 (UTC)

I wasn't refering to the UN list but actually to this list. Metropolitan 03:25, 2 April 2006 (UTC)

City lines, heck all SUPERFICIAL lines have become TOTALLY obsolete. Including (in America) some county lines, even state lines. If you live in the Bay area, you would know that the WHOLE region is interrelated. Forget lines, think inter-relations. Look where transportation links are inter connected, TV-radio stations and newspapers cover a REGION. The Bay area is a region, Greater LA (including ALL 5 counties) is a region.

IN ADDITION: the written report had Hong Kong and Shenzhen together and (in essence) they are different countries. Hong Kong doesn’t officially revert to China for another 44 years. But living in Shenzhen I'd take the subway to the border, the train to the (Hong Kong) island; ALL within 40-50 minutes.

AND, while we are talking China, how can there possibly be NO mention of the Chongqing Municipal area. China reports its population at 30 million.

P.S.: I know this it a tough situation. You guys are doing a good job trying to decipher. I just say, avoid lines, they are obsolete. Look at regions and inter relations. With the growth of the world, what you guys are doing is nothing short of re-defining the new POPULATION paradigm. Keep up the good work

Just look at this, all this urbanized land facing the bay, and you tell me this isn't one contiguous area!! Sparsefarce 23:38, 20 June 2006 (UTC)

(DNF, 11/12/06)I agree, the Bay Area is ONE continuous metropolitan area. It is pretty much a fact that all the officials in northern California still categorizing the Bay as broken up, have personal reasons to downplay urban sprawl in the area. If San Jose-San Francisco and Oakland are not one, then what is Newark and Northern New Jersey doing in the same metro as New York?!

Polaron, I prefer the other list, as mentioned on March 29th, 2006. However, numbers indicate that The Randstad in The Netherlands has at least 7.5 million people, and if the conurbations bordering The Randstad are also counted, that number will rise up to at least 10.5 million people. If those conurbations are not selected to be part of The Randstad, then the names and figures of the biggest conurbations in The Netherlands are the following: Randstad-7.5 million people, Brabantse Stedenrij (or Brabant-stad)-2.3 million, Betuwestad-around one million people.

Milano

added new population of Urban Area of Milan in the year 2006 by Italian Governement, the Demographia Group and World Gazzetter.

http://www.world-gazetteer.com/wg.php?x=1142792485&men=gcis&lng=en&dat=32&geo=-108&srt=pnan&col=aohdq&va=&pt=a

http://www.demographia.com/db-worldua.pdf

Melbourne

The wikipedia article on Melbourne says it has "a population of approximately 3.8 million (2006 estimate)"

The population given on this page is 3,663,000

I dont know how to edit the page so can someone else do it for me. Thankyou

PS. It was reported in a major Australian newspaper that the population infact burst the 4,000,000 seam a few months ago, but I would need to find a source and cant be bothered.

Also, I am not referring to the Melbourne in America, this is the Melbourne in Australia

Accroding to ABS, the mid-2006 projected population for Melbourne Statistical Division is 3,689,700 (using the model that gives the highest figure).Polaron 04:21, 10 April 2006 (UTC)

Uniformed list

Polaron, I have read many messages above, and I think everybody agree that the current list mixes different things (metropolitan areas, urban areas, municipalities). There is actually a much better list than the UN list you published here. It's a list updated every 10 years by the Geopolis research group at the University of Avignon, in France. These guys make a list of the 1000+ most populated urban areas in the world (read URBAN AREAS, the Geopolis list is not a list of metropolitan areas). They use the international definition for urban areas (not any two buildings more than 200 meters apart) and apply it to every country, disregarding national definitions which vary from country to country, and applying the same uniform definition to all countries. They use satellite images to determine the limits of urban areas, then they retrieve the number of people within these urban areas using national statistics at the neighbourhood level. It's an immense work they renew every 10 years, and I think their list would be much better than the UN list. You can find out more (in English) here. Only extracts of the list are available online, but I happen to have a paper version of the list, containing the 70 most populated urban areas in the world. Ask me if you're interested.

Here are the 10 most populated urban areas in 2000 according to Geopolis:

1. Tokyo: 30.9 million
2. New York: 26.5 million
3. Seoul: 21.1 million
4. Mexico: 19.3 million
5. Jakarta: 17.4 million
6. São Paulo: 17.1 million
7. Manilla: 16.5 million
8. Bombay: 16.1 million
9. Delhi: 15.1 million
10. Osaka: 15.0 million

PS: If you're surprised by NY figures, I think it's because looking at satellite pictures and using the 200 meters definition, Geopolis realised that New York and Philadelphia have now physically joined into a single urban area. It is also probable that in the future Baltimore and Washington DC will also merge with the NY-Philadelphia urban area. In Europe, Geopolis expects that in the coming 30 years there will appear a merged urban area from Brussels to Amsterdam, with some 12+ million people in it. In Japan I doubt the urban areas will merge, due to high mountains separating Osaka from Nagoya, and Nagoya from Tokyo. Hardouin 14:51, 11 April 2006 (UTC)

This is a very interesting list. I agree that it uses a uniform standard (which is great) but it is for morphological urban areas. It might be better suited for a separate List of urban areas by population article. Also, is this document easily available (even if not online)? One advantage of the UN list is that all data and methods are public domain. Also, figures can easily be verified by anyone since the data and sources are available online. Furthermore, correlation with census figures is easily done since it uses census defined areas for the most part. Adding up of separate figures is not done except in a few cases.

Also for the New York-Philadelphia urban area, it must also include the Hartford (Connecticut) and Springfield (Massachusetts) metro areas. Do you know if they also include Long Island (it looks like they do)? There is a big water gap of larger than 200 m between it and either the mainland or Manhattan. If so, do they make exceptions to the 200 m criteria? I've seen another morphological urban area definition (see [2]) that excludes Long Island. Polaron 16:04, 11 April 2006 (UTC)

The 200 meters criteria is explained in detail in the link I mentioned above I believe. Reread. In a nutshel: rivers, canals, freeways, and railways are discounted when counting 200 meters between two buildings. Hardouin 18:50, 11 April 2006 (UTC)

• • • This is BY FAR, the MOST asinine list I have EVER seen trying to document populations of a given area. If, indeed, this list is compiled by the, so called, 200 meter rule, than that rule is ridiculous. This list would be laughed at and have NO legitimacy.
    • NO CHINESE CITIES?
    • Yeah, and then, in the future, obviously Boston and Providence and Pittsburgh will hook up with New York and then New York WILL BE 50 MILLION. Do you guys REALLY think you're having a legitimate conversation? 75.3.234.103 12:48, 2 May 2006 (UTC)rick

I dunno, you seem to be doing well enough having a conversation with yourself. - ҉Randwicked҉ 13:15, 2 May 2006 (UTC) that's a zing!

200 meters works great for the OLD millenium cities. The new rapid transporation cities of the modern age get seriously penalized and misrepresented by this modlel

Hardouin, you've mentioned the urban area between Brussels and Amsterdam (April 11the, 2006). That conurbation will not have 12+ million people, it will be higher. Amsterdam is part of the Randstad Conurbation, with at least 7.5 million people. Just south of that conurbation, you find The Brabantse Stedenrij (or Brabantstad) with at least 2.3 million people. Both Randstad and Brabantstad are touching the conurbation of Betuwe-stad (with another million people). Then you have the Antwerp agglomeration, with at least 1.3 million, and the Brussels Agglomeration, with another 2 million people, at least. All those numbers combined makes a total of at least 14 million people. Besides, that conurbation is in the vicinity of The Ruhr-area in Germany. Furthermore, from the Antwerp-Brussels-Ghent conurbation, it is not that far to the Kortrijk-Lille agglomeration in the South, between the border of Belgium and France. And from there the Ile De France region (Paris agglomeration) is not that very far. Infact, perhaps it is not that strange to think that perhaps in the (near) future Greater London, Ile De France, Randstad, Brussels-Antwerp-Ghent, Ruhr-area and some other agglomerations/ conurbations are all linked with another. Maybe you have some information about that topic? It is, in my opinion, a very interesting case.

Possible model for this page?

Could everybody have a look at List of languages by number of native speakers and especially its accompanying talk page. They face the same problem as here, ie even the best uniform source is not 100% reliable. They use ethnologue as a base but if anybody thinks ethnologue is wrong or out-of-date in a particular case it can be overruled subject to consensus. The talk page shows lots of examples of this process. Jameswilson 00:12, 18 April 2006 (UTC)

Atlanta

According to the US Census Data and Metropolitan Atlanta Chamber of Commerce the current population of Metropolitan Atlanta is 4,708,297. Metropolitan Atlanta is a 28 country region in the state of Georgia, USA.

Randstad

I wonder why Randstad, Netherlands (7,5 million) isn't ion the list?

The UN report lists the urban areas of the Netherlands (Amsterdam, Rotterdam, The Hague, etc.) separately. Polaron 23:03, 6 May 2006 (UTC)

Common, international list of cities (with population)

I added this to the City page too, sorry for any cross-posting, but thought this is too important... I haven't found yet a common list of cities, including english name, local name, with/without special characters, country name (local and international), lat/long etc. It's really a shame that in the year 2006 someone interested in displaying a selection of cities has to consult a dozen of different sources to get - unfortunately - as well almost a dozen of different ways of writing the city name or country name, some with lat/long, some without, some with the population numbers, some without, some with local wirting, some without, etc... Believe me, I already spent hours and days trying to set-up a more comprehensive, unified, harmonious set... but not with great success. I think it should be a place like this one here (Wikipedia), where it would be best situated. -- Luftikus143 13:15, 9 May 2006 (UTC)

The nearest - and upon first look pretty substantive - thing is geonames.org. -- Luftikus143 07:22, 10 May 2006 (UTC)

Its useful for smaller places but unfortunately it goes by administartive divisions not urban areas so it doesnt work for the big cities - for example it gives Paris as only 2.138.551. Jameswilson 23:17, 10 May 2006 (UTC)

Saint-Petersburg

AFAIK, the figure of 5+ million applies to the federal region Saint-Petersburg (region #78) which includes not only the city proper, but also some smaller towns in the suburbs of the city such as Kronstadt, Pushkin and Pavlovo.--Achp ru 19:28, 27 May 2006 (UTC)

City proper here refers to the entire administrative area (gorod). You are correct -- it does include all the communities under the administration of the Town Council. Polaron | Talk 21:25, 27 May 2006 (UTC)

Seattle, Washington

I think that seattle should be put on this list. In the Wikipedia article on Seattle, Washington it says that the greater Seattle area is home to 3.8 million people. I have read outside sources that estimate it at 3,802,230.

Seattle (urbanized area) has a 2005 population of 2,959,000 putting it at #106. Polaron | Talk 12:38, 31 May 2006 (UTC)

UN World Urbanization Prospects 2005 revision

Data for the 2005 revision is now available online here. The full report including what definitions are used do not yet seem to be publicly available. There are some significant adjustments in a few definitions, particularly the Chinese cities. Also, at first glance it appears that figures now more closely resemble urban area populations rather than metropolitan area populations. It might be more appropriate to move the UN list to a List of urban agglomerations by population article and use a different source (maybe this one?) for a list of metropolitan areas. I made a table using the 2005 revision data here for reference. --Polaron | Talk 00:34, 18 June 2006 (UTC)

Polaron, I dont mind if you want to separate out "metropolitan" and "urban area" but I just have no confidence in that world gazeteer page for Britain. Use it as a starting point by all means if it seems to works OK for the places you know well enough to judge but I'll definitely reserve the right to "improve" the figures for places in Britain!

Anyway I think we should press on (this page has been stuck for too long already). Nobody has found an ideal list. Use whichever list you prefer and then we can change particular figures ad hoc afterwards.

I know you are reluctant to sacrifice internal consistency by mixing-and-matching from various sources but in the absence of one single convincing source I think its a sacrifice we'll just have to make - ie, in wikipedia-speak, it will have to be a dynamic list, subject to constant revision as people come along with better figures for a particular city. Jameswilson 02:37, 18 June 2006 (UTC)

I'll put in the UN figures from the 2005 revision as a starting point. I guess, as long as modified figures are based on figures by national census authorities then we can probably allow changes. Polaron | Talk 03:05, 18 June 2006 (UTC)

Oh yes of course, any changes would have to be justified properly. The point at issue is not so much the figures for each of the component towns in an agglomeration, but rather which towns should be included and which are "too far out".

In theory that is possible for the urban area definition - we should be able to agree on contiguity, but it strikes me its much more tricky for wider metropolitan areas ("travel-to-work areas") because there doesnt seem to be a common defintion. Jameswilson 22:17, 18 June 2006 (UTC)

Los Angeles

Shouldnt the Inland Empire Be added to the Metro statistics?

consistency with european list

This global list seems inconsistent with the list of largest European metropolitan areas. Specifically, the Dutch Randstad and the Italian Milano (both ~7.5 million inhabitants) are missing from the global list, and the number for, for example, the Istanbul area, differ. Baszoetekouw 09:59, 9 July 2006 (UTC)

The UN splits the Randstad into its separate components. Are you Dutch? Is that UN interpretaion correct? Is the Randstad one continuous urban area? If one drives from Amsterdam or The Hague to Rotterdam is it continuously built-up or is there agricultural land? Jameswilson 21:47, 9 July 2006 (UTC)

WE SHOULD HAVE A POLL ABOUT WHICH LIST TO USE!!!

It seems that many people here are having a problem about the current list. Why not make a STRAW POLL or something like it so we can have a concesual agreement on which list is suitable and the one we should use. So we can rest this once and for all.

I would appriciate if someone could help me how to organize the voting because I'm quite new here thank you.

Basically here are the choices.

   * PopulationData.net: all 1M+ unhabitants Metropolitan areas
   * Demographia.com: 50 Largest World Metropolitan Areas
   * Citypopulation.de: The Principal Agglomerations of the World
   * World Gazetteer: Metropolitan Areas
   * Current List (UN)

---

888

Philadelphia/Delaware Valley

I can't understand why that whenever I change the Delaware Valley Population and Rank accordingly to what that page says, which is about 5,951,797 people according to that page, someone comes by and puts it back and says "revert to Un data." Just keep it as is! I am doing it according to US Census data.-Andrewia 17:29, 29 July 2006 (UTC)

The entire list is from a single source. Changing individual figures would make the list even less reliable than it is now. If there is consensus to use a different source then we'll change it. For now, the UN list seems to be a good compromise. You can add a footnote if you like but don't change the main figure/ranking unless it is a transcription error. --Polaron | Talk 17:56, 13 July 2006 (UTC)

To make the list more accurate what should we do?

I see many people have problems with the list. I also have some doubts about this list, don't get offended, probably not because of the people who are editing this page, but of certain sources used to create the page. As stated in the beginning of the page "This is a controversial issue rather difficult to quantify, as unlike national subdivisions, there are no defined borders for such an area..... Commuting patterns and other aspects of urban behaviour which underlie the metro area concept differ greatly around the world. Thus all population figures for specific metros should be treated as interpretations rather than hard facts." it's a difficult job. Therefore rather than arguing with eachother, as logical people we should compare certain sources and come to conclusions before putting it on the site. Editing the page by just relying upon one website would be wrong, because none of these websites are %100 percent correct.

I think comparing the list of metropolitan areas with the list of cities will be helpful. When you compare these two pages you get to see some impossible statistics. Logically metropolitan areas should be more populated than it's city, due to the fact that metropolitan areas include cities and the suburb areas around it and even sometimes a couple of cities together and their suburbs make up the metropolitan area (example; Miami - Fort Lauderdale - West Palm Beach). Unfortunately for some cities this isn't that way. There are cities on the list bigger than their metros?! (Example; Seoul, Lima)

According to the main source used for the cities page (http://www.mongabay.com/igapo/2005_world_city_populations/2005_city_population_01.html) there are more examples to this, for example; Toronto and Istanbul. Some like these have been corrected on this page however, but by numbers like 200,000. Where is the source of these new numbers on the page, not provided with the source. It looks like these have been changed just the get the population of these metros just a notch up than their cities.

The main problem exists in the definition of the metropolitan areas. Some under the metros title are really metropolitan areas, some urban areas and even some are only city districts. The same problem occurs in the cities page. For example if we take a look at Lima we see that it's marked as official metropolitan area in the metros page, as it should be (7,186,000). But when we look at the cities page it's marked as province which is a larger area with 7,866,160 people. Neither cities should be showed bigger, nor the metros smaller than they actually are. I attempted to change Istanbul's stats in the metros page due to the fact that the province contains 11,332,000 people. I didn't even attempt to put it for the cities page (like how Lima was put). But it wasn't accepted. On the other hand in the source (http://www.mongabay.com/igapo/2005_world_city_populations/2005_city_population_01.html) Istanbul's city population is lower, how can that be? These major differences spoil this page, I think. Cities are labeled differntly like province, metoropolitan area, urban area, metropolitan municipality etc. For real and more accurate comparissons I think they should be classified as same as possible. I think first of all we should make definitions for such terminology before we attempt to change anything.

I hope for everyone's contribution to this discussion. Only then can we make a more reliable page. Thank you. --John9834 21:05, 20 July 2006 (UTC)

It is extremely difficult to make a list that is completely consistent without going into original research. Personally, the Mongabay list is not reliable since it does not cite sources nor defines what the boundaries of each entity are. One possible option is to use statistical metropolitan areas, i.e. areas as defined by a national census authority. The problem there is that many countries have no official definition for a metropolitan area. Also, not all countries define such areas in the same way.

My suggestion is this: Let's start from the current UN list and then allow changes to figures provided that an authoritative, reliable reference is cited. This reference must be ideally a national census authority. Other official government sources might also be acceptable. The reference must be clear that the figure is for a metropolitan area and we should be able to either determine which cities are included and excluded, or determine the surface area of the entity.

Another potential problem: should polycentric metropolitan areas be included? The UN list does not include such areas -- they are split into different areas. This is why the UN figures are generally lower than what some other lists use. --Polaron | Talk 21:28, 20 July 2006 (UTC)

I think we should remove the Mongabay link, as the main source. But once removed, what can we put in it's place? --John9834 22:04, 20 July 2006 (UTC)

Also how do we know that the UN page is reliable enough? Where do they get the numbers? For example I currently live in Turkey and such stats in the UN page do not match with the ones the Turkish Goverment states. --John9834 22:08, 20 July 2006 (UTC)

Definition of Lima has been changed but in the name of area section "metropolitan area" still remains. Plus the population estimate doesn't match with the cities page. --John9834 23:35, 20 July 2006 (UTC)

UN figures don't match exactly because they are projected for July 1. For example, Istanbul is listed by the UN as 8,744,000 (July 1, 2000). The Turkish census in 2000 was dated October 22 and gives a figure of 8,803,468. Is that too far off? We can always put official estimates if they are way off. The full 2005 UN report is not out yet but if you can find the 2003 report, it says that figures are based on past census counts and interpolated/extrapolated from that. For Turkey, the UN report lists censuses from 1950 to 2000 as sources.

Personally, it doesn't matter to me if the definition listed here doesn't match the definition used by the linked metro area article. You can remove the link if you want. That is currently true for almost all the entries anyway. As long as it is clear what the listed figure stands for then it shouldn't be a problem.

For the List of cities by population page, that is probably best treated as a separately as this article is hard enough as it is. --Polaron | Talk 23:54, 20 July 2006 (UTC)

Sorry but I can't agree on that because there's a reason. When the two pages are treated separately big errors occur (as I have explained above giving the example of Lima, which is corrected later on) due to the fact that the two pages are based upon different sources. --John9834 10:44, 21 July 2006 (UTC)

By the way the number given by the Turkish Government in 2000 was 9,085,599 and this is only the city population. The total population, meaning the province was 10,018,735, including the suburban areas. http://www.istanbul.net.tr/istanbul_istanbul_nufus.asp In Turkey, province and city is considered almost the same, due to the fact that each province contains one city and the name of the city is the same with the province. Also with the goverment's ignorance to this subject mixed it all up. The same problem exists here like the Lima example. In the two pages; the city list and the metro list, the same number (city) is given. I'm suggesting that we correct this. Like Lima we should use the province population. --John9834 12:06, 21 July 2006 (UTC)

Go ahead and make the changes you want as long as they are sourced and the definition made clear. If you know of a reliable source for current population estimates for cities alone, then by all means do overhaul that list. --Polaron | Talk 14:57, 21 July 2006 (UTC)

You said "current population estimates for cities alone". Why cities only? Isn't this the metropolitan area page? I did post a message for the cities page though. Probably you wanted to post this message there. --John9834 21:08, 21 July 2006 (UTC)

But I don't want to edit this page without your confirmation. Because I may make mistakes, even when selecting sources, anyone can. Therefore I will post my suggestions here on the discussion page, then I will put them on the article page. --John9834 21:13, 21 July 2006 (UTC)

I'm suggesting to use province (as I have explained above) for Istanbul, not city, as there's no difference between the two statistics (city and metro). Istanbul province is accepted as the metropolitan area most of the time. --John9834 22:13, 21 July 2006 (UTC)

Istanbul and Ankara

After editing the cities page now the population of the two cities looks bigger than their metros. I'm suggesting to change them. The source is official goverment statistics; http://www.tuik.gov.tr/PreIstatistikTablo.do?istab_id=229

That link doesnt work for me. Could you check the URL please? Jameswilson 23:41, 21 July 2006 (UTC)

I have no idea why it doesn't work. It works for me. This is the link to the main page. http://www.turkstat.gov.tr/ This is English but hard to navigate. Try the Turkish one. http://www.tuik.gov.tr/Start.do From there bring your mouse over to (left of the page) "Nüfus, Konut ve Demografi" and click "Nüfus İstatistikleri ve Projeksiyonlar". Go to the very bottom of the new page that comes up. Click "İllere göre yıl ortası nüfus projeksiyonları". Hope that helps. --John9834 23:52, 21 July 2006 (UTC)

Thanks. I'll look at that properly tomorrow. At first sight I'm not entirely happy with the whole-province definition. The "Köy" column on that page seems to imply there is still a large rural population in the province of Istanbul. Should we be including those? My Turkish is non-existent though. If you speak Turkish could you have a look at the official website of the Istanbul Metropolitan Municipality and see what population/area that gives? Jameswilson 01:43, 22 July 2006 (UTC)

What it means by "köy" isn't village population. There are luxurious villas and neighborhoods in those regions. --John9834 17:57, 7 August 2006 (UTC)

(c) issues

A while back, someone posted a question about the copyright status of this list. Please know that under the United States Supreme Court case Feist Publications v. Rural Telephone Service, factual lists can never be copyright-protected, so as long as we stick to the list (and not, say, editorial comments attached to the list), we can use whatever source we wish. --M@rēino 14:55, 27 July 2006 (UTC)

germany: Rhine-Ruhr

where is Rhine-Ruhr in this list? --LaWa 00:45, 5 August 2006 (UTC)

It is split by the UN into 3 separate urban areas: North (Essen-Duisburg), Middle (Dusseldorf-Wuppertal), and South (Koln-Bonn). --Polaron | Talk 01:02, 5 August 2006 (UTC)

3 Rhine-Ruhr Flag of Germany Germany 11,716,845 in 2005[2] well, thats how big this region is in the list of the largest european metropolitan areas.... but... whatever. 84.144.124.221 14:43, 5 January 2007 (UTC)

Crazy

This list is crazy. How can you exclude San Jose from the Bay Area. It doesn't matter at all what the UN or the census beurogh says because this is just obvious. Just drive south on el camino real (a street not even a highway) and see for yourself. There would be not one person out of the million that live in San Jose that will say that they are not from the bay area. Thats a better source than this census crap. look at google earth or any sattelite image. This is so obvious. The only way to tell you going into a new city is city limit sign. Anyways this is is a huge bunch of crap and should be deleted. More that half of all the info here is incorrect. When has Bogota ever been bigger than Lima. Never ever. Both the cities are both hundreds of years old and not once has Bogota been bigger than Lima. But then on this list it says that it is. There needs to be some administrator here to just delete the page and block anyone else from making it again.

P.S. To all the people trying to make their cities look larger than they really are; its not always good to have a really big city. Try living in L.A.

Vivaperucarajo 07:53, 6 August 2006 (UTC)

If you have an official source for Lima, please cite it and make the appropriate change. In the case of San Jose and San Francisco, both cities are strong employment centers and is why they are listed as separate metropolitan areas. The Bay Area is essentially a region composed of several metropolitan areas. These metropolitan areas do share commuters and is why there is a larger Combined Statistical Area. But that doesn't eliminate the fact that both San Francisco and San Jose are big enough employment centers to have their owen separate metropolitan areas. --Polaron | Talk 15:37, 6 August 2006 (UTC)

That's a bull crap. I live in San Jose but I work in San Freancisco. San Francisco and San Jose simply are connected in every single way. They may have different MSA but it does not necesarilly mean they are not in the same Metropolitan Area. That is why the census have CSA. Do not interchange the term MSA to a "metropolitan area".–88800--67.101.102.134 03:41, 8 August 2006 (UTC)888

By that definition, though, Newark shouldn't be included in New York City because it has its own employment center, neither should Yokahama in Tokyo, etc. 68.96.161.229 06:00, 7 August 2006 (UTC)

That goes into the little details of how the urban cores are delineated. The San Francisco and San Jose urban areas are separate because, while the urban areas do indeed touch, their total contact boundary is less than 3 miles and they were historically different metropolitan areas (Criterion III-A) making them eligible to be split. The situation is similar for Los Angeles and Riverside. See this for urban area criteria and this for urban area boundary maps. --Polaron | Talk 15:42, 7 August 2006 (UTC)

Riverside/LA is not the same. When you drive on the 91 freeway to Riverside, you're not driving through an urban area. They don't touch. 64.60.81.2 22:12, 15 August 2006 (UTC)

Do not be a fool. We need to REVISE THIS article. Everyone who lives in the Bay Area knows San Francisco and San Jose are in the same Metro Area. It seems you are the only one who is pushing for something different. You said they were different metro area historically. So why not put what it is that is current. San Jose and San Francisco simply has since merged and become one Metro Area.

Every Metro Area should be treated differently because there are simply NO standard period. So we should take it to account evrything locally. That is why the Citypopulation.de are the most appropriate as it gets its source right from each local nation. The Reason why there are no dominating city in the Bay Area is because many of the city limits are very small compare to others. For Example the whole city of San Francisco is just about the size of Manhattan. But If San Francisco were to have the city limit size like the whole New York City it would eat up Oakland to the east and most of San Mateo County to South and would be like the 4th largest city in the US. New York City (not the metro area) is just about the size of the Bay Area core including San Jose. So Boudaries are irrelevant as they differ a lot from each metro to another –88800--67.101.102.134 03:41, 8 August 2006 (UTC)888

Additional note: for Newark, it is considered a separate employment core but since it does not satisfy the criteria for the urban areas to be split, it is counted as a metropolitan division within the MSA instead. --Polaron | Talk 15:44, 7 August 2006 (UTC)

Los Angeles and Inland Empire

Census Bureau aside, I think it's clear that the 5 million-plus residents of the Inland Empire should count toward Los Angeles' total metropolitan population. 68.96.161.229 05:59, 7 August 2006 (UTC)

California is tricky. Despite dividing both the urban regions of the Bay Area and SoCal into several metro areas, I think many residents would agree that San Francisco and San Jose form a unified urban region as does L.A. and Riverside. The problem rest on the reliance the Census Bureau places on commuting patterns (the only thing that really matters in determining American metro areas.) As both San Jose and Riverside have substantial areas of employment they are treated as separate metros from San Francisco and Los Angeles.chazman 01:52, 13 February 2007 (UTC)

Updating the list

As we are in 2006, I suggest to change the list according to 2006 figures. We are one year behind at the moment, therefore far from real numbers. The UN page may be according to 2005 but we can take a look at official goverment websites for 2006 estimates for every city. --John9834 18:06, 7 August 2006 (UTC)

Polaron, I need you to comment on this please, because you made this article reliable, so I need your ideas. It would be rude otherwise. John9834 22:38, 17 August 2006 (UTC)

Sorry for not replying sooner -- was preoccupied with highway articles for a while. If there is a reliable source for 2006 population figures, then we should definitely use them. The problem is that we probably won't be able to get official 2006 figures for *every* city on this list. The UN list only has 2010 projections and nothing in between. I'm ok with updating only certain entries as long as it is clear that those entries are using a different year. --Polaron | Talk 23:40, 17 August 2006 (UTC)

May I suggest something, even though I'm not the decission maker here, seeing you guys are into this. We may add a new row to the list where we can put the date of the estimation; so that people can compare the metros with eachother even though their estimations are of different years. As the numbers will change by figures like 100,000 the rankings won't change very much. This is only a temporary solution, as I believe we will find 2006 estimates for all cities in some time. Like John has found (for the Istanbul and Ankara examples), we should look into official government websites for each city (of its country). They are the most reliable sources. Berkserker 21:42, 18 August 2006 (UTC)

To Americans

To Americans. The urban area definition (continuous built-up area) is not commonly used in the US where the main statistic has always been the metropolitan area (roughly equivalent to travel-to-work areas).

However the US Census Department has produced some figures using the urban definition on an experimental basis (see here (scroll down)).

This is useful for comparison purposes if nothing else, because the figures given for European cities such as London, Madrid, etc are generally given on the urban area definition, so its interesting to see how big US cities are when caslculated on this much narrower definition (ie excluding a lot of commuter settlements). See here for the US metro figures. Jameswilson 00:59, 18 August 2006 (UTC)

Redirect from List of urban agglomerations by population?

I don't really understand this redirect. "Agglomeration", or unbroken urban growth, would be the equivalent of the urban area, not the metropolitan area that is an urban area with its commuter belt added. thepromenader 16:13, 19 August 2006 (UTC)

The UN list is actually a list of urban agglomerations. Once we can make a reliable list of metropolitan areas then we can separate the two. Please help and change the figures here if you have authoritative references for metropolitan areas. --Polaron | Talk 16:23, 19 August 2006 (UTC)

Good lord, I would like to help but pure statistics is a bit out of my league. If I remember correctly, someone on one of the pages I am working on did suggest a source like the one you ask for - I will get back to you as soon as I find it. thepromenader 16:37, 19 August 2006 (UTC)

ALL WRONG

THIS IS ALL WRONG. sOURCES ARE OLD AND INNCORECT.

If you know of an authoritative, reliable, recent source, please share it with us. --Polaron | Talk 20:41, 23 August 2006 (UTC)

Milan and Naples

the metropolitan area of Milan is 4,451 sq. km the population 5,941,000

the metropolitan area of Naples is 2,397 sq. km the population 4,254,000

please also Rome needs to be corrected: 3,658 sq. km and a population of 3,653,000

the source is the official gazette of Società Geografica Italiana (Italian Geographic Society) 2001

Can you point to an official government source like istat.it that indicates the extent and population of these areas? --Polaron | Talk 20:32, 28 August 2006 (UTC)

Additional note: It looks like the Eurostat Urban Audit II program [3] uses the province as the "Larger Urban Zone" (LUZ) (a commuter based region that is adjusted to convenient political boundaries). But, except for Rome, these LUZs are smaller than what you listed. --Polaron | Talk 20:38, 28 August 2006 (UTC)

According to this page List of the 20 largest metropolitan areas of Italy the official census never publishes figures for the metroploitan areas. (I wonder why not - some political reason I suppose, like the UK census being forbidden by law from collecting native-language statistics in England).

That list you cited indicates that the province is usually just used as a substitute for the metropolitan area. That is probably what we should also do since official province statistics are easy to get. --Polaron | Talk 23:57, 28 August 2006 (UTC)

It would be convenient but unfortunately I think 3.7 million is too small. Common sense suggests we need a figure in excess of five million as Greater Milan is "generally stated" to be bigger than Madrid or Barcelona. Plus, what would happen when the province is reduced in a few years time as explained here Milan_Province. Monza is certainly part of the Milan built-up area but its going to get its own provincial status. Jameswilson 01:16, 29 August 2006 (UTC)

Anyway the Società Geografica Italiana is obviously an impeccable academic source BUT, for this page, we need figures for the contiguous urban area only, not the whole economic region. Looking at the maps, it aeems that we need a figure for an area which only goes out from central Milan as far as Gaggiano, for example. Does that correspond with the 4,451 km²? If not, do the Società Geografica Italiana have another definition area which stops at Gaggiano (assuming myn maps are correct)? Jameswilson 23:52, 28 August 2006 (UTC)

I've found a map about the metropolitan area of Milan, with many data, this is a photocopy from my old geographic book, Manuale di geografia per il biennio Edizioni R.A.R.E., Milano 1994 [[4]] The metropolitan areas of Bergamo, Como, Varese and Busto Arsizio/Gallarate/Legnano, as described on map, are contiguous with the Milan's one and they form togheter an unique metropolitan area, so they were summed on the table below, in fact in Italy we refer only to an unique one, named as the metropolitan area of Milan (the main city)...that is also the name of the picture on the book, as we can read over the map

The 4,451 sq km area is clearly a multi-core metro region (analogous to a CSA in the US). Maybe we should use the 290 commune, 2,558 sq km definition. That seems like the most reasonable definition to me. --Polaron | Talk 16:57, 29 August 2006 (UTC)

Just because metro areas are adjacent does not mean we should merge them. New York City and Philadelphia metro areas are adjacent but not merged because they are separate urban concentrations. We should stick to the most conservative definition possible to discourage people from putting figures for urban planning regions (often composed of multiple metropolitan areas) as a substitute for the metropolitan area itself. --Polaron | Talk 18:28, 29 August 2006 (UTC)

We are not speaking about great metropolitan areas of big cities, that are also so distant (Philadelphia is distant from New York more than 110 km) we are speaking about very near areas as, for example, New York and Neewark, in fact the second city is on the other side of the river Hudson and also in another state, but usually is comprised on the metropolitan area of New York. The major distances from the lombard towns below and Milan are comprised on a ray of 30/40 km. Besides I would to remind that the articles on Wikipedia have not to be based on opinions, but on sources, in this case the source is not an opinion but a school text, a book used on Italian school.

Let me give you another example then: Boston and Providence. The metro areas and urban areas are adjacent and there is some commuter interchange. Should we merge them? Then there is the problem of London. Using figures that represent multiple adjacent metropolitan areas leads to the 18 million London metro area figure. There's also Mexico City whose metro area is adjacent to two others. This can quickly degenerate into bigger and bigger definitions used in the list which is one reason why the UN list was used in the first place. --Polaron | Talk 19:17, 29 August 2006 (UTC)

Polaron, I agree with the Milan 2,500 km2 definition.

Note: Illustration. The 18 million London definition goes out as far (clockwise from north) Luton/Dunstable in Bedfordshire, all of Hertfordshire, South Essex including Chelmsford, Basildon and Southend, North and West Kent including the Medway Towns and Maidstone, Crawley in Sussex, all of Surrey, Aldershot/Farnborough in Hampshire, and eastern and central Berkshire including Reading, Buckinghamshire up to and including High Wycombe. I.e., all of the Greater London (administrative county) plus the area decribed as the Outer Metropolitan Area or the Inner South-East (of England, I mean). Maximum 40 miles from Central London, sometimes less where there is no railway line/motorway( eg the area of Essex at two-o'clock on my clockface is excluded)..

Ctr: The smaller London urban area definition used here includes most (but not quite all) of Greater London (administrative county) plus a very few bits beyond that boundary such as Watford and Epsom/Ewell, Chertsey/Walton. 15-20 miles from Central London to the end of the urban sprawl and the start of the Green Belt (a line which occasionally creeps inside the Greater London boundary, eg, in the south-east around Biggin Hill). For full definition see Greater London Urban Area. Jameswilson 23:04, 29 August 2006 (UTC)

so...Use the criteria that you prefere, but still I can't see Milan and Naples in the list. Why??? And why were not them considered before?

I think it was simply because the Italian Statistics Office doesnt collect those figurezs (see above) so they got missed out by the UN. Jameswilson 23:27, 31 August 2006 (UTC)

One Important Question.

Where in the current source did the UN claim the list as "Metropolitan areas of the world"? It seems they never reffered to it in that way. I think thats the reason why the list is having so many problems. I just want to clear if the UN did indeed make an official defenition of what is a "metropolitan area", (not agglomeration not uraban population but "Metropolitan area"). Like what the IAU did where they make a clear defenition of what a Planet should be. If the UN did not make such a thing then we should follow each Individual countries defenition of thier own Metropolitan Defenition. It does not matter if they varied from each coutry to country as long they are the official defenition of that country. This measurement is no different when a country is defining what their cities should be.

You are correct, the UN list is properly called a list of urban agglometations. We are trying to move away from strict adherence to the UN figures but we want to make sure each new figure is associated to an official definition. One difficulty seems that many countries do not have official definitions. If you are aware of official sources, please do share them. --Polaron | Talk 03:03, 29 August 2006 (UTC)

I agree this page name is misleading. But I disagree we should take official statistics when they are clearly based on a different definition. The main purpose of this page is for international comparison, an international league table if you like, so we need above all else to have like-for-like figures based on a common definition. Jameswilson 23:16, 29 August 2006 (UTC)

Eeach country has different defenitions of thier cities. Each countries has different defnitions of thier metropolitan areas. Whats the difference? I think we should not aim for comparison because that will lead to subjectiveness, hence, what is happening right now with the list. We simply want to provide information based on official description for each country.

Proposal to move all US cities to metropolitan statistical area definitions

The Office of Management and Budget in the U.S. defines metropolitan statistical areas (MSAs) based on Census Bureau data. These are the closest there is to an official metropolitan area definition in the US. The current UN list uses urban area data for US cities. I am proposing to change all US cities to conform to the MSA (and not the larger CSA concept) definition. If there are no objections over the next few days, I'll begin changing figures. --Polaron | Talk 03:12, 29 August 2006 (UTC)

Many Metropolitan areas fits more on the CSA groupings than MSA. On the otherhand there are also Metropolitan areas that fits more on MSA than thier CSA. We should untilize a list that would incoorporate both categories. (Of course we are just talking about the US). I have seen a list who have done this where they included all the CSAs together with the MSAs that are not included in CSAs and ranked them accordingly.www 06:45, 29 August 2006 (UTC)

The CSAs are composed of multiple urban cores so are not the usual metropolitan area concept. Also, the OMB clearly states in their definitions that MSAs should not be compared to CSAs. We should minimize the tendency to boost metro area figures and use the most conservative values. --Polaron | Talk 14:19, 29 August 2006 (UTC)

I do object to using MSA's here! Create that list by all means, but not here, unless you want to abandon the UN format altogether and change London to 18 million, etc, etc. MSA's are based on administrative boundaries not the actual physical limits of the built-up area. Perhaps we really do need two separate lists. Jameswilson 23:19, 29 August 2006 (UTC)

I'll leave the urban area data then. In any case, most official metro area definitions, where available, are built up using small administrative units (typically municipalities) or approximated by a province/region. The US building blocks (counties) are indeed somewhat large. --Polaron | Talk 23:40, 29 August 2006 (UTC)

Sorry, I was a bit terse there but I really thought it would be a mistake. On a more general note though, I think we may be reaching the point where we need to split this page and create two lists (one on an MSA-style basis, and one on an urban area basis). Although I personally prefer the latter, there is I think a demand for the former (as someone pointed out yesterday(?)). People from America and elsewhere who are used to MSA's come to this page expecting to find that. Reasonably enough, given the title of the page and the fact that thats what they are used to. But it would be a lot of work to put together MSA-equivalent figures for all the European cities. Jameswilson 00:47, 30 August 2006 (UTC)

On a related note, the Urban Audit LUZ data I mentioned in another section seems like a good approximation (on the conservative side) to the metro areas of many EU cities. The most recent tabulated data, however, are only for the year 2001. What do people think about using LUZ data. The main Urban Audit site is here. What do other people think about using this as a reference to standardize EU cities. There is also some information here. --Polaron | Talk 23:46, 29 August 2006 (UTC)

I seem to be in a very negative mood tonight, but the LUZ figures seem a bit strange for the two countries I know best. The Spanish LUZ figures are all for whole provinces - eg Santander 537,000 equals the population of Cantabria. In Britain though the figures for Leicester and Worcester (756,000 and 278,000 respectively) are smaller than the counties of Leicestershire and Worcestershire. Presumably some of the more distant districts of those counties have been excluded. I suspect the different approaches may have had more to do with ease of calculation (Spain lacks a district level of local government between municipalities and provinces/counties) rather than any "demographic" reason. Also, the list of cities is not comprehensive. In Poland just about every provincial capital seems to be included, but in other countries quite large cities are missing altogether - why include Worcester when you dont have Nottingham or Stoke-on-Trent? They would be below our cut-off point anyway but it doesnt fill you with confidence. Unless maybe participation in this EU project was optional for local authorities? Anyway do keep trying - I'm sure you'll find a list I like in the end! Jameswilson 01:24, 30 August 2006 (UTC)

Not all cities were included in the program. But I think all of the EU cities that would wind up in the Top 100 international list are there. I think that if the province is a reasonably good approximation, they do use that (e.g. Madrid, Barcelona, Rome, Naples). For Paris, they use the Ile-de-France region. For Essen, they use the Ruhr Area. For the others (London, Athens, Milan, Hamburg, Berlin) they build up the area similar to how MSAs in the US are delineated. The project is an attempt to create comparable statistics for cities across the EU. I think it is still in the preliminary stage so that's why only selected cities were included. In the absence of official figures, this might be the next best thing. --Polaron | Talk 01:38, 30 August 2006 (UTC)

This sums up our dilemma very nicely. You are used to the building-block approach, taking municipalities (or some lower-level administrative units) or census blocks and adding them up.

But every time I see that approach I automatically get negative and assume those figures cant be what we want here because I am used to the other system. Taking an aerial, nowadays satellite, photo to determine land-use and drawing a red line on the photo around the edge of the built-up area. Add up all the census-blocks (enumeration districts) which fall wholly within the line of course (for the sake of convenience), but where a census block doesnt correspond with the red line, and they hardly ever do, the red line takes priority and its a question of the statisticians wading through the original census returns house-by-house to work out how many people live within the red line.

To you this probably sounds ridiculous but we are used to the idea that the "real shape" of a town is the urban footprint rather than the administrative boundary. (This ia probably because administrative boundaries are changed so frequently in Britain). For example all our road maps show even small towns as grey blobs corresponding to the shape of the built-up area on aerial/satellite photos, rather than showing coloured municipal/parish boundaries with a symbol in the middle for the town centre.

This is why I was hinting at splitting the list to reflect the different approaches. For example, I was playing around with some maps of Brazil and it is clear to me that the Brazilian conurbations on the list are done on your approach (adding whole municipalities even if they includes rural hamlets). So by my criteria all the Brazilian figures on this page are too high. The MSA, LUZ, Brazilian system is perfexctly valid of course; its just different. Jameswilson 23:40, 30 August 2006 (UTC)

Yes, we should definitely create a List of urban areas by population or something like it. But for this particular list (of metropolitan areas), I think the LUZ might be ok to use. Metro areas are generally larger than urban areas for cities with good transportation infrastructure so the figures here being larger than the urban area figures is not really a problem.

Do you know of a good, reliable reference for international urban area figures? Some countries do tabulate urban areas (including the US) but many do not. There must be some international organization that tracks urban areas. If you have a reference, then let's create that urban area list. --Polaron | Talk 23:47, 30 August 2006 (UTC)

I dont know of one single source that brings them together. It should be no problem in Europe has all the countries should have accessible lists on their respective national statistics sites. As you say, the US we have too. I'll have a hunt around but most lists seem to be a mixture of both definitions when you analyse them closely (as with the UN list itself in my Brazil example).

LUZ's seem to be "economic zones" ("economic dominance" rather than commuter, travel-to-work, zones). The Santander example the LUZ covers an area of just over 5,000 km2. Does that seem right for a central city of 200,000? (Santander is a port, summer resort and financial centre which punches above its weight in terms of economic importance within Spain).

You'll know better than me how closely that matches the MSA concept, but I would suspect the the LUZ areas, at any rate in Spain, might still too big even for your (metro) purposes. Specifically, it seems to me that if all Spanish cities participated in the project, most of Spain would end up falling into one or other of the LUZ's. The only "gaps" would be the (ten?) provinces which happen to lack a large capital. From the figures it is clear tha the Santander LUZ directly adjoins the Oviedo one. Is that justifiable on an MSA basis? I dont think the population density (about 80/km2) in Spain is high enough to justify including so much of the national territory. Large parts of rural Spain are quite empty really (to a Briton admittedly). The overall density is less than, say, Ohio. Is 80% of Ohio's territory included within one or other of the MSA's. How much is left out? Of course Spain is only one country. Other countries' LUZ's might work better. And as you say it might be more rliable for the larger cities. Jameswilson 01:31, 31 August 2006 (UTC)

A good comparison to your Santander example might be Charleston SC. Its city proper is about 110,000 with an MSA population of 595,000 (in 2005). The land area of the MSA is about 6,710 km2. So the Santander LUZ is not unusual if viewed as an MSA. --Polaron | Talk 01:49, 31 August 2006 (UTC)

Fine, thats a good equivalent then. Jameswilson 02:27, 31 August 2006 (UTC)

Urban Audit II LUZ figures

For reference:

City	Definition	Population (2001)	Area (km²)
London	Greater London + 42 surrounding districts	11.62 million	8,920
Paris	Ile-de-France	10.95 million	12,080
Madrid	Province	5.37 million	8,023
Essen	Ruhr Area	5.36 million	4,434
Berlin	State + surrounding kreise	4.94 million	17,405
Barcelona	Province	4.80 million	7,755
Milan	Province + surrounding municipalities	3.90 million	2,767
Athens	including surrounding nomos	3.89 million	3,807
Rome	Province	3.70 million	5,352
Hamburg	State + 6 neighboring kreise	3.08 million	7,304
Naples	Province	3.06 million	1,171

So why do we not move for the EU cities to the LUZ definition? It is the only EU-harmonized defintion by a neutral agency (Eurostat) and accepted by all EU national statistics offices? JGG 23:08, 26 September 2006 (UTC)

Johannesburg

To anybody reading this article: please note that the population reflected here for Johannesburg is the CITY PROPER, not the metropolitan area, and is therefore wrong. Suggest you take this into account when using this info for anything. The figure excludes the East and West Rand, which have always been functionally linked to Jozi, but separated for administrative purposes. For a much more accurate figure, please see Thomas Brinkhoff's data [[5]]. The real figure is closer to 7.5 million. PS Soweto is AUTOMATICALLY included in Jozi as it is within the official city boundaries.

YAY! Somebody finally took note! Thank you - the inclusion of the East and West Rand is a much more accurate and real estimation of the Greater Johannesburg Area.

LIST ABOMINABLE

All lists have errors, but this "UN List" is an abomination!! I've been studying populations of cities for years every damn day, suburb by suburb, city by city, and this one is one of the worst i've seen. Citypopulation.de is one of the better ones, although these are my estimates for those cities based on satellite data, multiple sources, area lit by light, commuting patterns, etc.

Changes/Notes to citypopulation.de

• • Denotes Changes to citypop.de estimates.

Tōkyō Tokyo Japan 34,200,000 incl. Kanagawa, Tokyo, Saitama, and Chiba prefectures, excluding mountainous fringes areas of these (Chichibu and village areas), including Toride/Abiko area of Ibaraki prefecture. Many small cities dot the landscape exiting Tokyo N/NE/E, including them or not is very difficult to determine, and megacities like Tokyo have very strong pull. Including Maebashi-Takasaki, Oyama, Kiryu-Ashikaga areas would greatly increase these estimates. Kanto region less mountains and Mito and areas north has around 37.5 mln people.

• • Ciudad de México Mexico City Mexico 24,000,000 incl. Nezahualcóyotl, Ecatepec, Naucalpan including Toluca, as its completely urbanized out to Toluca.

São Paulo Sao Paulo Brazil 20,200,000 Official Regiao Metropolitana of Sao Paulo, plus a few suburbs, minus Salesopolis and low density rural fringe inside RM de SP. SP has a NYC like core with a LA style sprawl. If LA-like sprawl cities are included, Campinas and Baixada Santista must be added, bringing the total to over 25 mln, with potentially others like SJ de Campos and Sorocaba.

8 Shanghai Shanghai China 18,150,000 I support this number for Shanghai as there are so many undocumented people there without official residence.

10 ?Ōsaka Osaka Japan 16,800,000 incl. Kobe, Kyoto, not including Himeji, Wakkanai, or Nara areas. If Himeji areas is added, 17.5 million is easily achieved.

11

• • Jakarta Jakarta Indonesia 18,550,000 incl. Bekasi, Bogor, Depok, Tangerang.

Jakarta is under-reported and easily contains 18 million if not more.

• • Lagos Lagos Nigeria 11,100,000 Lagos definitely has more, maybe 15mln.

• • Krung Thep Bangkok Thailand 8,450,000 Bangkok has easily 10 million people when suburbs Rangsit, Pathum Thani, Park Kred, Nonthaburi, Bang Kapi, Samut Sakhon, Chachoengsao, MinBuri, Samut Prakan, Bang Na are included, as all these suburbs have direct commuter busses to the center of Bangkok.

35 ?Nagoya Nagoya Japan 8,050,000 Though Nagoya region is very populated, Toyota region people don't bother to go to Nagoya, and neither do Mie prefecture people....6 mln is a good estimate, 8 mln is good for connected urban corridor but unintegrated areas.

• • Lahore Lahore Pakistan 7,550,000 Potentially as many as 10 million.

Distribution by country inconsitent ranking basis

The list seems to alternate between ranking by number of 100 areas and by total population of top 100 areas. I'd think one should be decided upon?

Country	Number of top 100 metropolitan areas	Combined population
European Union	10	56,455,000
India	9	80,100,000
...	...	...
Indonesia	2	17,341,000
Germany	4	17,001,000

As the table headers are currently, and without explict definition, the table with the distribution of metropolitan areas by country should contain exactly the same top 100 metropoplitan areas as the main table, just grouped according to country. As a first consistency check, there should be 100 entries in each and the total population should be the same for both. And there shoudl indeed be no entry for European Union. An alternative definition would, of course possible, but needs to be explicitly stated, and linked ot the sources. Until the current implicit definition is changed, entries and rankings can be checked and corrected. (I did this for Germany as it has only 2 areas in the main table and will do so for EU).Tikiwont 08:11, 29 October 2006 (UTC)

Proposed list update

I'm planning to use figures that seem closer to what other lists of metropolitan areas use. The default would be the UN figures. The proposed exceptions are:

• U.S. cities -- Use data for metropolitan statistical areas. The top 5 cities won't change too much but some of the smaller ones would increase.
• EU cities -- Use data for Larger Urban Zones from the Eurostat Urban Audit II program. See stats in one of the sections above.
• Metropolitan planning regions for:
  • Seoul (already implemented)
  • Taipei (already implemented)
  • Johannesburg (currently uses city proper)
  • Lima (currently uses province)
• Urban area data for:
  • Moscow (currently uses city proper)
  • Kuala Lumpur (currently uses city proper)
• Province data for:
  • Istanbul (already implemented)

I think these are the ones that where people come in periodically and change the figures and where people have complained on the Talk page specifically about. I will begin changing figures in the next day or two so if there are additions/subtractions, please let me know. --Polaron | Talk 16:34, 21 October 2006 (UTC)

European Union

I'm surprised no one has noticed this, so I may be hopelessly confused. The "Distribution by Country" list lists the European Union as a country. Last time I checked, Europe isn't a country. Is it possible that this is because some Metropolitan areas in Europe are in more than one country? (ie: Rome, Vatican city)

Proposal for Metropolitan Area definition

Perhaps what this list needs is definition. Maybe we should define a metropolitan area as the area with a minimum population density (ie: the area around Paris with at least 100 people per square mile). This might cause some confusion because some areas would be merged together (BosWash), but at least it would give the list some order. And I know there is at least one credible up-to-date source out there that lists it in this way. So if we accept this proposal, our next question is "what is the minimum population density that defines a metropolitan region?"

That definition is closer to an urban area and is more appropriate for a List of urban areas by population. A metropolitan area typically includes adjacent non-urban territory that is socio-economically tied to the urban core. --Polaron | Talk 03:17, 25 October 2006 (UTC)

Distribution by Country errors.

I'm wondering how Algeria got onto that list when they don't have a city in the top 100.

Also in the Distribution by Country list Colombia only shows as having one city listed when there are two Colombian cities in the top 100 adding up to about 10,000,000+ in population.

Milano official Metro Area 24 November 2006: 7.4 Millions

Hi, there is a big mistake about Milan. the Urban Area of Milan, the real city is of 4.280.000 people. the sprawling, the Metro Area is of 7.400.000 of people, as was for the first time officialy calculated by OECD / OCSE just 3 days ago.

http://www.oecd.org/dataoecd/14/62/37720067.pdf

OECD Territorial Reviews: Milan, Italy

"Milan is often identified as a prosperous region and an international capital of fashion and design. Once a successful industrial city located in the northern part of Italy, Milan has grown into the core of a wider industrial metropolitan region that is home to more than 7 million people. The OECD Territorial Review of Milan recognises that Milan’s historical skills endowment and its advantageous geographic location could underpin its ambition to become a southern European and Mediterranean capital. At the same time, the Review demonstrates that Milan displays disappointing international performances and seems to have lost part of its long-established drive..."

To try and minimize the tendency to overestimate metro area sizes, the list uses the Eurostat Urban Audit II data for EU cities (which is a single urban core definition). Note that the official labor market region of Milan has a 2001 population of 2.98 million according to istat.it. See here. --Polaron | Talk 23:09, 27 November 2006 (UTC)

Holes In The Table?

Perhaps this is a silly question, but is there any particular reason why the table (at least when I looked at it; November 28th in the evening in the Western US) has so many holes for the population density? I was going to go through and correct this, but decided to post this in the talk page first. Ixnayonthetimmay 02:32, 29 November 2006 (UTC)