Hearing: Problems With U.S. Weather Satellites

1. Chairman Bridenstine proposes a simple solution; eliminate government weather satellites and simply procure the data from commercial providers. The only question I would ask is, if the commercial providers can already provide all the needed data, why is there a “gap”?

   1. roger than, Dan-O. The Chairman has been worshipping again at the Church of the Free Market where free enterprise solves all problems.

      And it does! Just ask the DoD about Blackstone. Pretty darn cost effective 🙂

      Sheesh. In the same way that people born in Indonesia will think that Islam is the True Religion, and those born in Georgia will know that Christianity is the True Religion, we now live in a country where folks think free enterprise is the cat’s meow. As I’ve said before: if you were born in 1400 you’d think the divine right of kings provided a natural and divine order to human society (not my idea, originally by Ursula LeGuin).

      Think out of the box.

      1. “Think out of the box”
         If the govt offered an X-prize of $50,000,000 to the first commercial venture to provide the same data as one of these “flagship” weather missions, they would have their data before JPSS-2 launches.

         1. 1) So let’s assume a company that wanted to win the prize contracted with Space X to launch the satellite they end up building to provide the data. But Space X charges somewhere around $50 million to launch satellites in orbit. Uh, suddenly that $50 million doesn’t sound like much of a prize, does it?

            2) Businesses often make their money off of scale. So a company like Ford (or GM, or Apple) will spend hundreds of millions to develop a product, and then they’ll sell millions of the finished product (at a few hundred dollars for an Apple phone or 20-30K dollars for a Ford or GM car) to recoup the investment and to generate a profit. Satellite instruments also cost hundreds of millions to develop, and yet only 2 (or 3 or 4) of them end up being made. So they end up costing hundreds of millions to buy.

            3) Initial investment on developing the instruments is only part of the cost. It also costs money (a lot of money) to obtain, process, analyze, and deliver the data for years after the instruments are in orbit. The tens of billions of dollars price tag for these programs factor in these “life cycle” costs. Those costs do not go away; commercial companies would have to perform (and charge for) it. And the people who do that work are not cheap.

            4) I’m not saying that what’s happening now is not rife with problems and issues and that maybe there’s a better, more cost effective way to do things (including, possibly, data buys from commercial companies). But blithely saying “hey, the government should offer a $50 million dollar prize and a commercial company will solve the problem” is awfully naive.

            1. I wasnt suggesting that:
               A company builds a satellite as a one-off just for this challenge OR that the spacecraft launches by iteslf on a LV OR that a ground system would be required for just the weather data. My guess is that IF there was a bucket of $$ for this type of challenge that companies like DigitalGlobe just might consider putting one or more of the types of instruments that the govt requires for this effort on their bus.
               If the govt offers a $50M prize, and it goes unclaimed, how much did that cost?
               If it DOES get claimed how much did that save?
               AND I have a lot more knowledge about JPSS, NOAA, and how to build weather satellites (or digital globe birds) than MOST people out there 🙂

            2. Weather and climate information are used by everyone and traditionally gathered at taxpayer expense and provided to the public at no cost. Because of the large volume and currency of the information charging individual customers for access would sharply reduce the number of users and thus its value to the nation. Making it a private monopoly would be a gift to an insider.

            3. Accommodation on a bus is one thing. Who pays for the instruments which get more expensive no matter how many “build to print” copies are made? Who pays for the data downlink of gigantic streams, storage, processing, dissemination, modeling, analysis/reanalysis?

               And dealing with the academic and government consumers is no picnic, either. They want what they want, they want to change their minds, and don’t care about cost or schedule. Private companies will put a price on that. “It costs $10 for parts, $40 for labor, and $5000 for putting up with your <bleep>”

               Clinton had a good idea with the weather sat consolidation plan. The interagency hatred and turf wars surrounding that plan were pathetic. Now, NASA and NOAA are claiming to play nice but the Suomi-NPP carbon copies are 3x the lifecycle price for less than half the capability.

               A pox on all their houses.

            4. The bloat for the cost of the JPSS instruments occurred during the NPOESS era; the JPSS program, which was created from the ashes, was not responsible for that. NPOESS was a tri-agency monster created from NOAA, NASA, and the Air Force, and it was the NOAA – Air Force interaction (or maybe I should say NOAA/NASA – Air Force interaction) that doomed that program. When JPSS was created, the gov’t went back to the POES/GOES paradigm where NASA acts as the procurement, integration, and launch arm for NOAA. And, although there is some friction in that arrangement (there always will be some), it has actually been working well (as it did in the POES/GOES era).

               The SNPP program was always a joint NOAA-NASA program (no Air Force) and, although it had problems, those problems stemmed from the fact that NPOESS was in charge of procuring the instruments (hence, the huge cost overruns on developing the instruments). Apart from that, SNPP always ran more smoothly than NPOESS, and JPSS is running, pretty much, as smoothly.

            5. “The bloat for the cost of the JPSS instruments occurred during the NPOESS era; the JPSS program”

               NASA claimed the development risk had been retired. They weren’t.

               “When JPSS was created, the gov’t went back to the POES/GOES paradigm
               where NASA acts as the procurement, integration, and launch arm for NOAA”

               Bingo. NASA was furious that its gravy train had been derailed and had its knives out from the get-go. After they sank the shivs, all is sweetness and light. Except that the satellites have fewer sensors than originally planned, there is still the possibility of a gap, and the costs are still eye-popping despite everything (sensors through spacecraft) being “build-to-print.”

            6. 1) The reduction in the number of sensors and the capabilities of the remaining sensors occurred because Nunn-McCurdy kicked in on the NPOESS program, long before there was a JPSS program. By the time JPSS came into being, the damage had been done.

               2) NASA was, more or less, an afterthought in the NPOESS tri-agency monster. The agency kicked in zero money to the program, and therefore wielded very little power and not much influence. The Air Force, as the procurement agency, had control of how the dough was spent. (I’d heard this bit before about NASA having the knives out. But I never saw that, and I worked for Raytheon, not NASA.)

               3) The procurement of the satellites was made under the Air Force’s SSPR paradigm. And every procurement that used SSPR was an absolute, unmitigated disaster; I can’t think of one that didn’t trigger Nunn-McCurdy. If you want to really understand how the cost of NPOESS blew up, you should start with SSPR (and, again, NPOESS was responsible for the cost of the sensors, not JPSS).

               4) The SNPP program was always under the control of NASA. When the GAO looked at that program, they said that the cost overruns were due to NASA’s partners, NOAA for the ground system and the Air Force for procurement and development of the sensors.

               5) Now NASA is responsible for procurement in the GOES-R program. So any budget grief and sensor issues on that program is on them.

            7. OK, so a company decides to create a one-off instrument, launches it on some satellite with space to accomodate it, and it works. They win the prize. Terrific.

               But the gov’t wants (actually, the gov’t needs) to ensure that it will work for a long period of time. And it needs to make sure that the instrument works and plays well with other instruments on the satellites that will host the one-off instrument’s successors.

               So, the gov’t says to the company, “Your next instrument needs to be built to the following specs to ensure that it can withstand the space environment for the 2 (or 3 or 5) year operational requirements of the mission. And you need to modify the connections to those on the designated weather satellite it will fly on. And you need to chop off X number of kilograms to be accommodated on the satellite. And you need to change the placement of the antenna because it interferes with another instrument. And…”

               And the cost grows. Now for NPOESS, the costs got completely out of hand for a lot of reasons, including but not limited to bad management.

               But, even under the best management circumstances and processes, building these things is not really that easy, and for better or worse it’s not cheap.

            8. So true. Weather satellites are such an obvious place where sharing resources really work. Certainly it’s true that the sharing mechanism (i.e., NOAA et al) is a bit moribund, mostly due to interagency rivalry if I read the tea leaves correctly.

               And more’s the pity, because these snafus give righteous ammunition to those from the CFE (that would be the Church of Free Enterprise). And their criticisms would be correct, although the offered solutions aren’t.

            9. Sharing the information would have been fine. The problems came with sharing both cost and management authority between NASA/NOAA on the one hand and DOD on the other, organizations with very different goals, strategies, and cultures. They ended up with a joint program office, a separate NASA program office running NPP, a high-level management oversight committee, and even a program executive to facilitate communication with the management committee. They should have taken one look at the Shuttle program and not made the same mistake.

         2. $50m? JPSS is an $11 billion program (plus $4b sunk costs.) The budget is $800m per year.

   2. Its called the FAR. Companies like Boeing can turn out earth observation satellites without any problems. I would venture to say other companies could as well.

      The problem comes in when you have to maneuver through the red tape of bureaucracy. The FAR causes delays like you see in SLS and Orion. Wx satellites are not totally new technology… and the FAR is not new either.

      The gap exists because the government needs a crisis like poor weather forecasting to justify its purpose. I know companies who would love to do this kind of work and even develop new technologies but look at the regulations and simply walk away.

      1. deleted

      2. FARs are fine for buying standard products like bananas or bags of cement. It’s true they make major procurements expensive and tedious. Congress appears unwilling to improve the process. So why not switch to an SAA? Sometimes the best solution to a program that is slipping is to improve its management and provide realistic funding. And the gap may not even occur since it assumes the existing satellite will fail immediately at the end of its predicted life.

         As to privatization, there are relatively few regulations restricting private investment in Earth sensing, but I doubt any private investor will step up to the plate without a guaranteed profit from the taxpayers.

2. the gap is caused by the way the government manipulates the market place. Until the governments of the world stop giving weather data away to the nightly news no one will buy it. Someone needs to figure a way for the governments to purchase data and not weather space systems.

   1. Best response in the entire thread but your response doesn’t fit the narrative.

3. I’ve been preaching about the pending WeatherSat and EarthResourcesSat Gaps for years from my soapbox on the street corner in Wyoming . I started enjoying and buying LANDAT imagery shortly after LANDSAT-1 was launched in 1972 and imagery was made available at cost from EROS Data Center in Sioux Falls SD thru the USGS. 9 x 9 inch film negatives were about $ 10.00 and 9 x 9 color positives were maybe $ 50.00 during the Carter administration. Then along came Reagan and suddenly LANDSAT products cost a petty King’s fortune…B/W negs for $ 350 and film positives went to $ 1200 or something. The data was privatized under the EOSAT corprate consortium program, controlled by Hiughes and RCA for profit. I always felt that arrangement and cost spike was wholly unjustified. I felt it wasn’t done so much to make money from a critical government service while giving a corporation a sweet inside deal , but rather to price the stuff beyond the reach of users who maybe were not ” entitled” to satellite imagery or had ” motives” for acquiring the stuff. ( E.G. – Remember who was Secretary of the Interior then : James Watt . The Reagan admin sure had some political apparatchiks with agendas in key positions ) . Thankfully , the EOSAT consortium was not renewed. LANDSAT was financially decoupled and the archives went public and were once again available at lowest reasonable cost in late 1992 just before the election. These days you get the files by downloading them. The older stuff is free. And still very very useful.

   In the intervening dark years 1985-1994ish Congress failed to appropriate the funds for replacement satellites and data center buildout. Failing satellites were deferred. The whole of the remote sensing space science effort was falling behind, collalteral damage in the GHW Bush recessionary cllmate and some surging anti-science and anti-environment rhetoric in Congress. Bottom Line: The LANDSAT program almost crashed under quasi-private oversight and Congressional fiddling. How foolish.

   Then came the NASA/NOAA weather sat debacle. The Weather Service was forced to get Atlantic hurricane imagery from Europe’s MeteoSat. Our followon satellite programs went FUBAR.

   From that experience I have to say hear and now that funding and operating a robust redundant fleet of weather satellites and earth resources satellites should be among Congress’ highest priorities, even mandated. These things did not exist when the US Constitution was written . I’m sure the Founding Fathers would’ve liked to see the federal government provided essential information in the public interest just as it decreed the government needs to assure first class mail to all citizens was the law of the land ( the centralized universal Postal Service was the big government tech inititiative of 1791, Pre-Steampunk ).

   Frankly , only governments on behalf of the people can afford to develop launch and operate these sorts of satellites. There is small potentiality for profitability. ( See EOSAT above) . Besides, by every argument vector you eventually come to the conclusion that real time weather observation and earth resources data is an Essential Public Service.

   Americans should be genuinely concerned these services are foundering. Contact you Congress rep about the issue. Mine hear from me a lot. They’re conservative Republicans from Wyoming , so I have to make my point twice as hard twice as often to get it thru their technophobic and Science Adverse craniums.. It’s that important.

   Try ti imagine your life today without those fleets of weather satellites, GPS, telecom sats – let alone the science sats that have such huge roles in people’s lives 24/7/365. IF they all went dark tomorrow we’d be heading back to the Dark Ages ourselves, in a matter of hours.

   1. Just wanted to note that Landsat 6 was funded and built in the “dark years” and launched in 1993. It was destroyed in a launch failure.

4. I think that will turn out to be an inefficiency issue. The problem is known by NOAA and its related government departments. However, solving that problem has run onto a bureaucratic sand-bank. It is possible that they’re still having meetings about who will comprise the committee that will decide the shape of the meeting table for the meeting of the working group who will determine the composition of the committee that will decide if the Space Shuttle can be used to repair the satellites (Yes, I know they can’t but no meeting = ‘not decided’ in bureaucratic circles) or whether new ones are needed.

   Like a lot of government agencies, it will likely turn out that NOAA’s processes are so sclerotic that they simply cannot move fast enough to get replacement satellites planned, authorised, designed and launched in a reasonable time-scale.

5. Here’s an interesting paper from MIT on why bringing DOD, NOAA and NASA together for the NPOESS program ended up adding much more cost in organizational complexity than it saved in reduction in hardware.
   It’s called “The Cost Impacts of Jointness: Insights from the NPOESS
   Program” http://systemarchitect.mit….
   One lesson: “Organizational complexity hindered the program’s decision-making process.”

   1. The software industry understands this very well, as recent books have shown. Throwing more programmers at a problem doesn’t always help.

6. It will be a while before they can design and build a satellite with carbon filters on the lens so they won’t be tempted to use the data for climate research.

Comments are closed.