# MultispeQ fluorescence

by cfastie | 15 Dec 22:54

Image above: Tristan captures Robert explaining the MultispeQ.

At the Public Lab Barnraising last month I got to meet Geoff and Robert from the PhotosynQ team at Michigan State University. The crash course I got in all matters MultispeQ brought me about 3% of the way to understanding what I might be able to do with one of the devices. I have found a few diagrams that helped me add a few percentage points to that score, so I am posting them here for easy reference. These diagrams are primarily about fluorescence, which is one of the more clever measurements made by the MultispeQ.

The PhotosynQ MultispeQ is a device that leverages two recent advances:

1. The ingenious methods of measuring photosynthesis which plant physiologists and biochemists have discovered in the last few decades, and
2. The more recent availability of cheap microprocessors, single color LEDs, photo diodes, and other sensors.

PhotosynQ is moving fast to put the capability of field equipment worth about US$20,000 -$30,000 (?) into the hands of hundreds of people for US\$150 per unit. The idea is to allow people everywhere to make precise measurements that allow very good estimates of the photosynthetic health of plants. A wide network of such devices could change the resolution at which we understand the environmental response of plants.
.

Geoff explains the MultispeQ to Tristan.
.
We have been having a conversation about one of the measurements that the MultispeQ can make. Most of the visible light that is absorbed by the pigments in plant leaves drives photosynthesis, but some of it just heats up the leaf, and a small fraction causes fluorescence -- it is radiated away as a new color of light.
.

I added the red text above based on the information in the caption, but I’m not sure I interpreted that information correctly. Some of the actual values vary a lot among different species of plants and for other reasons, so it’s probably close enough. Source
.

The above diagram represents just the light that is absorbed by photosynthetic pigments, so reflectance and transmittance are not included. There are only three important possible fates of the energy entering this system.
.

By manipulating a leaf a certain way and then measuring how much it fluoresces, you can learn a lot about that leaf’s potential to do photosynthesis. Because there are only a few possible fates of the light energy absorbed by plant pigments, if you control or measure all but one, you can estimate that remaining one. The general idea is to measure fluorescence and estimate photosynthesis. This estimate can be very accurate and can reveal differences in plant health that are completely invisible to the eye.
.

Laboratory plant experiment in which seedlings were treated the day before with herbicide. The lower image is an index based entirely on fluorescence and indicates slowing of photosynthesis when the plants still appear healthy (upper photo). This is the type of result many have expected from Public Lab NDVI cameras but have not achieved. The MultispeQ does not produce images like this, but instead produces measurements on individual leaves which could lead to the same conclusion.
.

The light fluoresced by green plants is lower energy (longer wavelength) than the light absorbed by the plant pigment. Chlorophyll fluoresces most strongly at 690nm (red) and at 740 nm (near infrared). The red light fluoresced by chlorophyll is easy to see if you shine UV light on an extract of leaves in a darkened lab. Outside, the amount of fluoresced light is trivial compared to the sunlight reflecting off the leaf, so neither the red nor the near infrared emission is noticeable.
.

Different extracts of plant pigments fluoresce at slightly different wavelengths, but this is a typical result for Chlorophyll a. Source.
.

A chlorophyll extract in alcohol shown under white light (above) and UV light inducing fluorescence (below). The red glow is fluorescence at the peak of 690 nm in the previous figure. Source.
.

Fluorescence from plant pigments is inconspicuous compared to reflection of sunlight from plant leaves. Source. Open the full figure and its caption here.
.
I still don't understand the photosynthetic pathways that allow the fluorescence measurement to tell us so much about photosynthesis. But I guess I'm better prepared to learn about that now. My interpretations here are from brand new things that I have learned in the past few days, and most of them are probably poor approximations of reality. So let me know about the glaring errors.

This is a really good collection of information about photosynthesis and spectral analysis.

My understanding was that NDVI has largely been used for wide area scanning, but it isn't so useful at a per-plant level. However, in a controlled environment (such as an indoor aeroponics station or a small garden at night), per plant analysis might be possible by trying controlled experiments to separately judge reflectance+transmittance, phosphorescence, and heat; that leaves photosynthesis as the free variable. It might take a bit more than a single camera and a single light source for such a battery of tests though. If we figure out how to do those tests by hand, the next step is to automate it.

Btw, some of the information and graphics have linked sources, others do not. It'd be great if all the sources could be filled in.

Woops I skipped the part at the top where it seems to suggest in advertisement language that the PhotosynQ gives you this and more!

Any chance you could remove the PhotosynQ advert or describe how PhotosynQ addresses all of the salient bits of information about photosynthesis contained in the majority of the note?

Is this a question? Click here to post it to the Questions page.

Thanks Bryan. You're right that NDVI rose to fame as a satellite and airborne tool, and has only recently been used more widely at close range. It can work well up close, but it's potential might have been overstated in some of our presentations. Distinguishing stressed from healthy plants requires a well tuned Infragram camera and driving it very precisely. It's far from plug and play. And NDVI might never be as effective as fluorescence at this. The MultispeQ makes a non-fluorescence measurement (SPAD) that is essentially NDVI using light transmitted through the leaf instead of reflected from it. So it will be enlightening to compare that with fluorescence based measurements. And in a development of great symmetry, there are plans to map chlorophyll fluorescence from space using sensors aboard the OCO-2 which were intended to map CO2.(Great early animations of CO2 generation and dispersal.) The only way they can see the tiny glow of plant fluorescence from space is to look at the wavelengths of Fraunhofer lines where little sunlight reaches Earth. Are scientists clever or what?

The images in the note that do not have links were scraped from behind pay walls.

I have added a link to the description of measurements which can be made by the MultispeQ. I am still coming up to speed on understanding how these work and what they are good for. This note was intended to organize some information about the background science related to just one type of measurement (fluorescence) made by the device. I am the wrong person to explain much more about this. I don't even know what wavelength the device measures to quantify fluorescence.

I think I am flattered that my attempt to describe the context of the MultispeQ struck you as an advertisement. No two ways about it, this is a once in a lifetime opportunity to hold the power of a Real Scientist in your own hand. Don't delay, become a beta tester today!

Is this a question? Click here to post it to the Questions page.

Good stuff. From my perspective NDVI is good for mapping vegetation and can provide some indication of overall health and is good at differentiating between vegetated and non-vegetated areas but fluorescence is a better indicator of plant function. The remote sensing community is optimistic that in the next couple years new satellite sensors and algorithms will be able to provide decent estimates of global gross primary productivity (GPP) based on fluorescence measurements. Currently different estimates of global GPP vary by a factor of two. Passive mapping of fluorescence (usually called solar induced chlorophyll fluorescence [SIF] in the remote sensing literature) is not easy since, as the image above illustrates, the signal from fluorescence is very weak and it's difficult to separate that from all of the reflected light.

NDVI can still be useful at close range but it is quite different from coarse resolution satellite images. At close range a single pixel might fall entirely on a leaf but from satellite the pixel is a mix of leaves, branches, soil, shadows.

Apologies if I was being pedantic. Now that I think about it, cleaning up the note isn't a great solution. Instead of changing this note, the pertinant information would be great fodder for a wiki which could be cultivated, watered, artificially selected, and spectroscopically analyzed.

I saw @liz email about a high level farming wiki. This sort of content might make a good offshoot of that. I'll take a look.

I agree that this material is more appropriate for a wiki than a note, but I wanted people to see it.

Such a wiki should be built by somebody at PhotosynQ who actually understands this stuff, and the wiki should live at PhotosynQ where lots of people will be looking for information on how and why the MultispeQ works.

Stripped out all the juicy plant bits and spectral mumbo jumbo. added it here: http://publiclab.org/wiki/multispectral-analysis-for-determining-plant-health

Formatting is ugly on the wiki page, but whatever.