Slashdot Mirror
The Sharpest Ever Global Earth Map
Roland Piquepaille writes "The GLOBCOVER project, started by the European Space Agency (ESA), has a very simple goal. It will create the most detailed portrait of the Earth's land surface with a resolution three times sharper than any previous satellite map. The image acquisition will be done throughout 2005 and use the Medium Resolution Imaging Spectrometer (MERIS) instrument of the Envisat environmental satellite. To create this sharp map, the GLOBCOVER project will analyze about 20 terabytes of data gathered by the European satellite. When it's completed, the map will have numerous uses, 'including plotting worldwide land use trends, studying natural and managed ecosystems and modelling climate change extent and impacts.'"
time to put something interesting on the roof for when the sat passes over
The image acquisition will be done throughout 2005 and use the Medium Resolution Imaging Spectrometer (MERIS) instrument of the Envisat environmental satellite.
Surely the High Resolution Imaging Spectrometer would be more appropriate?
Niles will be happy to hear she's orbiting the planet...
Obliteracy: Words with explosions
I can see my house!
shit... better throw a tarp over my missles...
Now when this is incorporated into maps.google.com I'll be even happier..
A computer without Microsoft is like ice cream without ketchup.
acquiring images with a spatial resolution of 300 metres
Surely the resolution will be better than 300 m, or am I missing something?
(BTW, I *highly* recommend checking out World Wind if you haven't seen it. It is one of the most awesome programs ever to exist, bar none.)
main(c,r){for(r=32;r;) printf(++c>31?c=!r--,"\n":c<r?" ":~c&r?" `":" #");}
It's true! Just look here.
Let's hope a vessel doesn't go too far west of North America.
I'm Trappped at Berkeley.
Will I be able to see the titties around the world's pools and beaches? If not, it all seems like a bit of a waste.
I better cover up my illegal plants..
Beat the computer, program your life.
Well but isn't this data for which I've paid with my tax euros already? Why does the public who financed it not get free access to that data?
While we're at it, can other Slashdotters perhaps point to links of freely available satellite imagery? Is there any kind of systematic coverage of the planet we live on which is freely available to everyone who does happen to live here?
Now, the real question will be how much ESA will want for a copy of this map, or portions of it.
"The image acquisition will be done throughout 2005 and use the Medium Resolution Imaging Spectrometer" So if they want to create the sharpest map, wouldn't it be best to use the High Resolution Imaging Spectrometer? Duh....
"To create this sharp map, the GLOBCOVER project will analyze about 20 terabytes of data gathered by the European satellite." How long is that going to take?
The estimate is that up to 20 terabytes of imagery will be needed to mosaic together the final worldwide GLOBCOVER map - an amount of data equivalent to the contents of 20 million books.
Why do writers insist on making these kind of useless comparisons? Is there any research that indicates the average book contains the equivalent of one megabyte of data? Especially one megabyte of imagery? Will this really help a layperson quantify a terabyte?
This just in: The human brain is capable of storing an amount of data equivalent to 68 quintillion index cards.
Yes, *that* Bob Vila.
So the most important question is how big does my sign have to be?
-- Thou hast strayed far from the path of the Avatar.
That area without the trees...it's called the ocean.
MERIS grabs data in up to 15 spectral bands with 16 bits per band per pixel. Its only has a 1/2 megapixel imager (842 x 691), but the RAW images are 17.5 MB.
Multispectral data is great for identifying ground cover (e.g, classifying the types of plants, health of plants, minerals, etc. on the ground). Sometimes, it's more valuable to know the materials on the ground than to see the geometric detail.
Two wrongs don't make a right, but three lefts do.
It'd be great if I could download freely the maps. I know, it's a huge effort, but so are my efforts when I work and pay my taxes.....
Great! I plan on burning a few messasges for my favorite politicians into the field behind my house. When do they pass over Waddington, NY?
You could download Nasa Worldwind software for free.
There are some issues with Landsat7 data, but hopefully they will get fixed soon.
Its awsome piece of software! offers 7m resolution globally and offers 1m resolution for USA.
On the other hand, ESA has always been stingy in giving access to data. It took them a while to release Titan images; as opposed to Nasa who makes them available almost instanteneously.
I guess thats the difference between the cultures!
- Sh!t
No- and I mean *NO* nude sunbathing for any of y'all until 2006. We just can't afford to take the chance.
I came, I saw, I left. It looked better in the brochure.
The fact is, there is more vegetation on the planet now than there was 100 years ago
I call bullshit.
State your references or admit you're pulling 'facts' out of your ass.
____
~ |rip/\/\aster /\/\onkey
I, for one, welcome our new gloabal mapping overlords.
Google Maps Europe at last?
:wq
It's a sharp map, you said? Sharper than the sharpiest sharp map ever made sharply of a sharp?
How about "higher resolution?" Or are you actually talking about a map that I can cut myself with?
Because honestly, Roland... your prose make me want to cut myself. And I stopped that in high school.
Reading the article, it really is 300m/pixel. This is 400x lower resolution than the 15m Landsat data that is available as a basemap in Keyhole, Google Maps, and other providers.
The reason this data is interesting is its 15-band nature and the amount of analysis and extraction that can be done from it.
For pretty pictures, there are plenty of better sources.
And what about Area 51, the Green Zone in Baghdad, and all of the nifty places on the earth that we don't typically get to see via satellite photos?
At 300 meter per pixel resolution, you're not going to see many details anyway.
There is a huge difference between the 'google maps' visual coverage, and this, which is a 'earth surface condition map'. For one, the resolution here is pathetic compared to that of the 'google maps', but it has a completely different goal. It is intended to show details about land, instead of whether or not someone mowed a 'hi mom' message into their lawn. Move along, you won't be seeing that any time soon.
Current multispectral stuff is more than 300m? Interesting, it would miss entire steams, even rivers and stuff.
Good point and that seems plausible, but is not entirely true. With a good pixel-mixing analysis you can resolve stuff inside the pixel. The key is having a clean spectral model for the terrain versus water and being able to say that a given pixel looks like its 90% trees and 10% fresh water. "Unmix" enough pixels and you can string them together to find streams smaller than 300 m wide that cross an expanse of forest or grass-land. A string of 10% water pixels in a 300 m resolution image is probably a 30 meter-wide stream. Moreover, the plants around water often different substantially from the plants in the drier surrounding areas -- making it even easily to infer the existence of creeks that are far smaller than the resolution limit.
Two wrongs don't make a right, but three lefts do.
How is 300 Meters sharper than the 15m NaturalVue data set from earthsat?
People who think they know everything really piss off those of us that actually do.
What about mapping the sea surface? This just adds more fuel to the argument that Europeans are biased in favor of land.
I'm not sure about world-wide, but it is true for North America.
The total amount of large-tree standing timber in the US has increased by 30% since 1950. US forestlands covered 732 million acres in 1920; today they cover 747 million acres.
we're pumping the atmosphere full of food and [the vegetation is] growing like mad trying to eat it all..
:)
Very funny post
I'll probably be modded down for this...
US forestlands covered 732 million acres in 1920; today they cover 747 million acres.
A gain of 15 million acres over 85 years.
Roughly 176,471 acres gained a year.
Meanwhile, worldwide, we are losing rainforests to the tune of 1.5 acres per second.
That works out to 47,336,400 acres lost a year...more than 268 times the rate forestlands are growing in the U.S.
____
~ |rip/\/\aster /\/\onkey
Holly Crap People RTFA, and if you don't understand TFA then STFU!
/.ers that compare everything to google, but for real scientists. K, I'm done venting now.
This is not imagary for you to be able to spot yourself fingering,mooning or otherwise exposing yourself to the sattelite. It is a Land Cover map, meaning it will be used for classifications . . ie, Forest, urban, Prairie, Water etc. The reason such a teribly poor, crappy, worse than google maps resolution is used, is becuase if you havew a very fine resolution, classification becomes difficult. In a city, for instance, every tree would become forest, even if right next to a house. Some degree of homogeneity is needed in the image to make classification meaningful. It will be a useful project, maybe not for stupid
Seriously, are we going to help create the maps? No. Are we going to be able to provide assistance in any way? No. Is there anything to see yet? No. So why not post a story once the maps are out?
You got your calculation wrong because you don't get the operating principle of the MERIS instrument. It is not a half megapixel ordinary camera that with some magic collects 15*16 bit per pixel (e.g. by taking 15 images in close succession with a changing filter wheel).
As the page on MERIS says, it is a 'pushbroom imaging spectrometer'. 'Pushbroom' means that instead of a rectangular field of view like a normal camera, it has a line-shaped field of view. An image is formed by continuously observing the single line and wait for the the satellite to 'sweep' the whole planet. 'Imaging spectrograph' means that the previously mentioned line is spectrally separated (with a prism or a grating) perpendicular to the line. You then get an image on your CCD where one direction corresponds to the distance along the line and the other is formed by the different colors.
The optics details of the system are here, an description of the CCD is here. If I understand it correctly, one line is 740 pixels wide and the colors are 'binned' in 15 different spectral bands. That gives you 740*15*2=22200 bytes per line. How much that would be per picture depends on the height of the image, which depends on how long you collect data. In principle you could have it sweep the whole planet non-stop.
karma police: arrest this man, he talks in maths; he buzzes like a fridge, he's like a detuned radio. [radiohead]
I better (Insert totally witty action here) (Insert hilariously unpredictable object here)!
Support Liberty, Support Ron Paul
That number does not make a distinction between old-growth forest which is bio-diverse and mono-culture treefarms which have very limited habitat for anything but genetically-engineered fast growth pine trees.
Starman97@Gmail.com (bring it on spammers)
Heh, why not also analyze traffic patterns and see exactly which roads get crowded and when, and by doing this over a period of several years, find patterns related to various things, like which day of the week it is, what time it is, what holidays are going on in various parts of the social, political, and religious world, and other such things. Then, combine all this information into a computer program that allows drivers to pick the best possible route for their drive, and continue adding realtime information from the satellite into the system on a daily basis. Cool.
Soon I'll no longer have to use the mirrors at Blockbuster to look down women's tops.
Dear Will, the plums were poisoned. -- Cheese Club
How many atlases would fit into 20 terabytes?
If an atlas has 200 A4 pages,
each being 96,9375 square inches,
fulfilled with maps at 600 dpi,
32 bits per dot,
and one dot is 4 bytes;
then one page is 140 megabytes,
that multiplied with 200 equals 28 gigabytes.
Then 20 terabytes equals to 20000/28 = 714 unique true-color atlases.
The privacy nuts that frequent this board think this is cool but will complain when they see a camera watching the floor of a store?
Can I read the license plate off the car of my ex-g/f's new b/f when it's sitting in her driveway?
Inject.
Maybe he was referring to this trend?
"God fights on the side with the best artillery." - Napoleon, Marshal of France - speaking truth to power
Mod this up! This is exactly the problem.
Most satellite maps have high detail pictures of american and european cities, and very low resolution of remote villages in afghanistan and the gobi desert. I wonder if this image will really map all of the world's surface.
"Give orange me give eat orange me eat orange give me eat orange give me you." -Nim Chimpsky
Unless you want the US to invade Brazil and force them to stop, I don't see why you're bringing this up.
Unless the US has been hiding it's mighty amazon rain forest from the world.
Democrats or Republicans. They are both taking us to the same place and they are not afraid of us anymore.
Will I see this on google maps by 2006?
Could Jesus microwave a burrito so hot that he himself cou
CARTOSTAT sent by India this week has a capability to take photographs with resolution of 2.5 mts.
By the brave folks of The Harvard Pep Squad!
Clearing up a couple of confusions seen above.... First, there are (at least) three types of resolution in remote sensing: spatial, spectral, and temporal. In other words, how small a thing can you see, how much 'color' information can you see, and how often can you see, any one area. Color is actually a misleading term here because many of the 'things' that a sensor sees are not visible to the human eye. Spatial Resolution: The best spaceborne non-military platforms are now at .5m pixel resolution. That means you can see the shadow of a person on a bright day, can get a good view of a large building, can map streets and traffic, and so on. There are military systems that are in the sub-10cm range and there are persistant rumours of systems in the 1-2cm range using adaptive optics.
Spectral Resolution: First, most sensors look at fairly narrow wavelength ranges. This means that a sensor seeing 'red' isn't seeing all of what YOU would call red, it's seeing a slice of that. Most sensors are themed - they are built so see wavelengths that allow fairly easy discrimination between natural features in one subject area. Landsat has (mostly) bands for discrimination of healthy and unhealthy vegetation from soil, rock, and water - it was originally a 'biomapping' satellite. One rock mapping band was added so it is somewhat useful for rock discrimination too. You can always add more bands to a satellite, but this dramatically increases download time, and many modern satellites are essentially limited by the power and time to download what they see: a high spatial resolution, high spectral resolution satellite would spend lots of time just downloading.
If a satellite has a few bands (up to say 20... no real rule here) it is called multispectral. If it has one 'all spectra together in one grey image' band that is called a panchromatic sensor. If it has a lot of bands (... more than 20) in one sensor that is called hyperspectral. As you add bands, download time increases and calibration sensitivity problems get severe. Most hyperspectral systems currently used are airborne.
There are also active satellites - these transmit radiation and then wait for the 'echo.' Canada's RADARSAT is an example. The Shuttle Radar Topography Mission (SRTM) is another example of an active system based on RADAR. The 'cutting edge' in active systems right now is in LIDAR - laser ranging for 3d - but these are pretty much airborne at the moment, though this will undoubtably change.
Temporal resolution means 'how often do you see the same area.' Some satellites are in stationary orbits - they look at the same thing all day - and some are in orbits that have return periods. Some are mildly steerable (pointable, for example) and can do oblique views for 3d. Landsat, for example, only sees the same area every few weeks (can't remember the return period to be honest....) whereas lots of weather sats just sit in one place.
So.... depending on what you mean by resolution, and especially depending on what you are looking at, different sats are useful.
Finally (... will he ever shut up ...) once you have a 'scene' for an area there is a huge amount that can be accomplished with image processing, including sharping, fusing with other types of imagery, construction of 3d views, ....
So....
the point of this EU initiative is to have a medium spatial resolution, medium-to-fairly-high spectral resolution image of a large area, calibrated properly so that image processing (e.g. to identify health in crops) is meaningful across the entire composite scene, so scientific studies can be done... ...
RandomRob
The Berkeley-Iowa Naked People Finder project by Margaret Fleck and David Forsyth in ~1996 was fairly good at deciding whether pictures had naked people in them or not. (They also did a G-rated version that looked for horses.)
Bill Stewart
New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
The image acquisition will be done throughout 2005 and use the Medium Resolution Imaging Spectrometer (MERIS)
Maybe it's just me, but if I wanted to make the "Sharpest Ever Global Earth Map," I'd use a High Resolution Imaging Spectrometer. Or maybe this is about job security, ensuring they can do it all over again next year. Ah, mediocraty.
Somebody needs to tell the ESA that doing a half-assed job is the American way, damnit.
https://www.eff.org/https-everywhere