She may not be the world's greatest expert, but she definitely knows whoever it is that might qualify for that honour. In reality, the top people have different skills and interests, and defer to one another's particular skills.
I'm well aware of the limits of my expertise, but nobody around here seems to have caught me out yet, and most of the people who post on the subject here are mostly woefully ignorant, to the point of being fooled by denialist propaganda.
Your posting and quoting methods make it *very* hard to tell what you are posting and what was posted by someone else. But I seem to see where you initially asked for the justification of being able to make
confuse accuracy with resolution). Then you ask the same question with
Why do you find this so surprising? Do you really not understand the basis of the isotopic analysis? If you can't measure the temperature of
The figure I am using is 0.05 Deg accuracy and resolution, as a ball park requirement, for indirect measurements, i.e. where there are no thermometers, of 1000s years ago. The bit is that it is indirect. Temperatures are only inferred.
This figure is because if the total effect you are looking for is only 1 or
2 deg, measurements are completely meaningless unless your measuring instruments are at least an order of magnitude better than the effect you are measuring, i.e. < 0.1 deg at least. This is a basic science principle.
I note that this is the second time someone has stated, well, go and find out, but don't seem able to provide any reference themselves, implying that they don't know either. Its not the basis that is at issue, I know the basis. I want to know *exactly* how such measurements can be so accurate considering that we are discussing average for the *whole* earth. This requires samples over many, many, points on the earth to get a valid, non biased value. I want to see a full statistical analysis of the errors to
*prove* that the 2 deg effect is not simply measurement noise. It certainly will be if, as you indicate, it is *only* 0.5deg accurate. You cant measure
1 deg effects reliable with 0.5deg measurement errors.
That's funny, I've been in science and engineering for 40 years and I've never heard that "principle" before.
Don't know what, the details of the research paper? No, I haven't even looked it up. But I don't automatically dismiss the idea that it is impossible to measure things from the past with a specified degree of accuracy.
That statement shows you *don't* understand the basis. Actually the temperature measured would relate to the surface of the water on earth mostly and not so much land. Otherwise you need to understand the basis of the effect to understand why it related to a global temperature.
Yes, which is provided by this measurement made virtually anywhere on earth.
You can get that simply by looking at the results.
I never said anything about the accuracy. I was asking about your numbers. You certainly *can* measure 1 degree effects with 0.5 degree accurate measurements. It can be done with worse accuracy than that. Stating it can't shows a lack of understanding of statistics.
Do not know if it is a science principle, but in meteorology you always want anything to have an accuracy of ten times better than the item you are calibrating.
I've told you how to answer your question. The task of tutoring you up to s econd year university thermodynamics is not one that is easily or usefully carried out in this kind of forum. E-mail me if you need more help. You've got my e-mail ( snipped-for-privacy@ieee.org).
:
The temperature at which water evaporates controls the O-16 to O-18 ratio i n the water vapour evaporated. The isotope ratio in the water being evapora ted would also have some effect, if it ever varied much, but there's a lot more water in the ocean than in the atmosphere. If you trapped enough of a particular isotope ratio you might move the baseline, but that doesn't seem to be a problem.
You've got to measure that ratio very precisely in the water that has been trapped at the time of interest, but that's just mass spectrometry, and if you have enough sample (and enough time) you can get as precise as you like .
and
f
in
,
If it isn't obvious to you, you do have defect in your understanding of wha t you claim to have been taught.
n
is
So what. Universal Gas Theory is a lot more mundane, and just works.
eir
the
Don't we all?
d
t
I've collaborated with electronic engineers who had worked at CERN. That's serious stuff. Jeroen Bellman works there, and provides a window into the c urrent technology.
I've no idea how much mass-spectrometer time was devoted to getting the iso tope ratios to the necessary precision, but it's entirely obvious that it i s perfectly practical to get to quite high precision. There's going to be a n upper limit dictated by the detector stability, but even that is going to susceptible to more elaborate measurment technique - alternating the sampl e of interest with a standard sample or whatever.
ant anything to have an accuracy of ten times better than the item you are calibrating.
It's nice if you can get that, but people measure stuff into the random noi se, and quite a lot of the work I've done has been about averaging enough m easurements to get well below the noise on each individual measurement.
For a lot of situations, the standard deviation of an aggregate of "n" meas urement is less than the standard deviation on a single measurement in prop ortion to the square root of "n". This gets messed up by 1/f noise, which i s why lock-in - modulate/demodulate - techniques have an advantage.
want anything to have an accuracy of ten times better than the item you ar e calibrating.
oise, and quite a lot of the work I've done has been about averaging enough measurements to get well below the noise on each individual measurement.
asurement is less than the standard deviation on a single measurement in pr oportion to the square root of "n". This gets messed up by 1/f noise, which is why lock-in - modulate/demodulate - techniques have an advantage.
I was posting about calibration in general. Does not necessarily have any thing to do with noise. For example calibrating pressure gages.
First, there is a difference in accuracy, resolution and precision. All three go into forming error in a measurement. Although Kevin seems to be asking about accuracy (but mentions resolution) what he really wants to know is the probable error range.
Those are measurements involving one thing under one condition with a device which should have very good precision, resolution and accuracy or there will be some question about the likely error of the measurement. In science 95% probability is a common threshold for any individual measurement error range. I expect in metrology a goal for the error range is 99.9% or maybe better.
When measuring a quantity in the field many measurements are taken and an average can reduce the random errors increasing the precision. Normally the resolution is not at issue. That leaves accuracy. I can't say much about this particular measurement. That would require a very thorough analysis of the methods and data used. Not having access to the data and not being willing to contact the authors about data or the methods, I'll leave it as an exercise for the student.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.