Whatever you measure, you always get numbers.™
But how do you make sure those numbers are meaningful?
First (and obviously) you have to know your research field and its methods.
Second you have to use appropriate instrumentation. And how do you know what is an appropriate instrument for photosynthesis measurements?
Do you look for nice pictures in brochures or websites with happy smiling scientists and their instrument in the field? No.
Do you believe in marketing speak like "most experienced manufacturer of infrared gas analyzers for measuring photosynthesis" or "We are considered a world leader in the manufacture and design of instrumentation for the plant and soil science markets"? No.
Do you look for technical buzzwords like "gold plated metal parts", "laser trimmed whatsoever", "reduced microprocessor response latency" or "hyper-quantum improved pre-amplifier gain"? No
The only way to know about the real quality of an instrument is to ask for verifiable specifications[ref]:
Unfortunately there is a lot of trickery in specification of inferior instruments to show the same numbers as in the LI-COR specifications. Just those numbers are either meaningless due to missing conditions or simply misleading.
Therefore you should plan your instrument purchase as careful as you plan your scientific experiment.
Or how some companies try to give unexperienced scientist the impression that their instruments are better then they acaudal are.
In science it is mandatory to give to a measured value also information of its error or confidence interval. Otherwise one can not judge the quality of the experiment.
The same is true for any technical specifications. Without additional information under what conditions specifications are valid they can not be verified.
It is very simple: Precision only tells you something about repeatability. This is a statistical measure and on short term measurements (minutes) depends only on the number of samples taken. Which means, with a sufficent large number of samples (n) the precision number can be made as small as wished. Remember from statistics: RMS = σ / √n
For example the following number does not tell you anything technically verifiable - only that somebody tries trickery on you:
Precision: 0.2 μmol mol-1 at 300 μmol mol-1
Meaningful precision has to be specify with the time of averaging. The right way to specify precision would look like this:
Signal Noise (Precision): 1 second signal averaging at 350 µmol mol-1 RMS: 0.07 µmol mol-1
But what you want to know in the first place is analyzer accuracy:
Accuracy: Maximum deviation: ± 5 µmol mol-1 from 0 to 1500 µmol mol-1
For most photosynthesis instruments available today on the market one does not find any specification of accuracy.
Well, this depends on the application. Capacitive humidity sensors are, compared to infrared gas analyzers, extremely cheap but have much less accuracy, stability and resolution compared to an infrared gas analyzer. Furthermore they are slow (the response time is in range of minutes). For ambient air humidity measurements they can do a good job if they are periodically (at last once a year) calibrated or replaced.
Nowadays they do not have a place in a proper photosynthesis instrument. They might give an estimate of a transpiration rate, but they certainly can not measure humidity accurate enough to be useful in calculation of the stomatal conductance or intercellular CO2 concentrations. For this a much better accuracy and resolution is mandatory.
In the context of sensor specifications it specifies how fast a sensor-signal follows the physical change it is measuring. That is usually specified as time to rise to 90% of final steady state value (called T90), measured from onset of step input change in measured variable. For poorly responding sensors you may find response quoted as T30 or T50.
Putting a sampling rate into the technical specification is just another trickery to mislead unexperienced scientists because this number only tells how many times the computer reads a sensor element but it tells absolute nothing about how fast the sensor itself follows a physical change.
For a example a typical capacitive humidity sensors has a response time of > 100 seconds. Now a computer can read the signal from such a sensor ten times a second (i.e. 10 Hz) but it still takes almost two minutes until the computer reads the right humidity.
On the other side, specifying a sensor (analyzer) with 10 Hz bandwidth means that a sensor reaches within 0.1 seconds 90% of the right physical value. That is 1000 times faster.
So, to relate a 10 Hz sampling rate with a sensor bandwidth is, just to name it friendly, a trickery.
H2O molecules stick to surfaces like tube walls and air filters. Therefore with tubes and filters between leaf chamber and analyzer any change in the chamber humidity can take minutes until the system reaches the new humidity equilibrium and is measured correctly.
Another disadvantage is that the response time of this setup depends on the flow rate of the system. The air has to flow through the leave chamber as well as through tubes and filter to the analyzer. The lower the flow rate is the longer it takes to get humidity equilibrium for correct measurement. With a gas analyzer directly coupled to a leaf chamber the response time is short and always the same, regardless of the chosen flow rate.
With all that said, we give this Rule of Thumb:
An instrument manufacturer is only as trustful as he is willing to give proper, i.e. verifiable specification.
We provide workshops for the LI-6400XT at your institute. Please contact us for more information.