Orbital surveillance satellites now exceed 1 inch resolution.

[RGClark]

The most recent public estimates of spy satellite resolution capabilities give them as about 10 centimeters, 4 inches. However, it is widely known that the most advanced astronomical space observatories lag what is currently available for military and intelligence satellites. The Hubble Space Telescope for example was derived from early technology surveillance satellites.
Then since the James Webb Space Telescope has a segmented 6.5 diameter mirror, very likely this at least is available now for surveillance satellites.
I discussed the capabilities for such a mirror for space-borne imaging in the post to sci.astro copied below. At 300 km altitude it would have better than 3 cm resolution, about an inch. Spy satellites frequently have elliptical orbits that can bring their altitude to half this at closest approach, so their max resolution will be perhaps half this, 1.5 cm to 1 cm.
The James Webb Space Telescope is however an infrared telescope. The question I had was whether the mirror smoothness tolerances required at visible wavelengths were available using the beryllium material used in the segmented mirror of the James Webb.

From this web page we may conclude that this is indeed possible:

ESO Press Photos 34a-b/97
12 December 1997.
First M2-Unit and Beryllium Mirror Delivered to ESO.
http://www.eso.org/outreach/press-re...hot-34-97.html

The ESO's Very Large Telescope (VLT) uses 1.2 meter beryllium mirrors for its secondaries. This requires visible wavelength smoothness since the VLT will operate at both visible and infrared wavelengths. The James Webb hexagonal mirrors are 1.3 meters in diameter. So we may conclude beryllium mirrors of this size could be polished to the smoothness required for visible light observations.

This question was raised by me in regards to astronomical planetary imaging: how soon could this be adapted to space missions to the planets? The James Webb telescope is a 4 billion dollar mission. However, a large part of this cost probably has to do with the fact of the high reliability required for this mission that has to operate far away from the Earth and therefore can not be serviced by human missions, and because of the fact the entire spacecraft's structure has to be optimized to keep the cryogenic temperatures required for the highly sensitive infrared observations.
Reductions in cost for similar sized planetary missions can be fueled by commercial imaging interests. It is clear there there would be commercial uses for Earth imaging at 1 inch resolution, though this would raise clear privacy concerns. The technology for producing such large foldable space mirrors has been patented so can now be licensed by commercial imaging concerns:

Deployable space-based telescope.
Abstract
A large aperture light-weight space borne telescope is provided which may be launched by a relatively small launch vehicle. A 6 to 8 meter primary telescope composed of, e.g., 30 segments arranged in two concentric rings is provided. Supplemental outer mirror segments are stowed behind and substantially perpendicular to the main mirror which is usable in the absence of supplemental mirror deployment. Deployment of outer mirrors segments provides a large aperture telescope with a large field of view. Other deployable components include a secondary mirror, a bus, deployable with respect to the optics portion, and one or more sun shade sheets or panels.
Patent number: 5898529
Filing date: Jun 20, 1997
Issue date: Apr 27, 1999
Inventors: Wallace W. Meyer, Robert A. Woodruff
Assignee: Ball Aerospace & Technologies, Inc.
http://www.google.com/patents?vid=USPAT5898529

Bob Clark

************************************************** ******
Newsgroups: sci.astro, sci.physics, sci.geo.geology, alt.sci.planetary,
sci.astro.amateur
From: "Robert Clark" <rgregorycl>
Date: 10 Jan 2007 09:12:09 -0800
Local: Wed, Jan 10 2007 1:12 pm
Subject: We will soon be able to resolve Mars microbes from orbit. ;-)

On another space oriented forum I noted:

"It took 20 years to increase the resolution by a factor or 10 over
Viking with the Mars Global Surveyor mission. But only 10 years to
increase the resolution over that of MGS by a factor of 10 with Mars
Reconnassance Orbiter.
Could we increase the resolution over MRO by another factor of 10 to,
gulp, 3 cm per pixel in only 5 years this time?"


Funny though, that rather off-the-cuff estimate of mine is close to
what is possible.
To resolve 3 cm in the optical from say a 300 km orbit would require a
6 meter mirror. The James Webb Space Telescope will have a 6.5 meter
mirror and is scheduled for launch in 2013. But it was originally
scheduled for launch in 2011:

James Webb Space Telescope.
http://en.wikipedia.org/wiki/James_Webb_Space_Telescope

So going by this rate, it'll be 3mm/pixel 2.5 years after that, and
300 microns 1.25 years after that, and ...
Hmm, in less than a decade then we should be able to resolve microbes
from space.

Admittedly though, the JWST is a 4 billion dollar mission. Also it
uses a beryllium metal mirror for infrared astronomy only. The
beryllium makes the mirror lightweight but it is unclear if you can
achieve the much more stringent smoothness requirements at optical
wavelengths with a metal mirror.
As for the data storage and transmission of the large files for such
high resolution images, data storage capacity and costs are doubling
and halving each year, respectively:

Bye-bye hard drive, hello flash.
By Michael Kanellos
Staff Writer, CNET News.com
Published: January 4, 2006, 10:00 AM PST
"Currently, NAND chips double in memory density every year. The
cutting-edge 4-gigabit chips of 2005, for example, will soon be
dethroned by 8-gigabit chips. (Memory chips are measured in gigabits,
or Gb, but consumer electronics manufacturers talk about how many
gigabytes, or GB, are in their products. Eight gigabits make a
gigabyte, so one 8Gb chip is the equivalent of 1GB.)
"Another driving factor in the uptake of the technology is cost: NAND
drops in price about 35 to 45 percent a year, due in part--again--to
Moore's Law and in part to the fact that many companies are bringing on
new factories."
http://news.com.com/Bye-bye+hard+dri...3-6005849.html

MRO uses the type of flash memory chips discussed here.

Also, interestingly NASA had planned a laser communication orbiter for
Mars for launch in 2010 before it was canceled:

Record Set for Space Laser Communication.
By Ker Than
Staff Writer
posted: 05 January 2006
02:11 pm ET
http://www.space.com/missionlaunches...aser_comm.html

Mars Telecommunications Orbiter: Interplanetary Broadband.
By Bill Christensen
posted: 05 May 2005
06:41 am ET
http://www.space.com/businesstechnol...om_050505.html

This would have allowed data transmission rates of a hundred times
greater than what is currently available.

It was the great cost overruns overruns that led to cancelling of the
Mars Telecommunications Orbiter, and great cost overruns also
threatened JWST as well.
That the costs for computer technology are dropping exponentially with
capacity increasing exponentially is no doubt fueled by the free market
in this sphere.
Conversely, that launch costs are staying static is no doubt because
the launches are controlled by large governments. When private
companies become the primary financer and purveyor of launches, the
launch costs will also drop dramatically.


Bob Clark

************************************************** ******

[mugaliens]

Ok... So optical technology might have progressed.

Yes, if true, it's amazing, but only compared to 50 years ago. Compared to 20 years ago, it's merely yet another incremental step.

[NEOWatcher]

Ok... So optical technology might have progressed.

Yes, if true, it's amazing, but only compared to 50 years ago. Compared to 20 years ago, it's merely yet another incremental step.

I agree...besides, even if we have micron resolution, how large of an area can we look at or where can we look?

RG, Why the post... are you worried? Impressed? Confused?

If I were to start being concerned about spying, it would be cameras in dressing rooms first...

[Van Rijn]

Any super spy space telescope would also need to deal with atmospheric turbulence. Sometimes, planes are the better and cheaper option.

[RGClark]

I agree...besides, even if we have micron resolution, how large of an area can we look at or where can we look?

RG, Why the post... are you worried? Impressed? Confused?

If I were to start being concerned about spying, it would be cameras in dressing rooms first...

Like most readers of this group I'm interested in achieving the highest possibile resolution for orbital planetary imaging. That segmented beryllium mirrors can achieve resolution down to an inch and better would indeed be great news. The problem is the cost. It's not likely were going to see a 4 billion dollar mission for planetary imaging.
That's why I was suggesting that commercial applications of this technology might bring down the costs to practical levels. But then you have the problem of the privacy concerns this might cause. I'm arguing however that the military and intelligence agencies already have this capability.



Bob Clark

[Cugel]

I don't think the military have developed such a device because it would be rather useless. What should they be doing with it? Tracking / locating small objects or people? That would be silly. Any object of such small size can easily be hidden (just put it in your pocket). Besides, the telescope would still be in orbit so you can observe any location for just a few minutes out of every 2 hours. And you can't look inside buildings. I just don't see a military application for this device.

[Doodler]

I don't think the military have developed such a device because it would be rather useless. What should they be doing with it? Tracking / locating small objects or people? That would be silly. Any object of such small size can easily be hidden (just put it in your pocket). Besides, the telescope would still be in orbit so you can observe any location for just a few minutes out of every 2 hours. And you can't look inside buildings. I just don't see a military application for this device.

At that resolution or smaller, its less about trying to read over someone's shoulder than simply knowing more certainly that it is what you're looking for. At centimeter resolutions, faces start becoming identifiable. Distinctive featers stand out. Glasses, a specific suits. If you're looking for a specific vehicle, you can identify it by specific traits, such as body damage, or rust patterns.

Consider this. You have a satellite orbiting with its primary observation footprint over a harbor. The satellite passes over every 8 hours. Over the course of several weeks of observations with sequential imagining, you establish a pattern of movements, enough to know what ships are in the harbor on any given day, if you're lucky, you catch them heading out or heading in, and you can begin to construct a timeline of movements from a series of snapshots. Regular patrols, deliveries, even what ships are actually moving, if you have other sources of intelligence to coordinate with. If you ever do want to put someone on the ground in a potentially hot spot, it would be nice to know who's going to be where.

These images aren't like a microscope, they are just EXTREMELY high resolution large footprint images coming back.

[NEOWatcher]

Consider this. You have a satellite orbiting with its primary observation footprint over a harbor. The satellite passes over every 8 hours.

What about a spy plane or drone that passes over every 8 hours... spots something... and goes back for another look? What's the difference?

[Cugel]

Consider this. You have a satellite orbiting with its primary observation footprint over a harbor. The satellite passes over every 8 hours. Over the course of several weeks of observations with sequential imagining, you establish a pattern of movements, enough to know what ships are in the harbor on any given day, if you're lucky, you catch them heading out or heading in, and you can begin to construct a timeline of movements from a series of snapshots. Regular patrols, deliveries, even what ships are actually moving, if you have other sources of intelligence to coordinate with. If you ever do want to put someone on the ground in a potentially hot spot, it would be nice to know who's going to be where.

These images aren't like a microscope, they are just EXTREMELY high resolution large footprint images coming back.

But you don't need centimeter resolution to identify ships.
The military is most likely spending its hard earned cash on more productive technologies such as electronics/wireless eaves dropping.

[Warren Platts]

What about a spy plane or drone that passes over every 8 hours... spots something... and goes back for another look? What's the difference?

Spy planes get shot down.

[NEOWatcher]

Spy planes get shot down.

But are still more cost effective and flexible than a satellite, especially when you consider drones. Who cares if a drone gets shot down once the picture is transmitted?

[ToSeek]

I get the impression that the engineers on my project would be extremely happy to know that someone else has pioneered deployable optical antennas in space prior to JWST. Currently that's being treated as one of the scariest issues for this mission.

[RGClark]

But are still more cost effective and flexible than a satellite, especially when you consider drones. Who cares if a drone gets shot down once the picture is transmitted?

You could make that argument about any imaging satellite. But we know that the military has sent up billion dollar satellites with Hubble class mirrors for imaging purposes.


Bob Clark

[NEOWatcher]

You could make that argument about any imaging satellite. But we know that the military has sent up billion dollar satellites with Hubble class mirrors for imaging purposes.

Yes, you can. But main point I'm trying to make, is that there are so many other technologies that can spy on us, that worrying about the resolution of a satellite is of little importance.

ToSeek has it right...if the military can do it, then it is a good sign that researchers can follow the same path and use the same technologies without worrying (too much) about the decision.

[RGClark]

I get the impression that the engineers on my project would be extremely happy to know that someone else has pioneered deployable optical antennas in space prior to JWST. Currently that's being treated as one of the scariest issues for this mission.

My guess is the same people who designed the segmented deployable mirrors for JWST also designed them for military and intelligence satellites.
As an example adaptive optics had been used by the military to remove atmospheric distortion. When this was declassified the scientists who developed it for the military could write about it and discuss its applications for astronomy.


Bob Clark

[sarongsong]

...I'm arguing however that the military and intelligence agencies already have this capability...

No telling what they have by now, considering these black budget figures:

May 22, 2006
..."In real (inflation-adjusted) terms the $30.1 billion FY 2007 request includes more classified acquisition funding than any other defense budget since FY 1988...when DoD received $19.7 billion ($29.4 billion in FY 2007 dollars) for these programs," wrote author Steven Kosiak. Secrecy News

but without any hard evidence, isn't this military hypothesis a candidate for the CT Forum?

[ToSeek]

My guess is the same people who designed the segmented deployable mirrors for JWST also designed them for military and intelligence satellites.
As an example adaptive optics had been used by the military to remove atmospheric distortion. When this was declassified the scientists who developed it for the military could write about it and discuss its applications for astronomy.


Bob Clark

Yes, but we knew that adaptive optics came from the military (just as we knew that Hubble was based on a spy satellite). By all accounts I've read, the JWST deployable mirrors are bleeding-edge technology - there's not a hint of "Oh, yes, we've done that for the military, we can adapt it for astronomy."

[Glom]

How big do surveillence satellites tend to be? How big are they likely to get?

Using the Rayleigh criterion, we could figure out the absolute limit to their resolution.

[djellison]

No discussion of the fact that at orbital velocity, and 3cm resolution - your pixel dwell time is 1/247500 of a second. You've either got to have quite extraordinary tracking, or downtrack TDI into the thousands of pixels. It's very very easy to do the maths and come out with a figure of a theoretically max resolution. It's a totally different story to explain how they achieve that resolution with the rest of the system.

Doug

[springa]

Huh, I thought that they were actually more powerful than that now. 20 years ago people were saying that reconnaisance satellites could read license plates, or even the print on newspapers. I assume, then, that these claims were wildly exaggerated!

[trinitree88]

Planes could read the numbers on a golf ball...Spaulding Dot 4 ..from ~ 72,000 feet in the late sixties. pete.

[cjl]

Not quite...

The main cameras in the SR-71 had roughly 4-6 inch resolution from 80000 feet.

[Larry Jacks]

Huh, I thought that they were actually more powerful than that now. 20 years ago people were saying that reconnaisance satellites could read license plates, or even the print on newspapers. I assume, then, that these claims were wildly exaggerated!

They were. And they are.

[djellison]

Of course - the entire license plate issue is a bit crazy when you consider that actually - you would see it 'edge on' from above - unless you were significantly off-Nadir.

Doug

[Larry Jacks]

Of course - the entire license plate issue is a bit crazy when you consider that actually - you would see it 'edge on' from above - unless you were significantly off-Nadir

Even if you looked at a license plate obliquely, you'd need a resolution of better than 1 cm to read it. Suppose you had a satellite with a resolution equal to the width of the letters and numbers on a plate, say something on the order of 4 cm. An individual character would look blob a single pixel wide and perhaps 2 pixels high. You couldn't read the character.

Super high resolution has some uses but fewer than you might think. At the same time, higher resolution greatly increases the data processing, storage, and transmission requirements. Suppose you had a hypothetical system capable of 3 cm resolution. You want to see a useably wide area (hard to look at things through a soda straw), say 3 KM wide. A single swath 3 KM by 3 KM at 9 cm resolution would require 270,000,000,000 bytes to capture the area using 256 level gray-scale (1 byte per pixel). You'd need to multiply that by a fair amount to handle color (a factor of 3 or 4 depending on color depth). Even with compression, that's a lot of data to process, store, and/or transmit.

[joema]

...since the James Webb Space Telescope has a segmented 6.5 diameter mirror, very likely this at least is available now for surveillance satellites...

Anything is possible, but in general I would doubt this, for several reasons:

(1) Every U.S. reconnaissance satellite is scrutinized by amateur satellite trackers. The orbits are computed and they are photographed in orbit. A new huge satellite with a deployed diameter 3x or 4x the size of current Keyhole satellites would likely be seen.

http://www.wired.com/wired/archive/14.02/spy.html

(2) The successors to optical KH-11/12 satellites are reportedly physically smaller. The primary goals are to increase stealth and maneuverability, not increase optical resolution with a gigantic mirror:
http://www.spaceref.com/news/viewnews.html?id=1122
http://cartome.org/nro-fia.htm

Even the older KH-7 reportedly once achieved a 75 mile (120 km) perigee: http://www.afa.org/magazine/June2003/0603kh7.asp

Given similar maneuverability, an optical spy satellite with the same mirror diameter as the current KH-11/12 (2.4 meters) could resolve nearly 1 inch at 500 nanometers, assuming perfect adaptive optics. That comes from Dawes' Limit: http://en.wikipedia.org/wiki/Dawes_limit

Angular resolution formula:

a = 250000 x W / d, where:
a = angular resolution in arc seconds
W = wavelength in meters
d = telescope diameter in meters

E.g.
a = 250000 x 500E-9 / 2.4 (HST and KH-11 mirror size)
a = .05208 arc seconds

Linear resolution formula:

s = tan (a) x d, where:
s = linear resolution in units determined by d
a = angular resolution in degrees
d = distence to object

E.g.
s = tan (1/(3600/.05)) x 100 x 5280 x 12
s = 1.5 inches resolution at 100 miles

The bottom line is you don't need a gigantic folding optical wavelength mirror (with attendant cost and technical risk) to resolve around one inch -- you just need to get closer. The satellite maneuvering technology for that has already been demonstrated.

[RGClark]

Anything is possible, but in general I would doubt this, for several reasons:

(1) Every U.S. reconnaissance satellite is scrutinized by amateur satellite trackers. The orbits are computed and they are photographed in orbit. A new huge satellite with a deployed diameter 3x or 4x the size of current Keyhole satellites would likely be seen.

http://www.wired.com/wired/archive/14.02/spy.html

(2) The successors to optical KH-11/12 satellites are reportedly physically smaller. The primary goals are to increase stealth and maneuverability, not increase optical resolution with a gigantic mirror:
http://www.spaceref.com/news/viewnews.html?id=1122
http://cartome.org/nro-fia.htm

Even the older KH-7 reportedly once achieved a 75 mile (120 km) perigee: http://www.afa.org/magazine/June2003/0603kh7.asp

Given similar maneuverability, an optical spy satellite with the same mirror diameter as the current KH-11/12 (2.4 meters) could resolve nearly 1 inch at 500 nanometers, assuming perfect adaptive optics. That comes from Dawes' Limit: http://en.wikipedia.org/wiki/Dawes_limit

Angular resolution formula:

a = 250000 x W / d, where:
a = angular resolution in arc seconds
W = wavelength in meters
d = telescope diameter in meters

E.g.
a = 250000 x 500E-9 / 2.4 (HST and KH-11 mirror size)
a = .05208 arc seconds

Linear resolution formula:

s = tan (a) x d, where:
s = linear resolution in units determined by d
a = angular resolution in degrees
d = distence to object

E.g.
s = tan (1/(3600/.05)) x 100 x 5280 x 12
s = 1.5 inches resolution at 100 miles

The bottom line is you don't need a gigantic folding optical wavelength mirror (with attendant cost and technical risk) to resolve around one inch -- you just need to get closer. The satellite maneuvering technology for that has already been demonstrated.

You made some good points, but of course the larger mirror would resolve even better if it were at the same close up altitude as the older Keyhole satellites.
Also, I have been able to find some cases where expected Keyhole satellites were significantly brighter than the usual Keyhole satellite when they were observed by amateurs:

Titan IV object observed.
"From: Ted Molczan <molczan@fox.nstn.ca >
To: "'SeeSat-L'" <SeeSat-L@cds.plasma.mpe-garching.mpg.de >
Subject: Object observed from recent Titan 4 launch
Date: Mon, 30 Dec 1996 02:07:07 -0500

Over the past few days, Anthony Beresford has provided me with several
observation reports of an object that I believe originated with the
Titan 4 launched from VAFB on 20 Dec 1996 at 18:04 UTC.
The object is precisely in the standard KeyHole western orbital plane,
but its orbital dimensions are not standard, and it appears to be much
brighter than a KeyHole (approx std mag = 3, 1000 km range, 50 percent
illuminated)..."
http://www.fas.org/spp/military/prog.../at_961230.htm

There are also some "stealth" satellites codenamed Misty that are designed to be difficult to observe or track visually or by radar. These satellites might not be seen by amateur observers.


Bob Clark

[joema]

...I have been able to find some cases where expected Keyhole satellites were significantly brighter than the usual Keyhole satellite when they were observed by amateurs...There are also some "stealth" satellites codenamed Misty that are designed to be difficult to observe or track visually or by radar. These satellites might not be seen by amateur observers...

Anybody who can afford an amateur 8" Schmidt-Cassegrain telescope with computerized guiding can download satellite tracking software and image LEO satellites as extended objects with fairly high resolutions.

This includes spy satellites. The image quality is surprisingly good. It's not a blob of light -- outline and structures are easily visible. Attached below are examples of ISS and shuttle imaged from earth by amateurs using a small telescope.

http://www.tsm.toyama.toyama.jp/cura...20010601-e.htm

Video of ISS passing overhead taken from earth (910k .mpeg). Note: -- this is NOT animation, but actual video captured from earth: http://www.tracking-station.de/image...07-02-16-s.m1v

The station is big, but it's about 230 statute miles up. A Keyhole-size satellite at 1/2 the altitude could be imaged with about the same resolution as the above ISS images.

So getting closer works both ways -- the satellites can see earth better, but telescopic observers on earth can also see the satellites better.

If a spy satellite three times the diameter of Hubble came overhead at 120 miles altitude, it would be imaged at fairly high resolution by amateur telescope observers.

Despite the talk about "Misty", there's no such thing as optical stealth for an extended, detailed object in bright sunlight. Radar stealth is possible, but amateurs don't track satellites with radar but visually.

[mugaliens]

That's way cool, Joema!

[Kullat Nunu]

How do they cancel atmospheric blurring? Extremely short exposure times?

So getting closer works both ways -- the satellites can see earth better, but telescopic observers on earth can also see the satellites better.

Future sky surveys like Pan-STARRS and LSST are a bit problematic for spy satellites since they can easily spot every one of them.