A behind-the-scenes look into the birth of a spacecraft; the research, development, fabrication, and mission objectives of NASA/JPL’s Earth science observatory, SMAP, and the challenges met in creating a new technology.
Part One: An interview with NASA/JPL’s SMAP project manager Kent Kellogg
In September 2001, a two-year study calling for the measurement of ocean salinity and soil moisture from low-Earth orbit was completed by NASA’s Jet Propulsion Laboratory (JPL) and submitted to NASA’s Earth Science Technology Office (ESTO). The title of the paper was “Spaceborne Microwave Instrument for High Resolution Sensing of the Earth’s Surface Using a Large-Aperture Mesh Antenna.”
Its focus was of “critical importance,” a mission that could accurately predict key ocean and land surface processes on a global scale accessible only from space. Using a 6-meter rotating deployable-mesh antenna with radiometer and radar sensors, the craft would “measure the microwave emissions and backscatter from the Earth’s surface.”
Carefully researched, illustrated, and thoughtfully crafted, the paper draws you into the discovery process and the long hours it takes to create a new and unique instrument. Clearly defining the urgency for the project, it lays out the benefits to be gained and what’s at stake. At the time of its release, the project was called “Osiris,” and its craft were called “Hydros and Aquarius.”
On Monday, March 9, 2015, this writer met with JPL’s Kent Kellogg, SMAP’s project manager and asked if he would take us into the journey of designing and fabricating the spacecraft that once had been called Hydros, and what it took to bring the project home.
- Q: Once the 2001 report on the urgent need for high-resolution ocean salinity and soil moisture measurements from low-Earth orbit was completed, a number of years passed, with many changes on the world stage affecting a go for the project. Can you take us back into that timeline and share some of the early developments that led first to Aquarius, and then to SMAP?
A: Hydros – a mission to measure soil moisture over Earth, which later became known as SMAP, and Aquarius were first proposed to NASA’s Earth System Pathfinder Program (ESSP) Office during the 2001-2002 timeframe. Initially, both Aquarius and Hydros were chosen to move forward, but only one was selected by NASA to continue, and Aquarius went on to launch in 2011. Hydros did receive some support for the program to continue, but with the budget constraints of 2006, the agency lacked the funding to move the soil moisture mission forward.
When both of these missions went into the 2001 – 2002 proposal process, their respective architectures were pretty much defined. Aquarius was a set, fixed non-spinning reflector (about 2.5 m in diameter). I think it was proposed to be a little bigger than they actually ended up flying. The Hydros mission proposed a 6-meter spinning antenna, which is what we actually flew.
But these are really two quite different missions. It’s not quite right to think that SMAP is an evolutionary mission that came from Aquarius. They actually have quite different radar instruments. The radar plays a much different role on Aquarius than it does on SMAP. And the radiometer, which is the sensitive receiver, has design characteristics also different than SMAP. So, although they both have a radar, radiometer and antenna, once you get below that, they really are quite different.
Now as far as the spinning architecture goes, one of the things that was of interest for SMAP was to be able to measure the entire Earth every two to three days. Soil moisture doesn’t change super rapidly, but it does change over the matter of a week or two. You can imagine a rainfall event, and then there’s a dry-down cycle, then another rainfall and dry-down cycle. So you want to sample those often enough so that you can observe that dynamic variation taking place between the forcing function (rain) and the dry-down cycle.
So the desire was to try to make a global measurement every two to three days. And that really dictates that you either have to have a constellation of satellites, which would be prohibitively expensive, or, you need a very wide measurement swath. Imagining it as a lawnmower, you want something with a very wide deck and blade path so you can get the job done with as few passes as possible.
That’s what SMAP’s spinning antenna does. It projects a beam at about a 40- degree angle off of vertical, or nadir, and then as we spin the antenna, the beam is spinning in a conical motion. Where the cone hits Earth, the diameter is a thousand kilometers. So, that’s how we’re able to get this two- to three-day coverage with a single spinning system.
We’ve used this type of scanning system on other missions, for example, QuickScat and its Seawinds instrument that measures ocean vector winds. They’re operating at a different frequency, that’s Ku-band. It uses a smaller reflector because it’s a smaller wavelength, but it uses the same kind of approach. It’s an antenna looking off at about a 40-degree angle, and spinning so it covers the entire Earth pretty efficiently.
- Q: So, the size of SMAP’s reflector was determined by the need for having a two- to three-day coverage?
A: The size was determined by a couple of things. The radiometer is the part of the instrument that makes the high-accuracy soil moisture measurements. The brightness temperature variation at L-band from wet to dry soil is about 200 Kelvin. It’s not a physical temperature; it’s the temperature of the naturally occurring L-band emissions from the land surface. So very dry soil looks to the radiometer to be about 300 Kelvin, and wet, or very moist soil, looks like it’s about 100 Kelvin. So, by measuring the brightness temperature, if you will, of a plot of soil, you can quickly estimate what the water content is down to about four percent accuracy. So the accuracy is pretty incredible.
Now the radar is a little different. Its accuracy is moderate compared to that of the radiometer. But what the radar brings to the party is high resolution. The radiometer is limited to the beam that’s generated by the real aperture of the antenna, in this case the 20-foot-wide reflector. But with the radar, we can generate what’s called a synthetic aperture. So by making a series of measurements as the satellite is orbiting overhead, we can have in effect a virtual antenna that’s much larger than our 6-meter (20 foot) antenna. And what that allows us to do is improve our resolution of the radar measurement down to as little as 1-kilometer.
The real aperture resolution of the radiometer, which is basically the spot size it reads on Earth, is 40 kilometers. So you can see, with the radar we get far better resolution, but it’s not as accurate. What we do in our ground processing is combine the radar and the radiometer measurements, and we end up with a composite measurement that has the accuracy of the radiometer, but its resolution is 9 kilometers instead of 40 kilometers. So for our soil moisture measurements, we don’t get the spatial resolution that’s inherently available with our radar, but we get something that’s at least four times better than having the radiometer by itself.
So, the size of the antenna gives us that 40-kilometer footprint of the radiometer. That’s one element that it gives us. The other element that it gives us is a large enough physical size so that we can generate the synthetic aperture with the radar. If the antenna size gets too small, the radar properties won’t allow us to generate a synthetic aperture.
- Q: So the reason you have so much better resolution with SMAP over land than Aquarius is the size of the reflector?
A: Yes. Absolutely. So, Aquarius is effectively a 2.5-m reflector. You can scale the difference in resolution if you compare our radiometer, with 40-kilometer resolution, with the Aquarius radiometer with something like 100-kilometer resolution. So it’s a little less than half the diameter as ours, which makes sense with ours being a little over twice as large as the Aquarius reflector size.
- Q: How important is the shape of the parabola of the reflector?
A: It’s pretty important. The way it works in principle if you recall from your geometry class, the unique feature of the parabola is if you draw a line from the focus of the parabola to the parabola itself, all lines leaving the parabola, if you can imagine a line bouncing off the parabola, all lines that bounce off the parabola are parallel to each other, if you start those lines at the focus of the parabola. And the same is true with reflectors. It’s a little bit like how your flashlight works. If you imagine a parabolic mirror with a light bulb out in front at the focal point, when the light is turned on, all the light rays bounce off the parabolic reflecting mirror, and instead of scattering in all directions like they would if the light bulb was just sitting out in space by itself, all the light bounces off the parabolic reflecting mirror, producing a nice crisp, focused flashlight beam.
The reflector antenna works the same way. We put the feed at the focal point of the parabola where its radiating up into the parabola, then all the RF (radio frequency) energy leaves the reflector collimated, or in parallel, so it’s focused into a fairly tight beam.
I used to be an antenna engineer, so I love these questions.
- What about the mechanics of the antenna’s rotation? The diagrams I’ve seen of SMAP’s rotation shows what looks to be the top one third of the spacecraft, including the radiometer and the radar, spinning on the same platform.
A: So what’s happening is, there is a platform on the zenith deck of the spacecraft that the antenna boom is mounted to, and the radiometer is mounted to that as well. You want to put the radiometer as close to the feed as possible so that you minimize the amount of loss between the feed and the radiometer electronics. All that spins together as an integrated unit. The reflector, the boom that holds the reflector out in front of the feed, the feed, and then the radiometer that’s attached to it. The radar is actually mounted to one of the spacecraft panels. In fact, it takes up an entire spacecraft panel. And it lives on the panel that’s actually facing deep space. So if you imagine the solar panels are on the side of the craft facing the sun, the radar is on the very opposite side looking out into deep space. The reason we do that is because the radar generates a lot of heat, which we can then dump out into space, keeping the electronics cool.
- Q: Does the radar also use the 6-meter reflector?
A: Yes. So we route the radar signals through a rotary joint that can transmit signals across a spinning interface, and then route those up through the feed. An interesting technological point is that you have this radiometer, which is a very sensitive scientific receiver. It picks up just about everything. And that radiometer is sharing the same feed with this very noisy, high-power radar transmitter. It’s a little like putting a library being filled with academicians and Ph.D. students right next to a rock concert and expecting them both to coexist happily.
- Q: How do you quiet the radar? Do you use filters to cancel the noise? How does that work?
A: We do a number of things. There’s a device called a diplexer that basically arbitrates the use of frequencies between the radiometer and the radar. What the diplexer does is basically act as a very sophisticated filter. It offers a very high degree of filtering for the radar signals that might otherwise go into the radiometer, providing a very low loss half for the radiometer signals that pass through the diplexer and into the radiometer. So we have this very high degree of filtering.
The other thing we do is have an electrical signal that goes from the radar to the radiometer that tells the radiometer when the radar is about to transmit a pulse. With the signal from the radar telling the radiometer that its about to transmit a pulse, the radiometer will momentarily put itself into standby mode using a receive/protect switch that gets activated. So, it covers its ears on top of the filtering that’s going on.
- Q: What kind of cycle are we talking about? How often does the switching occur?
A: SMAP’s radar pulses are transmitted around once every 300 microseconds. It’s all happening extremely quickly, thousands of times a minute. But the system’s designed for that.
Aquarius has much the same kind of issue. They have a different kind of radar and a different kind of radiometer. But, they were able to demonstrate how you could get these two instruments that you would think are inherently incompatible, and be able to use them with the same antenna and put them right next to one another. So, we did leverage that experience and expertise from them.
- Q: So, the alternate switching is like two gates opening and closing opposite one another?
A: Exactly! That’s a good analogy.
- Q: How do you balance the 14.7 rpm spin of the Zenith deck carrying its 20-foot antenna, boom, and radiometer, with the lower portion, the bus of the spacecraft, which must remain stable and locked in a position relative to the Sun and deep space?
A: So, the most important thing is that you want the spinning part of things to be very well balanced. If you look at a graphic of our observatory, you’ll see that there are three boxes that are mounted on struts off the spinning deck. There’s a large coffee can-looking structure with three boxes mounted on it along with the big reflector as well. We used those three boxes to coarsely balance the structure. If you look at those boxes, you’ll see that each one is at a slightly different offset from the structural cylinder that supports it. We changed that offset so some of the struts are very long with a small box mounted at the very end. Some of the bigger boxes are mounted much closer in on a set of much shorter struts. So what we’ve been able to do is actually use those boxes to balance out the structure.
The other thing we did is to put very small masses onto the big reflector itself, so that when it’s deployed, the whole system should be in balance. But as we designed the system, we were keeping track of the estimated mass so we could know how to position these boxes on their struts. We knew how much balance mass we thought we’d need around the reflector. Then as we actually built the boxes, we could weigh them and say do we need to change how we’re balancing the system? Mass was a very important property that we tracked very carefully throughout the whole development. So, we know through the careful work we did that the finished unit, which we could not spin up in 1g on the ground, should be pretty close to a perfect balance when we do spin it up in space.
- Q: Astro Aerospace, the company that fabricated the antenna, used a “spin table” provided by JPL to assist in balancing the mass of the antenna at their facility near Santa Barbara. The process was described as being something like balancing the wheel of a car using perimeter weights. Knowing that the antenna was made to function in the microgravity of space and not in 1g, how was the spin table used in the balancing process?
A: What the spin table gives us is a more complete measurement of the mass properties. It gives us the inertia as well as the balance point. If you think about what that means, if you were to unload a large box off the back of a truck, the first thing you might do is to lift one corner of it so you can figure out how the mass inside is distributed. You wouldn’t want to just pick the whole thing up and then find that the mass is way at the far end of the box at the very top, or you’ll just dump it over, right?
That’s the inertial property of the object, to determine how the mass is distributed in space with whatever it is you’re trying to measure, a strut, or an entire reflector.
- Q: The antenna developed by JPL and Astro Aerospace for SMAP, I understand this is a first, that a spinning reflector this large has never been attempted. Do any stories or anecdotes come to mind from the development phase of the SMAP array? What was it about the antenna that was out of the ordinary?
A: The launch vehicle uncertainty was a big part of that. They had to repackage that antenna several times when we were going through the early development period. This was also the first radiometric and radar application that they had supported. Their larger antennas are for telecommunication applications. And, the telecom people have a different set of “care-abouts” for antenna performance than I think science instruments have. With science instruments, every aspect of antenna performance, any issue that you can imagine, is a concern, whereas the telecom applications may be a little less stringent. And so, as a science mission, we were pouring over all kinds of performance nuances. We were having to ask them to repackage the antenna because of the launch vehicle uncertainty.
- Q: There’s no high-definition video in orbit to look at, so how can you be certain that it does deploy properly and have the correct geometry?
A: The ultimate proof of the pudding is when we turn on the instrument and get measurements. Once we got the antenna deployed, a couple of days later, we turned on the instruments, both the radar and the radiometer. We weren’t spinning yet, but we’re looking at Earth for the first time through this reflector. The algorithms that are used in ground data processing use the modeled RF (radio frequency) performance of the deployed reflector. And so, one of the things we saw right away was that the measurement results came out pretty close to what we expected. That says that the modeled RF performance of this reflector would have to be pretty close to what the actual performance is. Ergo, the antenna must be deployed properly, the surface characteristics must be pretty close to what we want, the pointing of the reflector has to be very close to what we want. So all that performance information that we’re able to glean from the radar and the radiometer allowed us to infer with tremendous accuracy how well the antenna was deployed.
In fact, the science team was telling me shortly afterwards that “you know, we think the antenna may be pointed a tenth to two tenths of a degree down,” below where they thought it should be. Well, a tenth to two tenths a degree is generally thought of as pretty exceptional alignment with your prediction. So the results just laid exactly on top of where we wanted them to be.
One of the problems with these large structures is that you can’t test them on the ground as an antenna, as an instrument. They’re just too big; you’ve got no way to support it. So all this is done on the ground by computer analysis. You take the shape of the reflector which you’ve measured using things like photogrammetry, you know what your feed pattern looks like, and you put that into a computer model and it predicts what the reflector beam pattern will look like when you get it in space. If you haven’t modeled it properly, or you’ve measured the surface incorrectly, or you’ve made some other mistake, you won’t find that out until it’s too late for you to do anything about it. In this case, with SMAP, our performance looks nearly identical to what we’ve modeled in our computers.
- Q: With so many facets to the mission, what were some of the major challenges you faced during SMAP’s development?
A: As the project manager, one of the things I have to worry about is cost. Our customer had a fairly significant cost constraint or target that they were imposing on us early on. One of the things we struggled with was how do we make the system function as it needs to, but also make it as simple as possible for affordability.
This was a case where, not just in the spin area, but in a lot of areas, we went through a fairly significant process of looking at an early design. We did this about a year before our preliminary design review. We went through the whole observatory top to bottom that had been developed to that point and decided we really had to simplify the system that we had because at that point, it was unaffordable to our sponsor.
The review board came in when we got to our preliminary design review and noted that we’d simplified the design significantly, and in the process, made it more reliable. We’d been worried about the cost part of it, but when you take out complexity, you’re left with a more reliable system. So, that was a benefit of going through the exercise, one of our challenge points.
The reflector itself probably turned out to be a bigger development challenge than we or Astro may have appreciated going into it. Some of that had to do with being faced with not knowing which launch vehicle we would be launching on. That’s pretty important; because as you proceed with the design, you need to know how much room do I have to work with inside the fairing? What will the loads and forces be on the instrument during launch? So, when you don’t know your vehicle, you have to make very conservative, even pessimistic assumptions, which end up driving your development cost.
In Part Two of THE EYES OF SMAP, we’ll meet with Astro Aerospace as they share the challenges of creating SMAP’s 20-foot spinning reflector and boom, the unique materials they’ve developed, balancing its mass, determining its inertia, and testing the unit to deploy with perfect alignment and geometry in the extreme environment of space.
Missions » SMAP »