NGA assembles jigsaw puzzle of diverse commercial, government geospatial data sets


Chirag Parikh
Deputy Director for Counterproliferation
National Geospatial Intelligence Agency

Chirag Parikh joined NGA in 2016, after serving as the director of space policy at the National Security Council where he briefed President Barack Obama and emphasized growing threats to U.S. space capabilities. Many national security space experts say those conversations led to an important government-wide review and increased funding for a classified space protection campaign. Parikh spoke with SpaceNews correspondent Debra Werner April 18 at the Space Symposium in Colorado Springs.

What is your role at NGA?

I am responsible for missiles, from short-range to intercontinental ballistic missiles. I’m responsible for space and counterspace developments as well. I have the best portfolio in the entire building: anything from the Syrian chemical weapons attack to North Korean nuclear issues to Chinese antisatellite tests. It’s always interesting.

What are NGA’s space and counterspace roles?

We provide the indications and warnings, the analysis to be able to say what’s out there. We are obviously worried about North Korea, where they are moving missiles. We are worried about potential antisatellite developments. We are trying to effectively link air, land, sea and space. The geospatial context on the ground may give you some type of indications and warning of something happening in space. We look at patterns of life behavior. We try to predict how, when and where things will happen. When we are able to make that kind of predictive analytics, we can work with [the National Space Defense Center] and the combatant commands to train their intelligence cadre and their operators. We can say, “If the missile is erect, you may need to worry about it. If it’s stored away, you may not.” We are informing them now because they will not have time to call the East Coast in the middle of a war to ask for analysis. We have to have that ready for them before that day, if and when a war extends to space.

I didn’t realize that was your mission.

In the 2010 national space policy, we did get some change of context. When I was at the National Security Council, I kept making the argument that you have to consider what is happening on the ground. In air warfare, we provide geospatial context and awareness. Why wouldn’t we do the same thing for space? We do. But space is more nascent at this point.

Are antisatellite [ASAT] weapons largely in space?

There are a number of antisatellite weapons. They include direct ascent ASAT missiles. That’s what the Chinese launched in 2007 and again in 2013. You also have jammers and high energy lasers that can blind optics or even cause structural damage. You have cyber attacks. You have attacks on ground sites. You have orbital antisatellite weapons that could host a variety of different things: missiles, lasers and jammers. The vast majority of these things are on the ground. That’s where NGA has that critical role. We need to be able to provide that geospatial context. Where is it? What is it doing? What is its readiness status? How many are out there? Do I need to worry about 60 of these or two? That makes a big difference in how we operate in space. Most weapons are on the ground and it’s important for us to be able to keep track of the ground to be able to protect the things in the space, like space imaging satellites, which we really need it for all the other mission areas the U.S. government has beyond just space defense. We use space for GEOINT. And we use GEOINT for space. Commercial space has a huge role to play.

Why do you want commercial images with lower resolution than government satellites, like those of Planet?
You need to understand the qualities of a system: resolution, spectral, temporal, resiliency. We fly in certain orbits but there are times we are not overhead places. Planet provides a daily scan of the Earth. That has inherent intelligence value. Are you going to be able to see something very minute? Probably not with Planet imagery. But it can tip and cue you to go to a national system with a better resolving capability. One of the biggest challenges the community has is trying to get away from the word “or” with respect to commercial and potentially international partnerships too. Capability is not just about resolution. Resiliency is important. I would much rather have Planet imagery than nothing on orbit.

But you would prefer to have Planet and DigitalGlobe imagery.

And national technical means and international and DoD and Synthetic Aperture Radar [SAR]. There are companies looking at radar in different bands of spectrum. There are companies looking at hyperspectral for commercial purposes, for mining and agriculture. From a counter-proliferation perspective, a GEOINT perspective, there is inherent value. We have all these different elements of the spectrum we can utilize. DigitalGlobe’s WorldView-3 satellite has a short-wave infrared sensor. That’s pretty unique for a commercial capability. You can see through the smoke and haze. You can see fire from a forest fire or fire from a war zone.

Or if someone struck a chemical weapons stockpile.

You have different opportunities to use the different layers of the spectrum to gain different levels of context. Sometimes it’s confirmation. Radar allows you to image, day, night, through clouds and all weather conditions. That’s got great value. We are interested to see if a commercial radar company can find a business case. Then we could become a customer just like anyone else.

Is that what you want?

We’d love that. But having my policy background, I have to say, it should be the responsibility of the government to help enable industry too. I think we made the right decision with DigitalGlobe and Ikonos and Space Imaging. We said, “We will give you guaranteed up-front money to enable this market to occur.” Look at what it’s done.

Plus, it would cost much less to support these new companies.

Right. This should be a question for the government. Should the government play a bigger role to enable a commercial SAR market? Or should we let them figure it out and then come in when their commercial marketplace is set up?

I can think of five companies working on commercial SAR.

Not all five are going to be around. The company who wins is going to be the company with the best data analytics on the ground. The company who can make sense of disparate pieces of information. Any company that says, “I want to sell you my set of pixels,” will probably not get much attention as opposed to a company that says, “I’m bringing in radar, panchromatic,  multispectral and short-wave infrared. I’m mixing it with geotagged social media accounts.” Now that’s interesting.


Now I can buy a service. Or I can say, “Layer all those images on top of each other on top of that geotagged and tell me what you see.” Or, “Just give that to me in a stack and I’ll see what’s there.” It’s almost like you are going to the buffet and you are getting the dish as opposed to the ingredients one at a time.
That’s our problem, we are resource-limited in being able to go through all this information. We need help with algorithms and automation and scripts and analytical services.

Are you buying those?

Yes. We are getting to an inflection point. A lot of these companies like Planet and BlackSky are full of IT professionals, not space people. They are developing algorithms that do change detection. Why am I going to take my analyst’s time when I can pay you to keep track of this airport and tell me how many 747s are there every six hours? I’ll do some quality checks along the way, but in the end, you are giving me a spreadsheet. I didn’t buy images. I just bought the analysis that came from geospatial.

Does NGA do that?

We are now looking to do that as more and more commercial capabilities go beyond just selling images. There are companies who want to sell the analysis. NGA is looking to be able to partner with companies to buy analysis. And it gets even stranger. It’s not just analysis, it’s also the algorithms. If you create a change detection algorithm, I could probably use that same algorithm, tweak it here and there, and ingested into my classified network. The whole paradigm is changing.

Is it changing fast?

It is. Sometimes people have a feeling of change fatigue because so much is happening so fast. But the parallel I’ve heard is where the NSA was back in the mid 90s to the early 2000’s. They had very technical, very expensive sensors to do signals intelligence. Then this thing called the internet blew up and all this data was out there. How do you get your arms around that? It became a big data problem. They spent the better part of a decade coming to grips with this gluttonous amount of data. We are probably at the beginning stage. There are commercial companies who do data analytics on the internet. Let’s buy the data analytics of GEOINT not solely the imagery.

I thought your analysis would be largely automated.

Not nearly as much as we need it to be. Developing the algorithms and the automation is very difficult. You have a lot of false positives and a lot of false negatives. We pride ourselves on being right. In the intelligence community, you are only talked about when you are wrong. We are at the nascent stage. We want to make sure we get it right. We are changing our hiring practices for instance. We aren’t just hiring GIS individuals or geography majors anymore. They are very valuable still to the organization. We are also hiring a lot more data scientist. We are creating a data corps. We are hiring computer programmers, computer scientist, analytic methodologist.

Are you saying bring on the feast and we’ll figure out how to use it all?

Absolutely. We collectively will figure out how to use it all. It’s a big jigsaw puzzle that we need to put together to bring the best of GEOINT to bear not just for space purposes but for the entirety of our mission set.