Imaging Notes magazine has published high-resolution satellite pictures taken by commercial operator Digital Globe of a “secret” tunnel for China’s nuclear submarines. The tunnel is easily located on Google Earth with a bit of sleuthing.

Here is a Washington Times story about the image, here is the actual image by Imaging Notes, and here is the page documenting the site by the Nuclear Information Project, which includes an account of how the tunnel was built.

I’ve turned the images into overlays at the correct location on Google Earth. Here is the KMZ file, posted to Google Earth Community.

Suddenly, last month’s thought experiment is no longer so abstract.

Time Magazine reporter Christine Gorman reports on her global health blog that Google Earth is being used by US poultry veterinarians to double-check the accuracy of coordinates for hen houses, in case an outbreak of the H5N1 avian flu virus requires quarantining decisions for poultry farms.

(Yes, she writes “double-check”, not “check”. If Google Earth were the only geospatial tool used for finding poultry farms in something as important as the defence against H5N1, I’d somehow be a little worried, as the geolocation and map data isn’t flawless, and isn’t expected to be, for a free product.)

> RoboGeo, a geocoder for photos, sees an upgrade to v3.2, adding the ability to write location data to the EXIF headers of Adobe’s DNG digital negative format.

> One way for smaller linguistic groups to even the playing field is to release their own toponomies for Google Earth. Case in point: Irish and English place names, separated, for Ireland. (qv. the tactics of a Catalan a few weeks ago.) (Via)

> Londonist discovers Digitally Distributed Environments, calls it their “new favourite blog”.

> Film buff Mark Allen locates 11 of his favorite film locations in Google Earth, and meticulously details where each shot was filmed. Absolutely fascinating — I had no idea Blow Up was filmed in London’s Maryon Park, for example, I only wish he’d given us the KML files instead of the coordinates. Another reason to adopt the gref attribute for anchor tags.

> HSA shows off screenshots (alas no actual KML files) of a product that manages boat traffic in the Great Barrier Reef using Google Earth. (Via Connotea’s GIS bookmarks)

> All Points Blog catches a description of the Google Earth team in action in Time Magazine, and also calls Wired magazine on a GIS rookie error.

> Biomapping.net adds “emotion” maps from areas outside London — Siena, Rotterdam, and a spot in Finland. If only this project used input that looked less like random numbers and more like real data.

Responding to yesterday’s archeology link, David Haskiya of Sweden’s Office of Antiquities writes that they too have been been experimenting with Google Earth.

His office uses an Oracle spatial database to keep tabs on all of the country’s historical monuments, from the Stone Age on, and they’ve now written up some java that converts all this geodata into KML. David’s made a sample available, of Stockholm’s historical monuments: I’ve wrapped it into a KMZ network link ready for download (KMZ).

Each placemark links to the corresponding record’s webpage, and while the info is in Swedish, it’s worth taking a look, as at the bottom of these pages are links to scanned primary sources. It really feels like there is some amazing potential for data mining here, with this sample just scratching the surface.

The transformative power of geobrowsing the database is unfortunately proving to be something a problem, says David (translated from Swedish):

The theft of church objects and the plundering of historical remains means public authorities are a little nervous about making the data public in this fashion. We continue to discuss internally about what kind of restrictions should be implemented.

(If you speak Swedish, here is more context.)

Currently, searching the database is possible only by userid/password, but available to all at museum libraries around the country. Making this data available in Google Earth is thus less of an issue than whether the data should be publicly available at all. Sweden is a large country with few people and thousands of remote archeological sites. Making the location of each site public would make the task of would-be robbers too easy, goes the argument.

On a technical note: The database looks accurate to 1/1000 of a degree. In the center of Stockholm, this causes the placemarks to stick to a grid; Google Earth can clearly support much more precise positioning.

Given access to such a database, I could imagine researchers truly benefiting from being able to map search results dynamically, such as, for example, Viking settlements by decade. i can’t wait for Google Earth to add that time browser:-)