Slashdot Mirror

Not to be confused with the state of the art WorldView series of commercial imagery satellites...

They could have purchased TomTom, for example and had everything up and running immediately.

I think it's a bit more complicated than that. From Apple's mapping attribution page:

© 2006-2012 TomTom
Business listings data © Acxiom, 2012.
Map data © AND.
Property parcel data for USA. © CoreLogic Inc., 2012.
Satellite imagery data © DigitalGlobe, 2012.
Map and postal data © DMTI, 2012. This software contains Postal Code OM Data copied by Apple under a sub-license from DMTI Spatial Inc., a party directly licensed by Canada Post Corporation. The Canada Post Corporation file from which this data was copied is dated 2012.
Business listings data © Factual 2012.
Map data © Getchee, 2012.
© INCREMENT P CORP., 2012, http://www.incrementp.co.jp/gc01info/e/legal01.html.
Map data © Intermap, 2012.
Map data © LeadDog, 2012.
Business listings data © Localeze, 2012.
Mapping data for Australia and New Zealand. © MapData Services Pty Ltd., 2012, PSMA http://www.nowwhere.com.au/lic/NowWhereLic.htm.
Map data © MDA Information Systems, Inc., 2012.
Neighborhood data © Urban Mapping, 2012.
Map data © 2012 Waze.
âoeReviews from Yelpâ Yelp, 2012.
(CanVec)
© Department of Natural Resources Canada. All rights reserved.
http://www.geogratis.gc.ca/geogratis/en/index.html
(CGIAR-CSI SRTM)
CGIAR Consortium for Spatial Information, http://srtm.csi.cgiar.org/
Flickr Shapefiles Public Dataset, Version 1.0, http://www.flickr.com/
(GeoNames)
GeoNames and contributors, http://www.geonames.org.
(GlobCover)
© ESA 2010 and UCLouvain, http://www.esa.int/esaEO/index.html

National Aeronautics and Space Administration, http://www.nasa.gov

Contains Ordnance Survey data © Crown copyright and database right 2012. Contains Royal Mail data © Royal Mail copyright and database right 2012. http://www.ordnancesurvey.co.uk/oswebsite/

(OSDM)
© Commonwealth of Australia, 2012. This data has been used with the permission of the Commonwealth. The Commonwealth has not evaluated the data as altered and incorporated within this software, and therefore gives no warranty regarding its accuracy, completeness, currency or suitability for any particular purpose. http://spatial.gov.au

(OSM)
OpenStreetMap contributors, http://www.openstreetmap.org/
(StatCan)
Statistics Canada, http://www.statcan.gc.ca
(TIGER/Line® fi

Digital Globe and GeoEye now operate commercial imagery satellites. That's where Microsoft and Google get their imagery for areas where they don't have close aerial coverage. DoD buys a lot of their info. Best commercial resolution is 45cm. Which, realistically, is enough to find most threats that can be seen from above.

Digital Globe has an analysis of Osama bin Laden's compound in Pakistan..

Digital Globe and GeoEye now operate commercial imagery satellites. That's where Microsoft and Google get their imagery for areas where they don't have close aerial coverage. DoD buys a lot of their info. Best commercial resolution is 45cm. Which, realistically, is enough to find most threats that can be seen from above.

Digital Globe has an analysis of Osama bin Laden's compound in Pakistan..

"Fact 1: this was an old nuclear reactor without a satisfactory containment solution;"

It was also apparently scheduled to be shut down and decommissioned this year because it was the oldest reactor on the site. The irony there is pretty bitter.

"Fact 3: backup generators and batteries were supposed to deal with Fact 2;"

The plant was apparently thoroughly earthquake-proof with two layers of backups to deal with the possibility of a major quake, and they worked fine for that purpose. However, the plant was apparently NOT adequately engineered to handle a tsunami, which A) wiped out plenty of infrastructure on the site, including apparently the backup diesel generators (refer to these DigitalGlobe satellite images showing the damage [PDF]), and B) that fact is freaking stupid! Earthquakes + tsunami go hand-in-hand along Japan's east coast. Building nuclear power plants close to the coast is pretty common for cooling reasons, and *nobody* thought about what the effect of a major tsunami would likely be on ground-level backup systems and connections to the broader power grid????

"Fact 4: you can only have so many on-site backups;"

True. Which is why having some off-site portable diesel generators ready to go makes sense, so that they could be transported on-site if necessary in an emergency. They did that, but I've read some reports that there were initial problems with getting compatible plugs during the hookup, which is bad planning.

This is an extreme situation, no question, and it doesn't necessarily reflect badly on reactors with better designs. But, sheesh, this is a place where there is a good statistical chance of a major earthquake of this type in 50 years of operation, and when the event actually happens, the performance under stress is rather mediocre. Not poor, because they haven't had serious loss of containment, but the performance does not inspire confidence, because even this relatively minor disaster (so far) shouldn't have happened like this. Hydrogen explosion blowing away part of the reactor building? Yeah, that really makes people feel safe. And an explosion like that can't possibly make the control of reactor systems any easier.

You're right that no matter what happens some people will conclude nuclear power is universally evil, but this is a pretty big failure of a system in a country that probably knows more about earthquake preparation than any other in the world. There are going to be a lot of lessons from this, and people better listen.

Here's a sample image. Yes, that's from orbit.

Each satellite images about 1 million km^2 per day, so in 250 days, they can image the entire planet at high resolution. But they'll do the populated parts of the US more often (they can aim the cameras for each pass), so they will pick up many of the balloons.

Microsoft Bing is buying all the data, so it's going on line. The data rate is about 50GB/hour. Start programs looking for red dots.

Here's a sample image. Yes, that's from orbit.

Each satellite images about 1 million km^2 per day, so in 250 days, they can image the entire planet at high resolution. But they'll do the populated parts of the US more often (they can aim the cameras for each pass), so they will pick up many of the balloons.

Microsoft Bing is buying all the data, so it's going on line. The data rate is about 50GB/hour. Start programs looking for red dots.

The next generation WorldView-2 satellite from DigitalGlobe will have a 'Coastal' band (450 - 500 nm) in addition to the Blue, Green, Red and Near-IR bands, specifically to improve this capability (it will have additional Near-IR bands as well).

Radar, as other correspondents point out is not much use in water, but it has been used to study desert sites, such as Ubar in Southern Oman. (This may be in TFA but I still can't get to it...)

That got modded insightful ? Get real!

"No it doesn't make sense. At a time when the internet provide dozens of different way to get that specific information, be it in several other on-line aerial-photo mapping softwares, or on various other online source,"

I don't buy that. Sure there is a lot of stuff on the internet but super high res satellite photographs of sensitive government installations ? Give me a break. That wont happen until every tom dick and harry gets his own satellite.

There are multiple sources of satellite and aerial photographs, especially when we are talking about public areas like government buildings. Someone already mentioned digitalglobe as a source google themselves might use (and which others use -- it draws on multiple sources apparently itself), and this site lists several different sources for satellite imagery which is available to the public. So yes, there are a lot of toms dicks and harrys out there with satellites. Welcome to the 21st century.

"And besides, it's just security through obscurity, and we all know very well how much that strategy works well."

Your trying to draw a parallel between two completely different fields with different goals and purposes. A government installation is not "open source software" that everyone gets a chance to get a peek see and everyone by and large is benevolent when looking at the source. When you have a country's defense on the line and a lot of baddies want to maim and kill people, obscurity is one of the best weapons. What next ? Show people on the witness relocation program on national television ?

No, the problems and solutions are not at all as disparate as you claim, by your own admission. Government facilities are indeed places where everyone gets a peek and by and large everyone is benevolent looking at the source. If you are talking about less public areas, like military installations, you would be surprised perhaps how much the public is allowed to see. Some military bases are pretty much completely ope to civilian traffic, and those which aren't often have very close perimeters. Even places like Area 51 regularly attract civilians who are able to record an awful lot of information about location of buildings, security measures, and activity. A telephoto lens and/or a telescope or set of binoculars can reveal an awful lot with little chance that the observer will be observed. And those people mostly are not trying to blow the place up.

The argument about security through obscurity is not about the moral question of keeping secrets. It is an indictment based on the fact that any security plan that cannot withstand scrutiny is weaker than one which can. And in this case the attempt to restrict information from one source when there are many others including local surveillance is, besides being a fool's errand in itself, indicative of a fear that the security measures in place will not measure up to an actual attempt. It is also very much in the same vein as the "fig-leaf" faux security that has marked the "war on terror" in general; it is clearly a justified criticism.

I don't know how the hell issuing a cease and desist so as to hold onto your digitalglobe reseller account could be construed as evil.

If Digitalglobe (who are the providers of Google's content on google earth) decided Google were breaching their TOS and decided they'd be better off keeping their imaging to themselves then everyone loses, including anyone using local.google.com and Google Earth.

Seems to me that Google are trying to keep a good thing going, and being IMHO reasonably respectful towards the Gaia project's authors.

You can see much better high-resolution imagery here and here (the images at Eurimage have been sharpened a bit and have better contrast). Don't forget to check out the imagery of Biloxi - which has basically gone the way of Banda Aceh.

Digitalglobe has a larger image. http://www.digitalglobe.com/images/katrina/new_orl eans_dwtn_aug31_05_dg.jpg

If you don't care about the entire world at once you can get 60cm, yep 0.6meter res shots by using Digital Globe's http://www.digitalglobe.com/ QuickBird satellite built by Ball Aerospace & Technologies http://www.ballaerospace.com/

Here are the specs: http://www.digitalglobe.com/product/basic_imagery. shtml/

If you don't care about the entire world at once you can get 60cm, yep 0.6meter res shots by using Digital Globe's http://www.digitalglobe.com/ QuickBird satellite built by Ball Aerospace & Technologies http://www.ballaerospace.com/

Here are the specs: http://www.digitalglobe.com/product/basic_imagery. shtml/

Very difficult: DigitalGlobe (then EarthWatch, with EarlyBird), Space Imaging (Ikonos), Orbimage (Orbview) all lost high resolution satellites before they became operational - these early high-res satellites were in fact based pretty much on off-the-shelf (but space qualified) parts. Landsat 6 was lost on launch and Landsat 7 suffered a crippling failure about a year ago. (BTW the EarlyBird launch on a Russian Cosmos, was said to have cost about $7 million, not 50 thousand). Having got it up there, and got it working, you then need a ground segment too, which is decidedly non-trivial.

DigitalGlobe has a amazing gallery of tsunami images, including several PDF analyses with before and after images of Meulaboh, Gleebruk, Banda Aceh and Sri Lanka.

DigitalGlobe has a amazing gallery of tsunami images, including several PDF analyses with before and after images of Meulaboh, Gleebruk, Banda Aceh and Sri Lanka.

oh wow... check out the first after-picture (the first in the series). That bend in the road near the top of the image, there is still one tree left... amazing that everything around it got flattened, and one tree was left standing.

DigitalGlobe have the satellite image before and after the tsunami in Sri Lanka.

The U.S government has always been able to do this with satellites licensed by them - it is called 'shutter control'. In the case of Afghanistan, they preferred to buy the output of Ikonos (the only Very High Resolution satellite then operating) not only to keep it from anyone else but because the data was actually useful - coverage is always a problem with satellite data. Not only that - I think you'll find private companies operating satellites like QuickBird and Ikonos will sell you exclusive rights to imagery if you are willing to pay enough.
You can't keep the imagery from the govt, however. And did you know they will not let those companies sell imagery over Israel or the occupied territories at less than 2 metres resolution, as against the 60 cm capabilities of QuickBird for example.
All this legislation does is prevent using the FoI Act to circumvent restrictions already in place - for US-based companies. And they're not worried about weather sats or Landsat, I'm sure...

I know at least the Quickbird satellite can be turned around to snap pictures of celestial objects. I think there is a picture of the moon taken with Quickbird on the www.digitalglobe.com website. What I really want is a high-res picture of another satellite - like maybe hubble?

First of all, Ikonos was launched in -99, not -95.

And even so I think its old news: Quickbird 2 has been "in the air" for over a year now with superior resolution (61 cm panchromatic).

Commercial spy satellites like Quickbird don't take pictures of every square of earth. Instead they photograph sites users request photographs of.

And they might photograph the same site at different points in time.

So unless you noticed something like that an important Soviet base was completely missing from the set, you wouldn't be able to tell what images were still classified.

--Pat / zippy@cs.brandeis.edu

Looks my 'cache' is ok according to their usage guidelines. (Even if it wasn't I still might be able to pull off the whole 'fair use' thing.)

the following is taken from https://www.digitalglobe.com/?goto=gallery/downloa ds :

Usage Guidelines

ATTRIBUTION

Credit: DigitalGlobe

Copyright © DigitalGlobe. All rights reserved. Online and news media distribution or publishing requires permission from DigitalGlobe.

Permission is granted to electronically publish, publish in hard copy and broadcast these satellite images if proper attribution to DigitalGlobe is provided.

The words "digitalglobe.com" must be hot linked to the DigitalGlobe web site (www.digitalglobe.com) for each and every use. Any other use of materials -- including resale, distribution or reproduction, or for purposes other than noted above -- without the prior written permission of DigitalGlobe is strictly prohibited. DigitalGlobe makes no claim or representation, and accepts no responsibility, regarding the quality, nature, or reliability of any materials publishing these images or Web sites linking to these images.

Here are the direct links to the images:

http://www.digitalglobe.com/imgs/basic/pan/thumb_p an_61cm_ff.jpg

http://www.digitalglobe.com/imgs/Standard/ms/thumb _ms_2pt8m_ff.jpg

Or if they go down too; Google:

http://images.google.com/images?hl=en&q=+site:www. digitalglobe.com+www.digitalglobe.com

Here are the direct links to the images:

http://www.digitalglobe.com/imgs/basic/pan/thumb_p an_61cm_ff.jpg

http://www.digitalglobe.com/imgs/Standard/ms/thumb _ms_2pt8m_ff.jpg

Or if they go down too; Google:

http://images.google.com/images?hl=en&q=+site:www. digitalglobe.com+www.digitalglobe.com

And belive me, Space Imaging's investors put up way too much money to let the Pentagon commandeer a satellite that could potentially bring them very lucrative returns on their venture capitol. The Pentagon couldn't afford to pay these folks enough to shut them up. Besides, launch failures happen way too often to divert Occam's Razor to point to any other conclusion. The launch system is still the most risky and unreliable rung on the ladder to space. Given number of nearly identical imaging satellites on the way, it will be the launch systems that will sort out the winners from the losers.

I must agree with your comment above comment. The launch business is very risky no matter how many precautions one can realistically take. Part of the success will come from luck.

I want to point out that QuickBird 1 will be in a non-sun-syncronous orbit to permit customers the option of viewing at various times of day. I don't know the orbit drift to say it is possible to view the same location more than once in the same day though.

Incidentally, a good overview of the applications of remote sensing can be found at Earthwatch. I'm sure the other companies have something similar.

~afniv
"Man könnte froh sein, wenn die Luft so rein wäre wie das Bier"

As one of the systems engineers responsible for designing the optical and detector subsystems used on both the OrbView-3 and OrbView-4 satellites, I can tell you what these reconnaisance systems can and can't do.

Resolution translates to two seperate functions: detection and identification. At 1-meter resolution, these cameras can detect an automobile, a tree, or an individual steer. It's possible to detect even an individual sunbather (if she's got a dark tan and is contrasted against white sand), but unless you knew it was a sunbather beforehand, you wouldn't be able to tell if its a he or a she, or even tell it apart from a vacant beach chair. That would be identification. To identify an object, its image must span several pixels on the detector, if not dozens. Therefore, Ikonos could detect a police car, but it couldn't read the call letters on its roof to identify that car.

Further, most of these systems are limited by the power available to point the satellite accurately and to run the camera. Without a monstrous solar array to provide enough power to continuously run attitude control, transmitters and detectors, the spacecraft must rely on the energy stored in its batteries. Thus, the cameras are limited to less than a half an hour of imaging per 90-minute orbit. Less if the sattelite is required to reorient itself in order to photograph noncontiguous regions. By no means can these systems take daily global high-resolution photos of the entire earth.

These satellites generally orbit somewhere between 700 and 800 km above the earth, in a polar orbit inclined nearly 90 from the equator. This translates to a "ground speed" of about 7 km/s. Most operators aim for a "sun synchronous" polar orbit, meaning that the spacecraft passes the equator at the same times every day, once going northbound (ascending node) and once headed south (descending node). The OrbView satellites ascending node crossing time is 10:30 am, for example. All of these numbers add up to the fact that at best, these cameras can only take one good picture of you each day. If someone were really motivated, they might be able to take two or three, but only one could be from directly overhead.

This is true also because the detectors on these cameras are linear photodiode arrays, made by folks like Kodak. These detectors are almost identical to the ones found in your scanner or office copier. And they work pretty much the same way: the satellite carries the detector overhead at a constant rate, and samples the array thousands of times per second. The successive linear images are piled up to create a 2D photograph. It's even possible to take two "scans" of a region from different angles, and then process this information on the ground to create a stereo 3D image. If you have the money to pay for it...

Each of these systems carries more than one array. The 1-meter resolution arrays are panchromatic or black-and-white, and are sensitive to the entire range of visible and Near-IR light: wavelengths of 450 to 900 nanometers. There are also 4-meter resolution multispectral arrays, with pixels 4x as big as the panchromatic arrays. These are four arrays have color filters to distinguish radiance in four color bands, generally chosen in the same bands as used on LandSat systems: 450-520nm, 520-600nm, 625-695nm, and 760-900nm.

OrbView-4 is unique among these systems, carrying a hyperspectral detector. Hyperspectral detectors generally employ an imaging spectrometer: the telescope image is passed through a very narrow slit, focussed on a diffraction grating, and the dispersed light is then imaged on a 2D CCD array. This scheme provides hundreds of linear images covering a continous series of narrow wavelength bands. OrbView-4 will cover the range of 450-2500 nm, extending from blue deep into the infrared. This kind of detailed information sounds like a gold mine, but there is a slight risk - it's the first of its kind to provide this sort of information, so there's no established demand. In fact, scientists are still figuring out what to do with this much data...

What are the theoretical limits on these kinds of systems? One satellite in geostationary orbit, with unimaginatively expensive optics and an energy-gluttonous data system could indeed continuously monitor one-third of the globe. But in practice, a system like this would cost GigaDollars and prove very unreliable. The theoretical limit of resolution is approximately 10 centimeters, without using some sort of technique to instantaneously correct for atmospheric disturbances. And then there's the whole question of shipping all that data down to the folks who don't have time to look at it all...

And belive me, Space Imaging's investors put up way too much money to let the Pentagon commandeer a satellite that could potentially bring them very lucrative returns on their venture capitol. The Pentagon couldn't afford to pay these folks enough to shut them up. Besides, launch failures happen way too often to divert Occam's Razor to point to any other conclusion. The launch system is still the most risky and unreliable rung on the ladder to space. Given number of nearly identical imaging satellites on the way, it will be the launch systems that will sort out the winners from the losers.

I'm happy to entertain any questions you might have - be sure to de-spamproof my email address.

As one of the systems engineers responsible for designing the optical and detector subsystems used on both the OrbView-3 and OrbView-4 satellites, I can tell you what these reconnaisance systems can and can't do.

Resolution translates to two seperate functions: detection and identification. At 1-meter resolution, these cameras can detect an automobile, a tree, or an individual steer. It's possible to detect even an individual sunbather (if she's got a dark tan and is contrasted against white sand), but unless you knew it was a sunbather beforehand, you wouldn't be able to tell if its a he or a she, or even tell it apart from a vacant beach chair. That would be identification. To identify an object, its image must span several pixels on the detector, if not dozens. Therefore, Ikonos could detect a police car, but it couldn't read the call letters on its roof to identify that car.

Further, most of these systems are limited by the power available to point the satellite accurately and to run the camera. Without a monstrous solar array to provide enough power to continuously run attitude control, transmitters and detectors, the spacecraft must rely on the energy stored in its batteries. Thus, the cameras are limited to less than a half an hour of imaging per 90-minute orbit. Less if the sattelite is required to reorient itself in order to photograph noncontiguous regions. By no means can these systems take daily global high-resolution photos of the entire earth.

These satellites generally orbit somewhere between 700 and 800 km above the earth, in a polar orbit inclined nearly 90 from the equator. This translates to a "ground speed" of about 7 km/s. Most operators aim for a "sun synchronous" polar orbit, meaning that the spacecraft passes the equator at the same times every day, once going northbound (ascending node) and once headed south (descending node). The OrbView satellites ascending node crossing time is 10:30 am, for example. All of these numbers add up to the fact that at best, these cameras can only take one good picture of you each day. If someone were really motivated, they might be able to take two or three, but only one could be from directly overhead.

This is true also because the detectors on these cameras are linear photodiode arrays, made by folks like Kodak. These detectors are almost identical to the ones found in your scanner or office copier. And they work pretty much the same way: the satellite carries the detector overhead at a constant rate, and samples the array thousands of times per second. The successive linear images are piled up to create a 2D photograph. It's even possible to take two "scans" of a region from different angles, and then process this information on the ground to create a stereo 3D image. If you have the money to pay for it...

Each of these systems carries more than one array. The 1-meter resolution arrays are panchromatic or black-and-white, and are sensitive to the entire range of visible and Near-IR light: wavelengths of 450 to 900 nanometers. There are also 4-meter resolution multispectral arrays, with pixels 4x as big as the panchromatic arrays. These are four arrays have color filters to distinguish radiance in four color bands, generally chosen in the same bands as used on LandSat systems: 450-520nm, 520-600nm, 625-695nm, and 760-900nm.

OrbView-4 is unique among these systems, carrying a hyperspectral detector. Hyperspectral detectors generally employ an imaging spectrometer: the telescope image is passed through a very narrow slit, focussed on a diffraction grating, and the dispersed light is then imaged on a 2D CCD array. This scheme provides hundreds of linear images covering a continous series of narrow wavelength bands. OrbView-4 will cover the range of 450-2500 nm, extending from blue deep into the infrared. This kind of detailed information sounds like a gold mine, but there is a slight risk - it's the first of its kind to provide this sort of information, so there's no established demand. In fact, scientists are still figuring out what to do with this much data...

What are the theoretical limits on these kinds of systems? One satellite in geostationary orbit, with unimaginatively expensive optics and an energy-gluttonous data system could indeed continuously monitor one-third of the globe. But in practice, a system like this would cost GigaDollars and prove very unreliable. The theoretical limit of resolution is approximately 10 centimeters, without using some sort of technique to instantaneously correct for atmospheric disturbances. And then there's the whole question of shipping all that data down to the folks who don't have time to look at it all...

And belive me, Space Imaging's investors put up way too much money to let the Pentagon commandeer a satellite that could potentially bring them very lucrative returns on their venture capitol. The Pentagon couldn't afford to pay these folks enough to shut them up. Besides, launch failures happen way too often to divert Occam's Razor to point to any other conclusion. The launch system is still the most risky and unreliable rung on the ladder to space. Given number of nearly identical imaging satellites on the way, it will be the launch systems that will sort out the winners from the losers.

I'm happy to entertain any questions you might have - be sure to de-spamproof my email address.

Earthwatch lost a similar remote sensing satellite, Earlybird, in Dec. 1997 I think. At least it made it to orbit. I don't remember the reason, but I think the batteries somehow drained too fast (some instruments were turned on too early or something like that) and therefore knocked out the comm system.

Fortunately it was insured, and the money is being used to finish the next generation satellite. Hopefully it will be successful later this year.

~afniv
"Man könnte froh sein, wenn die Luft so rein wäre wie das Bier"