Friday, July 11, 2008

Various geocode systems for Thailands subdivisions

For all my purposes I am using the geocode system from the Ministry of Interior, with the two digit numbers for the provinces, four digits for district, six for the subdistricts. It's simply the easiest and most common system, and unlike all the others it goes down till the village level and apparently even includes the municipalities. And also parts of the codes are covered by the national standard TIS 1099-2548. But there are other geocode systems for the country subdivisions, which are not just limited to Thailand.

The most well-known geocoding system for country subdivisions is of course the ISO standard 3166-2, which in fact is identical with the province codes of TIS 1099, only the 3166-1 country code TH is added to disambiguate it with the codes from other countries. The only specialty is the code of Pattaya in that standard, TH-S, as Pattaya forms a special administrative area, which is a local government structure but outside the municipal (thesaban) system.

The US Federal government uses its own standard FIPS 10-4, which confusingly has the same structure as the ISO codes, but the numbers for each province is totally different. Another problem with it - after 4 new provinces were created in the early 1990s, not only new codes were added, but also the codes for the provinces altered were changed. The FIPS codes were also used in a first draft for the ISO standard (ISO/DIS 3166-2:TH), but probably after intervention from the Thai government the TIS system was finally used. Both ISO and FIPS only cover the first level of country subdivisions.

Gwillim Law both in his book as well as on his website proposes his own geocode system named HASC (Hierarchical Administrative Subdivision Codes), addressing several problems other geocode systems have. For Thailand he has added codes till the district level, though it won't be difficult (just a lot of work due to the sheer number) to add the subdistricts or even the villages as well. Just too bad that the most serious flow he sees in the ISO and FIPS system, the time lag from changes in the entities till they get an official code, also applies to his HASC codes - not because of the bureaucratic work to be done but simply because he seems to be too busy. Wiang Kao still has no HASC code, even though I had notified him of this new district quite some time ago already.

The United Nations develop their own system of geocodes, going down till the second administrative level. Named Second Administrative Level Boundaries (SALB), it also includes maps of the boundaries, though only for few countries so far. Sadly they claim copyright on both the maps and the codes, and access to their website is only possible after registration. And of course any commercial use must be licensed, so I doubt the SALB codes will get much popularity. For Thailand, it has codes for the years 1990-2003, but no maps yet. Due to the five year time lag, it also does not contain Wiang Kao. But what is useful is the fact, that they also list the changes, i.e. the new provinces of the 1990s. But same as FIPS they chose to give different codes to e.g. Udon Thani before and after Nong Bua Lamphu was split off. Also, the list of historic changes of the 2nd level entities is "under process" - they'll have to list the 142 new districts created since 1990 there.

To also list a probably obsolete set of codes, in the national census 1990 the National Statictical Office had used codes which look similar with the TIS 1099, but in fact all the numbers, both for provinces as well as for the subdistricts, are different. There'll be a blog posting dedicated to this coding system in the future.


lenni said...

I experienced the same problem: No worldwide geocode standard available that is
- updated regularly
- consistent
- free
- and covers more than just the first level. But even the second level covers sometimes huge regions. In many countries the third level is best suited for area addressing and my personal favorite is HASC since this code is able to have even a fourth or fifth level.

Andy said...

The only problem I have with HASC is the fact that Gwillim is apparently so busy with updating the data he still hasn't responded about my mails.

One problem I have with all the codes mentioned - they are not complete for Thailand. There are more subdivision that just the central administrative Changwat, Amphoe, Tambon, Muban - there seem to be some codes for municipalities (thesaban) in the DOPA/TIS-System, but no complete list yet. The TAO as one level below municipality have no codes. And no coding goes into the past completely, covering abolished entities like the Monthon (till 1932), or for example the former province of Thonburi (till 1971) has no code in any system either.