While this would certainly put an end to the poor siting problems discovered by the surfacestations.org project, I can’t help but think almost everything related to climate can be solved with money:
Here’s the letter:
PDF with attachments here: USHCN_Letter_-_FINAL_-_7-29-10_SECURED
I also can’t help thinking of this image when 100 MILLION DOLLARS is used:
Now don’t get me wrong, I support a modernized network, but $100 million? That’s a bit steep.
It works out to $100,000 per weather station.
When I visited NCDC in April 2008…
Day 2 at NCDC and Press Release: NOAA to modernize USHCN
…they told me the USHCN-M cost was supposed to be around $25,000 per weather station.
Which looking at the USHCN-M equipment below, allowing for government inflation, sounds about right:
USHCN-M station at Greensboro, AL
But $100K a piece for what you see above? I don’t think so.
See: What the modernized USHCN will look like
Hell, I’ll do it for 10K a piece and do a better job than NOAA ever could.
h/t to Joe D’Aleo
Very interesting discussion. My first thought would have been that this is not the right approach. We should not be building weather-measuring stations, with concrete and all that stuff. We should be designing a very inexpensive and very sturdy thermometer with GPS and automatic reporting, powered by solar power(?), that can be dropped from a helicopter or such. Instead of getting 100 stations, drop 10,000 thermometers. If they’re $100 apiece, that’s a million dollars, not counting the helicopters. When half of them fail or are eaten, drop some more. You’d lose some metadata, but get vastly better coverage.
How to take a dump for an NIH study:
http://scienceblogs.com/mikethemadbiologist/2010/09/how_to_take_a_government-appro.php?utm_source=networkbanner&utm_medium=link
Gee, two of the three experts from “the three professional organizations we contacted” said that in their personal [not official] professional opinions, NOAA/NCDC USHCN Version 2 data “has value”, one saying it’s “the best we have” – but which can also easily mean not a ringing endorsement of past U.S. surface station related “GMT” reconstructions; and possibly leaving one expert not very impressed at all; along with me, as to this particular evaluation’s contribution to establishing the credibility of all non-satellite “GMT” reconstructions. FAIL?
Hi Anthony. The $75 K you don’t account for is for (i) good old kickbacks, (ii) a really big air conditioner unit to situate nearby, and (iii) blacktop paving of the site.
old44 says:
September 22, 2010 at 4:47 am
Steven mosher says:
September 21, 2010 at 3:45 pm
Anthony the land survey expense alone is 3000 per site.
so even before you buy anything, you are down to 7K.
Dear Steve, that adds up to $10,000 per site – $100,000,000 divided by 1000 sites equals $100,000 per site. Back to grade 3 for a little remedial arithmetic
########
no old, I am afriad you need to go back to school.
The claim was made that it could be done for 10K per site.
As I pointed out, the cost of site SURVEY is 3000 per site. So anybody who thinks they can do it for 10K per site, has forgotten that they have to survey the site and see that it meets standards. That involves travel, hotel, food, car rental, etc. hence, if they thought they could do it for 10K per site, 3 K of that is eaten up by survey.
so they are left with 7K for
1. equipment
2. equipement transportation
3. labor to install.
4. spare parts
5. Taxes ( 7%)
even if you could do it you could not operate or maintain the site, collect the data, etc
watch the blue angles fly over head, look at the computer in front of you, or listen to your ipod, or talk on your cell phone, and understand that some of us here have decades of experience in designing, costing and building and supporting the things you have no idea how to build.
James Sexton says:
September 22, 2010 at 5:03 am
Steven, I believe you’re thinking within the proverbial box. The scenario you’re describing is what got it all messed up in the first place. But if we still have to go that way, sure contracts are a nice way to go. As far as the comm goes, one would have to use a variety of forms of communication. One size won’t fit all. Fortunately, the old and new system will have a standardized unit of measure called “temperature”. NOAA says the system is broke. This statement implies they’ve identified problems and problem areas. Yes, compliance is always a difficulty, auditing should not be, we’re still paying NOAA personnel. They should have a function or two in the process.
#####################
Sorry, I spent years thinking out of the box. That was kinda my title. There are effective ways of innovating and ineffective ways. The most difficult innovations are those that try to innovate AGAINST infrastructure, since infrastructure has long term payback requirements. if you like I can detail all the innovations that I have succeeded in bringing to market as well as those that failed. The failures are almost always those that try to innovate infrastructure. In anycase innovating with the government is very difficult because there values and time horizens are different.
Since there is a state (Oklahoma) that already has a statewide system (120 monitoring sites) that has been in effect since 1993, it would seem logical to use that system as a model. Initial cost was $2,700,000 for 108 stations = that’s $25,000. I haven’t found out the operating costs yet, but I assume that it is within the state university budgets.
But, if Oklahoma has this type of network for education, research and support of ag and public safety, don’t other states? How much of the work is already done and available and probably partially funded by federal grants? And, I’m sure that the data is used by the National Weather Service and the Severe Storms Lab in Norman.
http://www.mesonet.org/index.php/site/about
david says:
September 22, 2010 at 5:11 am
I thought we already had weather sites. I thought we already had people to monitor them. I thought we already had supervisors? All existing costs of current system must be subtracted from new system. All current budget must be subtracted from new systems proposed budget. Is it not new equipment in better locations within same sites, or on gov land with no purchase price?
###################################
these are new stations. in the current system the stations have assigned engineers and scientists. in the new system the requirement is 99% data availability. So, that’s new staff. The current system will stay in place for calibration against the old.
The land in cases will have to be leased. that is in the program cost. people who expect to argue with me should do themselves the favor of reading the document
“Production is four things, paint, paint brush, painter and picture. Do each of those four things correctly and no problems. If you have a problem it is in one of those four areas, every time. Do it again 1,000 times.” unfortunately in this case its wood, and wire and concrete and computers and batteries and solar panels and wind mills and different in every case. no learning curve unless you send the same crew to every site.
And with installation happening over years I suspect the learning curve will be flat.
Still, since they have experience putting in 100 or so, the learning curve ( if any) is known.
“All costs, taxes pension, medical and payroll for an hourly employee equall about a 55% increase on the hourly rate of say 30 per hour. ”
Unfortunately if you want to sell your services to the government you must use their accounting principles. its how they manage contractors and get cost parity from suppliers. Cost parity is important because they want to decide issues on risk.
two contractors offering the same thing (meets the spec.. NO INNOVATION) same cost ( everybody uses the same accounting system and shares labor rate information)
so all decisions come down to risk: can you deliver on time. So a low ball bid is suspect. same thing: same cost: win on risk abatement. simple.
“the supewrvisors and data people are already in place in the current system. ”
Acually not. You would need to read the program overview to understand the staff required to integrate into the existing system. Its like this:
everyone is already working 100 percent. They just signed a time card that says so.If you add a new program, you have to add staff. If you lose a program, you lose staff.
this is not operated like a business.
“Maintance should be minimal at sites, reporting data gathering mostly automated.”
You didnt read the links I posted. the current system is manually operated. but with 1000 stations they propose a level of automation. that system needs to be built and tested. that system needs to be manned. heck we man satillites. My friend had this great job flying a satillite ( dont ask, top secret, blue building in sunnyvale) anyways.
Maintenance is based on experience with systems already fielded. As I said above I suggest anyone who wants to can go look at the ACTUAL data coming from the system and see how often the system has issues. like batteries or sensors going bad, or fans stopping, or wind damage. Then you have calibration. we all whine here about calibration. it takes money. and time.
Liz says: September 22, 2010 at 1:05 pm
Since there is a state (Oklahoma) that already has a statewide system (120 monitoring sites) that has been in effect since 1993, it would seem logical to use that system as a model.
Most states have a network, but the international values? There are about 40 or so stations here.
Missouri Historical Agricultural Weather Database
http://aes.missouri.edu/sanborn/weather/sanborn.stm
This model is used to query the Missouri Agricultural Weather Database to obtain hourly and dailyweather data from the Commercial Agriculture Automated Weather Station Network.
Well I don’t think $100M will be anywhere near enough to cure the problem.
After all the City of Los Angeles admits to having received $110M from the Federal Government (taxpayers) “Stimulus funds”, and by their own admission they were able to create a total of 55 public employee (union) jobs with that money or $2M fpr each public employee union job; well actually those jobs are only temporary; because when the $110M is all gone thse people will once again be unemployed.
So how is a mere $100M going to support the countless thousands of otherwise unemployed “Climate Scientists” who seek to make their living off the Global Climate Disruption gravy train ?
Liz says:
September 22, 2010 at 1:05 pm
Since there is a state (Oklahoma) that already has a statewide system (120 monitoring sites) that has been in effect since 1993, it would seem logical to use that system as a model. Initial cost was $2,700,000 for 108 stations = that’s $25,000.
#######
yes in 1990 dollars.
and one cant forget the donated time and services.
of course they charge for the data.
Steven mosher says:
September 21, 2010 at 3:45 pm
no old, I am afriad you need to go back to school.
The claim was made that it could be done for 10K per site.
WHO claimed?
Please contact Phil Jones and Wei-Chyung Wang – they can make up the data for you. This will actually be better than the original data would have been (you lay people may want to call it ‘value added’ data).
If you have any questions concerning data probity or integrity – an independent inquiry can be organized for a small additional fee which will guarantee full scientific integrity.
Mosher,
About the charge for the data. Oh boy do they charge for the data. Take a look at the charges for the data:
http://www.mesonet.org/index.php/site/about/data_access_and_pricing
Everybody here would be screaming if NCDC charged this much for data. You all want the best quality data, but you don’t want to pay for it. Personally, I think the cost in my taxes of a fraction of a cent per year is well worth it. This is opposed to the OK charge of $1000 per CD for three years out of date data or $12000/yr for near real time data.
Rattus says:
“You all want the best quality data, but you don’t want to pay for it.”
Sorry, that’s backwards. We certainly pay for the best quality data. But what we get has been adjusted, tweaked, massaged, manipulated, and beaten into submission until pseudo-reality is forced to conform to climate model output.
How can you say that you need another $10 million/year for 10 years to improve something that, overall, you don’t think needs improving? If it ain’t broke, don’t fix it. If it is broke, admit it and tell us what, in it’s broken state, it is still good for and what, in it’s broken state, it is NOT good for. But that would admit that some of it is broken, which means that 95% certainty goes out the window, which is non-admissible, which means it is NOT broken, so we don’t need the money, but we WANT the money, so …
My head already hurts and I have no idea what the justification is. Except that we think it is cool to get more money.
Steven mosher says:
September 22, 2010 at 12:39 pm
…….”watch the blue angles fly over head, look at the computer in front of you, or listen to your ipod, or talk on your cell phone, and understand that some of us here have decades of experience in designing, costing and building and supporting the things you have no idea how to build.”
———————
True, yet can you concede, I paid for it. Your expertise.
Without us peons, government contracts don’t exist.
Which comes first, the chicken or the egg.
A lot of numbers have been tossed out – I presented the cost of a 1991-1993 project that installed automated weather recording/transmitting in all counties of a state. I did not restate the cost in current dollars. I’m sorry, but I assumed that most people here would have realized that inflation had occurred and would be able to mentally adjust the number using whatever factor they were comfortable with.
The OK Mesonet involves a state agency (OK Climatological Survey) and two state universities. In addition, another state agency (Public Safety) lets the data be transmitted over its telecommunication system. So, the cost is covered by the state and its taxpayers. There really isn’t any true donated services. We paid for it. I have asked for annual operating costs, we’ll see if I get it.
As to the fees for the data – the real time data is available to anyone with a computer. The archived data is sold to companies who would have a need for it. Think electric companies planning wind farms. State agencies, schools etc get the data for their needs, including research. And I guess you were shocked at the price, so you didn’t see the “fee waiver” section for smaller research projects.
But, my purpose of writing about this system is to show that there is an automated system in place that has practical uses (farming, public safety, emergency management, general weather info) as well as collecting data for climate research. It is an excellent website that includes pictures of all the sites. Proper siting of weather instruments is a concern of Mr. Watts as well as many others.
It would be wonderful if whatever is done with updating the current NOAA system that they would be just as transparent.
This is government spending, not real money, $100m would be a steal – but you need to double that at least.
Steve,
Your experience with government contracts and wrap-rates and the like notwithstanding, I think you are missing Anthony’s point, just as Pachuri et al miss the point on CAGW. Do all the math and simulations and column adding you want, but the end result is not reasonable.
This “fix” for the climate data is not a fresh-start program but an add-on. Incremental costs are quite differnt. Sites, equipment can be improved. Current staff are in place, along with data handling facilities.
This full-cost accouting is the way they do it and insist on doing it, perhaps, but is the end result reasonble? I’ve worked for multinational organizations, and I’ve seen this many times. You don’t want to do it but don’t want to say you don’t want to do it? Write it up with a “precautionary principle” as a guide. $100 million and you can’t say it won’t be more? This brick isn’t going to fly.
Definitely a procedure that kills action while stimulating activity. If nobody moves, nobody gets hurt. Looks good up the ladder, though. Lots of meetings. Bosses get to say they have addressed the issue, but right now, well, there is a problem.
I actually don’t think those prices are too outrageous. It’s a little simplistic to just divide the total cost by the number of sites and call it a day. Since this is over a ten year period, you’re ignoring maintenance, calibration (which should be done annually if not quarterly), data gathering costs, network costs, server and hardware costs, etc etc. They’ll get hit by lightening, run into by lawn mowers, hit by tornadoes and hurricanes, and damaged by hail and high winds.
An AWOS-3PT, which has no redundancy and would rely on having power brought to the install location, can run $100k-$125k just for the install.
Of course, the AWOS would have a nice tall tower for more accurate temp and wind data and has much more stringent citing requirements (build a parking lot next to your AWOS and have fun explaining to your pilots why it’s been decommissioned by the FAA). I’m surprised at those pics. Placing a temp sensor that low and you get lots of higher temps due to radiant heat. Even an AWOS will show temps that are 2-3 degrees F too high if you simply don’t mow around them, especially at night. Are these going to be installed at locations that will have staffing in charge of maintenance? Just having multiple temp sensors doesn’t change the fact that the sites will need weekly/bi-monthly attention.
Steven mosher said: (September 21, 2010 at 5:17 pm)
“but you really cant tell ANYTHING without a specification.”
I am confident that Mr. Mosher would agree that you can’t do specification without documenting the system requirements. Only then can you create and evaluate the specification.
Note that the system requirements include not just equipment and communications, but processes, data bases, staffing, maintenance, reliability and the development processes. Validation and Verification should be stringent.
Requirements should be made public.
I quickly scanned the comments (as well as the article) and did not see a link for the full report. For those who can’t find it, here’s a link.
http://www.oig.doc.gov/oig/reports/correspondence/2010.07.29_IG_to_Barton_Rohrabacher_STL-19846.pdf