links for 2010-08-04

Wednesday, August 04, 2010 (19:02 UTC)

Permalink | Comments (0) | Del.icio.us | Connotea

Heading west for a bit

Friday, July 23, 2010 (16:30 UTC)

I'll be traveling through Xinjiang these next few weeks — to Ürümqi, then by train to Kashgar, or at least what's left of it before the old town is completely razed. From there, I hope to make it to Tushuk Tash, the world's tallest natural arch — you can stick the Empire State building under it. With some luck, I'll bring back 360-degree panoramas of all these places.

Considering that all of Xinjiang was completely without internet most of the past year (there were some troubles) I likely won't be online all that often. Needless to say, if something happens to Google Earth in the meantime, you won't read about it here. See you in a few weeks.

Permalink | Comments (1) | Del.icio.us | Connotea

Newly discovered crater is one of Earth's youngest impacts

Friday, July 23, 2010 (05:21 UTC)

Researchers scouring Google Earth for impact craters have discovered a new one in Egypt, National Geographic reports. Dubbed the Kamil Crater, it is small but very special, because it really is new, in geological terms — just a few thousand years old. So new, in fact, that the elements have not yet been able to erode the ejecta rays. On site, the researchers have been able to collect thousands of space rocks.

These findings were published just yesterday in the journal Science. The full text article requires a subscription, but the supporting online material does not. This material includes satellite images of the crater that contain coordinate information. So without further ado, here's the crater on Google Maps:


View Larger Map

It's a real beauty, and it really is in the middle of nowhere. The imagery we see currently in Google Earth/Maps was collected on May 21, 2006.

[Update 0546 UTC: Some more crater links:
This page explains that the crater was first noticed in February 2009, and confirmed on-site in February 2010. Here's the official page for the crater, labeled Gebel Kamil, at the Meteorological Society. Some guy on Ebay is purporting to sell pieces of the ejecta found at the crater site. It sounds dubious (what's with the piaster notes?), but the page has a good aerial image of the crater:


(Click to enlarge)]

Permalink | Comments (1) | Del.icio.us | Connotea

Google's new, improved map borders - how do they fare?

Wednesday, July 21, 2010 (17:27 UTC)

Google Lat Long Blog has announced new, improved borders for the default dataset of Google Maps (and soon Google Earth). This is much-anticipated news, because many have begun to see Google's border choices as a proxy for international recognition of sovereign territorial claims, with perceived errors loudly contested by aggrieved parties, sometimes even by governments.

As Google's announcement clarifies, the improvements are of several types — increased resolution, changed symbology, and changed borders:

Increased resolution:
In some cases, the resolution of the border is improved, increasing accuracy. Google's post shows an example of this in the Pamirs. From a quick visual inspection, higher accuracy is also visible in the border between Azerbaijan and Armenia. The border between Vietnam and China is also much more detailed — the rougher version of this border prompted a complaint by Vietnam in March 2010 that several of their border towns fell into the Chinese side of the map. That has now been fixed:

Before:
laocai-old.jpg

After:
laocai-new.jpg

There are likely many other instances of such improved borders around the world. (It would be great to get a list or KML file of these changes, just as Google does with imagery updates, for the sake of transparency.)

Changed symbology:
Google writes that some borders, while not changing shape or resolution, have been given a different symbology, to more accurately reflect their status. As an example, Google proffers a section of the border between Somalia and Ethiopia, which is now accurately shown as "disputed".

A quick perusal of another disputed border in Africa, however, leads to a problematic change in the symbology. The Ilemi Triangle is claimed by Sudan, Ethiopia and Kenya, but is de facto controlled and administered by Kenya. The current view in Google Earth recognizes this:

ilemi-ge1.jpg

ilemi-ge2.jpg

(In fact, it might even be better to show the Sudanese and Ethiopian borders of the area as orange, as that is where the "line of control" is for Kenya. For a guide to what the colors (red/orange/yellow) mean, see Google's explanation.)

In Google Maps, however (and I assume soon in Google Earth), it appears as if the disputed triangle has been handed to Sudan:

ilemi-new.jpg

While this change surely is unintentional, the removal of the Sudanese line and the "hardening" of the Ethiopian line can easily lead to mistaken interpretations.

Changed borders:
Google's announcement uses a disputed island straddling the Uruguay-Brazil-Argentina border as an example of how new information has led to qualitative changes in map borders. There is another example in South America — the border between Chile and Argentina has become more accurate, but also, surprisingly, has changed shape. Back in 2007, the Chilean border town of Villa O'Higgins was accidentally placed in Argentina and labelled as such. The error was soon corrected with the border that until now looked like this in Google Earth:

ohiggins-ge.jpg

The new version of the border, already visible in Google Maps, looks like this, with further changes up and down the line:

ohiggins-new.jpg

I trust that this latest (third) Argentina-Chile border dataset is to everybody's liking, as it is not marked as disputed:-)

One place where nothing has changed (as far as I can tell) is in the depiction of the disputed borders between China, India and Pakistan. Arunachal Pradesh is still shown as disputed, with both Chinese and Indian claim lines given equal weight, just has it has been since 2009.

Some Indian bloggers, however, on reading Google's announcement and checking India's borders, are discovering anew Arunachal Pradesh's disputed status, and are not liking it one bit. Trak.in:

I, for one can understand Jammu & Kashmir to be shown as disputed (although, I think only northern part should have been show as disputed), seeing Arunachal Pradesh under dispute is not something I can digest.

Pluggd.in:

Seems like the error in data processing is part of their ‘batch processing’ mistakes? Or an attempt to please Chinese government [J&K dispute is well understood and most of the international news sites shows J&K as disputed area, but not Arunachal Pradesh].

Relax, guys, nothing's changed. In fact, Pluggd.in noted themselves back in October 2009 that the the international version of Google Maps shows Arunachal Pradesh as disputed. You might not like it, but you can't act surprised.

One thing Google could do, however, to make their map of Arunachal Pradesh more accurate, is to draw India's claim line in orange, not red, because India is the de facto administrator of the region, whereas China "just" claims it. (Same goes for Aksai Chin, which China controls but India claims.)

Preah Vihear:
In February 2010 Cambodia vocally protested the depiction of its border with Thailand in Google Earth, especially around the temple of Preah Vihear. (All the gory details are here.) Basically, the demarcation line did not show that the area around the temple is disputed, instead opting to show the maximalist version of Thailand's claim:

Now, it looks like this in Google Maps:

preah-new.jpg

It's an improvement. Still, the road coming in from the left is completely inside Cambodian-controlled territory, so the dotted dispute line should be longer. Also, there is a Cambodian claim line and a Thai claim line — we already know the contours of the Thai claim line, so why not just add the Cambodian claim line and show both in red in Google Earth?

Elsewhere, you can see that the resolution of the border between Thailand and Cambodia has been much improved. Here is the change around the Ta Muen temple complex, which straddles the border further to the west:

tamuen-old.jpg

tamuen-new.jpg

This and other temple complexes along the border have become more heavily contested in the wake of the controversy surrounding Preah Vihear.

Perfection is elusive when mapping, but it is good to see that Google is spending resources on improving the depiction of borders in Google Earth and Google Maps.

[Update 2010-07-22: A Google spokesperson writes:

"I just wanted to clarify one point regarding the Ilemi Triangle: as part of this update, we have revised this area to depict the 1938 "Red Line." This line is considered to more accurately reflect which country controls which portion of the area."

Several independent online resources all state that Kenya is the de facto administrator of the entire area, but I will read up more thoroughly on this dispute before reporting back.]

Permalink | Comments (0) | Del.icio.us | Connotea

links for 2010-07-15

Thursday, July 15, 2010 (19:02 UTC)

Permalink | Comments (0) | Del.icio.us | Connotea

links for 2010-07-14

Wednesday, July 14, 2010 (19:03 UTC)
  • Poly9's product "Poly9 Globe" is self-described as a "cross-browser, cross-platform 3D globe which does not require any download." I suspect it is more of a presentation tool, though, and thus would improve on the maps currently in iMovie (which lack Great Circle animations between places!).

    Still, who knows what it might turn into when connected to Placebase, the other mapping company Apple has bought recently (though neither seems to have direct access to a repository of high resolution imagery, such as what Google and Microsoft have). I'll bet on one thing, though: Whatever comes out of this acquisition, it won't require Flash to run:-)

    One more thought: The more 3D globes the merrier — Google Earth and 3D Google Maps have been at it all alone in the mainstream web mapping space ever since Microsoft abandoned its 3D virtual globe in the latest iteration of Bing Maps.

Permalink | Comments (4) | Del.icio.us | Connotea

Ogle Earth on the radio

Wednesday, July 14, 2010 (08:51 UTC)

On the occasion of Google Maps and Google Earth's 5th anniversary, CBC-Radio Canada's show The Current broadcast a 20-minute segment on the geopolitical impact of these revolutionary new maps over the past few years. The host Jim Brown interviews me and Michael Frank Goodchild, Professor of Geography at UC Santa Barbara.

Listen along at the link above, or download the mp3.

It turned out to be a fun show, with the host focusing on the day that Google accidentally showed Indian-administered Arunachal Pradesh as being Chinese, and the fallout from that errror.

Permalink | Comments (0) | Del.icio.us | Connotea

links for 2010-07-13

Tuesday, July 13, 2010 (19:03 UTC)
  • Oops: Ogle Earth saw it's fifth birthday... last week.
  • Article about the Ogle Earth post calling for an end to the censored Google Maps for China. The author gets a response from Google:

    "Google Maps is one of the products still under review--in particular since the government recently announced new mapping regulations that we also need to take into account," a Google spokesperson said in an e-mail. "We hope to reach a conclusion on this soon; until then, we will continue to operate ditu.google.cn in accordance with local Chinese law. As we've said, we are committed to ensuring that our products in China are not censored."
  • "Government pays for licences after tech giant given free information."
  • or "How Google’s open-ended maps are embroiling the company in some of the world’s touchiest geopolitical disputes." Features Ogle Earth's coverage of several of these disputes.
  • No self-respecting paranoid totalitarian regime can be without at least a few underground airfields. Here they are in Google Earth.

Permalink | Comments (0) | Del.icio.us | Connotea

Azerbaijan to Google: Nakhchivan is (still) ours

Tuesday, July 13, 2010 (16:13 UTC)

Azerbaijan is upset again with Google. Several villages in the Autonomous Republic of Nakhchivan, an internationally recognized exclave of Azerbaijan, are labeled in Google Earth as belonging to Armenia. The Azeri Press Agency (APA) writes:

Google shows Nakhchevan’s villages and regions as Armenian area

[ 13 Jul 2010 12:30 ]

Baku. Aynur Valiyeva – APA-ECONOMICS. Azerbaijan’s area was perverted in Google Earth’s maps again.

According to APA-ECONOMICS, Nakhchevan’s Ordubad, Sadarak regions and Nehrem village of Babek region were shown as Armenian areas.

According to head of press service of Ministry of Communication and Information Technologies Mushvig Amirov, Azerbaijan has already applied to Google’s management about it. “If Google does take measure about our application, the ministry intends to apply to international regulation bodies in accordance with this problem.”

Here's the original in Azeri, which is longer, and which alludes (according to Google Translate) to this sort of thing having happened before, and that Google corrected the mistake last time after an appeal by the government, only to reintroduce these errors recently.

Indeed, we've heard all this before. Back in January 2010 Ogle Earth published the complaint by Azerbajian's government to Google, and confirmed there was a labeling error for Nakhchivan:

gearth-nar.jpg

That popup is no longer shown in Google Earth, because the entire region is no longer labeled. (This is likely as result of the reorganization of the default layers a few months back). Instead, we have popups for the villages of Nakhchivan. The capital (also called Nakhchivan) is labeled correctly:

ge-nakh201007.jpg

But the other villages are not:

nakh2010-ordubad.jpg

It's a clear and straightforward error, which Google should correct (again), especially considering the strained relations between Armenia and Azerbaijan, and the fact that Armenia currently occupies other parts of Azerbaijan.

On a side note: I'm not sure the threat to take this case up with "international regulation bodies" has much promise, considering that there is no such thing when it comes to names and borders. The UN will recognize countries and the International Court of Justice will adjudicate border disputes between countries if asked, but there is no international law that obligates a private corporation like Google to be accurate or fair in its depiction of the world on its maps. Of course, it is in Google's best interests to be accurate and fair, but that's another matter completely.

Permalink | Comments (1) | Del.icio.us | Connotea

Stop censoring Google Maps in China - kill ditu.google.cn

Sunday, July 11, 2010 (06:06 UTC)

Clearly I still have a lot to learn about what you can and cannot get away with in China. A few weeks ago, when Google replaced its automatic redirect of google.cn to google.com.hk with with a clickable image that functionally amounted to a manual redirect, I was convinced that this cosmetic change would not cut it as a ploy to get Chinese authorities to renew google.cn's license to operate in China.

But in the intervening weeks, there were signs that a compromise might be reached: The existing license number appeared on google.cn, as required by Chinese law; and links appeared to three apolitical and China-facing locally hosted Google services — music, products, and translation. This way, Google (and those Chinese authorities favoring Google) could argue that google.cn was not an empty husk used to skirt Chinese law, but a functioning website providing real services, just not search anymore.

And this argument has just won the day. Face has been saved all round, the license renewal has been granted, and everybody gets a solution they can live with.

But missing conspicuously from this newly happy arrangement is any mention of the localized version of Google Maps, ditu.google.cn. It's not linked to from google.cn, but it sure is on the same domain, and served from within China to Chinese users, on servers that need to be in compliance with Chinese law.

ditu.google.cn is, however, linked to from the mainland-China optimized google.com.hk search page that Chinese users land on when they manually redirect themselves from google.cn. (Click on the yellow button once there). If you switch to the Hong-Kong optimized version, you get a link to maps.google.com.hk instead.

The plight of ditu.google.cn has already been documented in detail here on Ogle Earth. Briefly, in May new stricter rules issued by China's State Bureau of Surveying and Mapping required the re-submission of all China-based internet maps for approval. The draft list of 23 approved applicants released at the end of June included no foreign companies, hence no Google, though competitor Baidu is on the list. ditu.google.cn's future is up in the air.

There are differences between ditu.google.cn and maps.google.com.hk: The HK version lets you overlay user-generated content from Panoramio, Youtube, Wikipedia and webcams; The CN version does not, as required by Chinese law. The HK version shows the disputed border areas between China and India inside dotted lines; on the CN map it all belongs to China, as required by Chinese law. (Other changes are cosmetic — the HK version will prioritize local placenames outside China, and its buttons use traditional Chinese characters.)

Objectively, then, the HK version of Google Maps is better in every respect. The CN version of Google Maps is censored to comply with Chinese law.

But didn't Google tell us back in January that it was "no longer willing to continue censoring our results on Google.cn"? And, surely, queries on Google Maps lead to results? By continuing to offer ditu.google.cn, Google has not yet fully implemented its decision to stop censoring in China.

Pace Jeff Jarvis, What Should Google Do?

Well, now that we know that Chinese authorities will tolerate a manual redirect to an uncensored search engine outside the great firewall, why not adopt this model for Google Maps? Kill ditu.google.cn and redirect users to a Simplified Chinese version of maps.google.com.hk (the map tiles remain the same). The HK version of Google Maps is already accessible to mainland Chinese users, just as Google.com and Google.com.hk have always been.

Or better yet, stop kowtowing to China's State Bureau of Surveying and Mapping and let them kill ditu.google.cn by keeping Google off their approved list. Then everybody gets their way — and Google will truly be able to say it no longer censors its China-facing services.

Permalink | Comments (0) | Del.icio.us | Connotea



Search Ogle Earth:
Ogle Earth documents how Google Earth and other neogeographical tools are affecting science and geopolitics. By Stefan Geens. Email me.
Last tracked here:
Get updates via email:

Recent comments

Sep 2 / sagar on
Earth on the mobile phone:
Sep 2 / sagar on
Earth on the mobile phone:
Sep 2 / sagar on
Earth on the mobile phone:

Core news sources (just now)

All Points BlogDirections Media blogs GIS
Digital UrbanInnovations in visualization
The Earth is SquareNASA World Wind developer
Geoblogger.euEuropean GIS aggregator
Google Earth BlogFrank Taylor's news, tutorials and tips
Google Earth DesignRichard Treves blogs about neogeo design
Google Earth CommunityThe official BBS
Google Earth HacksLarge unofficial community site
Google LatLongGoogle's official "geo" blog
Google Maps ManiaGoogle Maps mashups
Google PlanetPushing the API envelope
IMINT & AnalysisOpen source satellite intel
James Fee GIS BlogGIS by an ESRI user
Le Technoblog du LACFrancophone tech and neogeo blog
MapperzNews about mobile and web maps
The Map RoomMap aficionado
Planet GeospatialGIS blogs aggregated
RenaLIdFrench cartography news
Using Google EarthA Googler's tutorials and tips

Archives