[time-nuts] GPS usable for weather forecasting?
jimlux at earthlink.net
Fri Mar 29 23:50:53 EDT 2013
On 3/29/13 2:36 PM, Chris Albertson wrote:
>> COSMIC and (coming soon) COSMIC-2 also do GPS occultation.
> Yes, but COSMIC is not a constellation of 12 satellites and it is not as
> cheap either. These guys want to put up 12 satellites at a total cost of
> only $160M
COSMIC-2 is a constellation of 6, with more to be added later.
occultation is pretty heavily used now.. When you talk about weather
forecasting, they talk about "what percentage of the variance is reduced
by adding source X", and I seem to recall that for GPS RO it's something
like 10-15%.. around the 3rd or 4th most useful measurement.
There is also a huge amount of RO type data collected by GPS receivers
on the ground (after all, when that GPS satellite comes over the
horizon, you're looking through the atmosphere)
If you go to PlanetIQ's website they have this statement:
"The satellites that currently collect global GPS-RO data are degrading,
while planned replacements have encountered delays and funding issues,
and are insufficient to meet the stated needs of users around the planet. "
Hmm. I sat on a review board for the GPS receivers for that "planned
replacement" a month or so ago and I don't know that this fairly
represents the situation. Everyone has funding issues and everyone has
delays, but in general, I'd say that COSMIC-2 is better off than a lot,
and at least they're building hardware as I write this, which is more
than PlanetIQ is doing.
Also, that whole "stated needs" is a pretty fuzzy thing. If you surveyed
100 atmospheric scientists and said "would you like more data" they'll
all say "you bet". If you said, "would you sell your first born child
to get that data", they'll say "hmm. maybe I can do ok with what I've
got". Obviously, reality is somewhere in between.
I guarantee that every single proposal that comes in to a funding agency
has a very convincing argument how there's millions of people who will
benefit from funding this, and if you don't fund it, puppies and kittens
will spontaneously die. I've written those pages myself.
The website says that COSMIC-2 won't be online til 2017. I note that the
COSMIC-2 website says "six satellites into low-inclination orbits in
late 2015" on a Falcon-Heavy, I think.
It would be interesting to see how PlanetIQ will get into orbit earlier.
The website doesn't say.. basically, they say "if we get enough money,
we could do this and sell the data", which is fine, but even in these
cubesat days, it's not like you can get a launch in a year. Getting
your command and telemetry frequencies coordinated takes longer than that.
Fascinating the BroadReach is involved.. after all, they built the
receivers for COSMIC, they're building the receivers for COSMIC-2, etc.
They're the dominant (if not only) builder of flight qualified
metrology quality GPS receivers, based on the JPL designs (Rogue, etc.).
The TriG flying on COSMIC-2 is a JPL design, being manufactured by BRE.
An interesting business opportunity, but a lot of paper, and not much
substance. I also didn't find where the $160M comes from. That's "sort
of" believable, if a lot of things go the right way.
Keeping this somewhat time-nutty... It would be interesting to see what
kind of measurement performance one could get with cheap GPS receivers
like they use on cubesats. The TriG is pretty high performance, but it
definitely isn't cheap. The problem is that to make decent RO, you're
looking at those sub-ns kinds of measurement precision we talk about
here. it would be really interesting to figure out a way to get a low
multipath antenna comparable to the usual choke ring, but in a form
factor compatible with a 3U or 6U nanosatellite.
It might be that if you got all the electronics in a 10cm cube, the
multipath is not an issue: there's no structure to reflect the signals.
I note also that the COSMIC-2 RO measurements are made with a phased
array of a dozen helical antennas, so they can "slice" the atmosphere to
improve the measurement. That's sort of counter to a cubesat (although
PlanetIQ doesn't say cubesat.. so maybe their satellites are bigger than
More information about the time-nuts