PAR:
Photosyntheticaly Active Radiation is measured in W/m² or J/(m²*s) and it is a measure of the total energy of all of the light that hits a surface in the wavelength interval from 400 nm to 700 nm. It is a quantitative measure, not a qualitative measure as it only tells you the total energy of the light you have measured. See section about the McCree curve(s) for the origin of PAR.
PAR can in broad terms be seen as an updated version of the older way of talking about the intensity of the lights used for aquariums where the wattage of the light was used. The difference is that when wattage was used about fluorescent, metal halide or halide lights it referred to the input energy, where PAR is the photon energy output in the above mentioned range. But if you know what the PAR of a given light at a given wattage is, then for most LEDs there should be a fairly linear correlation between the PAR and the wattage.
A word of warning about the wikipedia on PAR, as of today there are a few errors in the article and off topic sections…
McCree curve(s):
It is hard to talk about PAR without also looking at the McCree curves. They were first presented in the article “Test of current definitions of photosynthetically active radiation against leaf photosynthesis data“ (Redirecting) [paywall] as it has help define the spectral range in which we measure photons in order to determine the PAR value of a given light. The McCree curves are an average of 22 different plants response to photons with different wavelengths, and have been a guideline for scientists for decades. There have been made some modifications to the McCree curves and I suggest looking at Dr. B. Bugbees video, link here ( ). As to the importance of the McCree curves with regards to plant yield and morphology I suggest reading the following easy to read article “The McCree Curve Demystified” (The McCree Curve Demystified).
Quality of light:
As mentioned above PAR is a quantitative measure of light, not a qualitative measure, but what do we mean when we say that?
PAR is only telling us how much energy is radiated in the PAR interval, it does not tell us how the photons are distributed with regards to wavelength. Looking at the McCree curves we can see that plants, to some extent, use all photons in the interval, and looking at the solar spectrum (see figure 3 in “Spectral-Temporal LED Lighting Modules for Reproducing Daily and Seasonal Solar Circadian Rhythmicities” (PDF) Spectral-Temporal LED Lighting Modules for Reproducing Daily and Seasonal Solar Circadian Rhythmicities), it is clear that the suns spectral power distribution is fairly close to the McCree curve for the action spectra of the plants photosynthesis.
To illustrate why it is important to know the spectral power distribution of the light in addition to the PAR value, let us look at two different spectra in relation to the McCree curves.
First we take a look at the very popular RGB lights that many in the aquarium hobby are still using. They have 3 types of diodes, namely Red, Green and Blue and still the light looks white to us. This can be illustrated with the experiment “Tricolor R/G/B Laser Diode Based Eye-Safe White Lighting Communication Beyond 8 Gbit/s” (Tricolor R/G/B Laser Diode Based Eye-Safe White Lighting Communication Beyond 8 Gbit/s - Scientific Reports). When mixing the 3 laser beams you end up with what looks white light in the area where they intersect (see Fig. 1 in the above article). Granted that this is RGB lights taken to the extreme, but it will still serve as an example of why PAR does not tell us anything about the quality of the light. It also serves as an example as to why we can not trust our eyes to judge the quality of the light. You can take a look at the following article “ Tutorial: Theoretical Considerations When Planning Research on Human Factors in Lighting “ (https://doi.org/10.1080/15502724.2018.1558065), to get a basic understanding of how our eyes see colors. In this context the interesting part is figure 2 in section 2.2 as well as the text in that section.
Getting back to RGB light in relation to the McCree curves. When looking at RGB lights spectral power distribution compared to the action curve one can see that there are large portions of action curve where the RGB spectral power distribution is very low or plain missing. This means that the plants will most likely be missing parts of the wavelengths they need for some parts of their development.
The fix in this case is to add some white light with a spectral power distribution that is closer to that of the sun, while still adding the RGB on top of that if that is the look one wants in the, in this case, aquarium. Increasing the PAR on a RGB light is easy up to a point (limited by the maximum output, but more fixtures could be added), and the same can be said for decreasing the PAR.
Second example is, and here it gets a bit biased, the VTC series from Yujileds Sunlight Technology - Yujileds. Looking at the spectral power distribution from these LED, one notices that we have something much closer to the McCree action curve. This does make a light like this more suited for growing many different types of plants, but despite this it does not mean that the colors of the plants are to the viewers liking. Another example is the article “Shaping and Tuning Lighting Conditions in Controlled Environment Agriculture: A Review “ ((PDF) Shaping and Tuning Lighting Conditions in Controlled Environment Agriculture: A Review).The challenge in horticulture is to maximize the output while minimizing the input, but that is most of the time only something that works if you are handling plants that have very similar light requirements. Unfortunately most aquariums does not contain monocultures and often the plants come from different environments, some are from predominantly shaded areas, and others grow in bogs where almost only the roots are submerged but the part of the plant that is above water is exposed to full sunlight. Only a few of the plants are truly freshwater plants, but with the water depths that are found in most aquariums we can disregard the attenuation of light in water (link to an article on this at the bottom).
So which is better? You could argue that it depends on your goals, but as a minimum you should provide the plants with the spectral power distribution they need for a natural development. This means that if you like the look of RGB lights then make sure that there are enough photons outside the RGB peaks to ensure this, in other words WRGB with the majority of the light coming from the white .
Measurements of PAR:
In reality there are only a few solutions for getting a proper PAR reading from all light sources in the PAR range, all of them involve getting a scientific grade instrument. I will link to the two manufactures I know to make good instruments. The first of them is LI-COR. The important things to look at in the link here are the two figures showing the Normalized response in photon units and the Response in energy units (LI-190R Quantum Sensor | LI-COR Environmental).
The other is Apogee, a cheaper alternative to LI-COR but still very good, where the link provided only show the equivalent of the LI-CORs photon units (Quantum Sensors - PAR Meters | Apogee Instruments) . As can be seen, the two curves are very much alike.
There are many other instruments on the market, but many of them does not make their quantum yield public and given the fairly small price difference I would suggest getting a sensor from Apogee. LI-COR are a bit harder to come by, but the tech support that you get from LI-COR is outstanding. Tech support at Apogee is very good. I would in the end recommend Apogee based on price and, in my mind, slightly better quantum yield, as well as the fact that it is less of a hassle to get the sensor from Apogee.
What about using your mobile/tablet for taking PAR readings?
In order to understand how you can use an app for getting PAR readings we first need to look at how the camera sensor in mobile devices are constructed. As we saw in the article in the above section ‘Quality of light’, the human eye have only three types of cones that detect colors, and those colors are Red, Green and Blue, and this is what is being used in camera sensors. There is no need to record pictures in anything but those three colors as the brain gets “tricked” into seeing upwards to 100000000 colors just by using those 3 colors. Take a look at the following article “Camera Sensors: What Are They and How Do They Work?“ Camera Sensors: What Are They and How Do They Work? | FUJIFILM Exposure Center – USA for a quick walkthrough of how light is collected and detected by camera sensors. Granted this is for camera sensors, but to my knowledge most, if not all, mobile devices that have some type of camera in them, use the same principals for their sensors.
What does that mean for PAR readings that are made using a camera sensor? Well it means that the camera sensor will miss the photons that lie outside the cutoff range for the filters on the pixels. Remember the spectral power distribution from “Spectral-Temporal LED Lighting Modules for Reproducing Daily and Seasonal Solar Circadian Rhythmicities” in the Quality of light section above and then compare it to the sensitivity of the rods in our eyes from the article “Tutorial: Theoretical Considerations When Planning Research on Human Factors in Lighting “.
This means that right from the start you are missing photons that a proper PAR sensor would register. Obviously this is not a problem if the light you are measuring is a RGB light where the spectral output from the diodes exactly match the cutoff filters on the pixels, and that the software is compensating for the actual area of the sensor that the particular parts of the spectrum hits. But everything outside will not be counted and in order to get an estimate for those photons that software will need to make some assumptions about the light source. And here we get to the problem, because what assumptions have the programmer made about the light you are trying to get a PAR value from…
In order to get meaningful values the programmer will have to know the spectral output of a very wide range of lights and then have those in the software. You will then have to input the exact light source in the software (app) to get the numbers. But why not just look up the numbers from the database that was used for making the software?
The solution is not to get better apps, the solution is getting the manufactures to actually measure their lights and make the data public.
A few more things about light:
This section is dedicated to touch on a few other properties of light that are of some important when we look at lights for aquariums and the demands of the plants that will be using that light. So a quick intro to CRI, CCT and how big a difference there is with regard to the spectral power distribution that plants at different locations in the wild receives.
CRI is short for Color Rendering Index,
Is telling us how well the light is able the reproduce the colors of an object compared to what we see under natural sunlight. It should be said that is only true for light with a CCT (see below) at, or greater than 5000K. If the CCT of the light is lower than 5000K it is compared to the light from a source that is actually radiating at the given temperature (halogen or incandescent bulbs).
With CRI it can be said that the higher the CRI is (on a scale [0;100]) the better the light is at reproducing the natural colors. One should remember that we are moving into the realm of esthetics so just because a light have a high CRI value it might not be what you are looking for looks wise. There is a good chance though that the light with a high CRI, is fairly close the natural sunlight. I have seen examples of lights with a high CRI that were not that close to natural sunlight, but can not remember where I saw that, so this is anecdotal.
Some lights also lists Chromaticity group and diagram, but this is not something I have looked into, so I will refrain from talking about that here.
CCT is short for Correlated Color Temperature,
It is a way of trying to classify the apparent color of a light source, by comparing it to the radiation from a black body radiating at a given temperature (a black body is a theoretical object used in physics). All it is tell us is that the ‘typical’ photon in the light source we are looking at are is emitted at the same wavelength as that of a black body with a given temperature.
Spectral power distribution as a function of latitude and atmospherically composition.
In general it can be difficult to find sources on how the spectral power distribution varies throughout the day and with respect to latitude. I have one where they look at the diurnal variations in two different locations in the USA. Variation of outdoor illumination as a function of solar elevation and light pollution - Scientific Reports The interesting figure here is figure 1. The changes seen in the twilight regime is due to Rayleigh scattering. Nothing in this figure should come as a surprise, except the power distribution during night time. If any lesson should be taken from this, it should be that if people insist on having night light on their aquariums it should be green, not blue as most night lights are. And a possible explanation why the moon is made from green cheese…
I have a couple of extra articles in the further reading section about the above topic.
Further reading:
Flowgrow (Flowgrow), a German website where I have found that the plant database is fairly good, and far better than the one found at Tropicas webpage. Here you can see the geographical localities of where the plant can be found as well as other useful information. It is possible to view most of the content in English, but not all.
“LED Light Sources and Their Complex Set-Up for Visually and Biologically Effective Illumination for Ornamental Indoor Plants” LED Light Sources and Their Complex Set-Up for Visually and Biologically Effective Illumination for Ornamental Indoor Plants Further reading on human eye sensitivity and plant growth
“BASICS IN SOLAR RADIATION AT EARTH SURFACE-REVISED VERSION #2” https://hal.science/hal-02175988/document the practical chapter in this document is chapter 6 where they go through, among other things, the spectral power distribution as a function of latitude. The entire document is a very practical document when it comes to get a better understanding of how seasonal differences are produced.
“Patterns in the spectral composition of sunlight and biologically meaningful spectral photon ratios as affected by atmospheric factors” https://www.sciencedirect.com/science/article/pii/S016819232030143X variance in the solar radiation tropical zone vs temperate zone, Discussion on how the changes in the spectral power distribution drive different growth cycles, as well as an indirect way of comparing the spectral power distribution
“Absorption and attenuation of visible and near-infrared light in water: dependence on temperature and salinity” https://ir.library.oregonstate.edu/downloads/mc87ps04g - This is a fairly practical paper, and using their results you can check for yourself that in our typical setups, we do not have to worry about attenuation.
Photosyntheticaly Active Radiation is measured in W/m² or J/(m²*s) and it is a measure of the total energy of all of the light that hits a surface in the wavelength interval from 400 nm to 700 nm. It is a quantitative measure, not a qualitative measure as it only tells you the total energy of the light you have measured. See section about the McCree curve(s) for the origin of PAR.
PAR can in broad terms be seen as an updated version of the older way of talking about the intensity of the lights used for aquariums where the wattage of the light was used. The difference is that when wattage was used about fluorescent, metal halide or halide lights it referred to the input energy, where PAR is the photon energy output in the above mentioned range. But if you know what the PAR of a given light at a given wattage is, then for most LEDs there should be a fairly linear correlation between the PAR and the wattage.
A word of warning about the wikipedia on PAR, as of today there are a few errors in the article and off topic sections…
McCree curve(s):
It is hard to talk about PAR without also looking at the McCree curves. They were first presented in the article “Test of current definitions of photosynthetically active radiation against leaf photosynthesis data“ (Redirecting) [paywall] as it has help define the spectral range in which we measure photons in order to determine the PAR value of a given light. The McCree curves are an average of 22 different plants response to photons with different wavelengths, and have been a guideline for scientists for decades. There have been made some modifications to the McCree curves and I suggest looking at Dr. B. Bugbees video, link here ( ). As to the importance of the McCree curves with regards to plant yield and morphology I suggest reading the following easy to read article “The McCree Curve Demystified” (The McCree Curve Demystified).
Quality of light:
As mentioned above PAR is a quantitative measure of light, not a qualitative measure, but what do we mean when we say that?
PAR is only telling us how much energy is radiated in the PAR interval, it does not tell us how the photons are distributed with regards to wavelength. Looking at the McCree curves we can see that plants, to some extent, use all photons in the interval, and looking at the solar spectrum (see figure 3 in “Spectral-Temporal LED Lighting Modules for Reproducing Daily and Seasonal Solar Circadian Rhythmicities” (PDF) Spectral-Temporal LED Lighting Modules for Reproducing Daily and Seasonal Solar Circadian Rhythmicities), it is clear that the suns spectral power distribution is fairly close to the McCree curve for the action spectra of the plants photosynthesis.
To illustrate why it is important to know the spectral power distribution of the light in addition to the PAR value, let us look at two different spectra in relation to the McCree curves.
First we take a look at the very popular RGB lights that many in the aquarium hobby are still using. They have 3 types of diodes, namely Red, Green and Blue and still the light looks white to us. This can be illustrated with the experiment “Tricolor R/G/B Laser Diode Based Eye-Safe White Lighting Communication Beyond 8 Gbit/s” (Tricolor R/G/B Laser Diode Based Eye-Safe White Lighting Communication Beyond 8 Gbit/s - Scientific Reports). When mixing the 3 laser beams you end up with what looks white light in the area where they intersect (see Fig. 1 in the above article). Granted that this is RGB lights taken to the extreme, but it will still serve as an example of why PAR does not tell us anything about the quality of the light. It also serves as an example as to why we can not trust our eyes to judge the quality of the light. You can take a look at the following article “ Tutorial: Theoretical Considerations When Planning Research on Human Factors in Lighting “ (https://doi.org/10.1080/15502724.2018.1558065), to get a basic understanding of how our eyes see colors. In this context the interesting part is figure 2 in section 2.2 as well as the text in that section.
Getting back to RGB light in relation to the McCree curves. When looking at RGB lights spectral power distribution compared to the action curve one can see that there are large portions of action curve where the RGB spectral power distribution is very low or plain missing. This means that the plants will most likely be missing parts of the wavelengths they need for some parts of their development.
The fix in this case is to add some white light with a spectral power distribution that is closer to that of the sun, while still adding the RGB on top of that if that is the look one wants in the, in this case, aquarium. Increasing the PAR on a RGB light is easy up to a point (limited by the maximum output, but more fixtures could be added), and the same can be said for decreasing the PAR.
Second example is, and here it gets a bit biased, the VTC series from Yujileds Sunlight Technology - Yujileds. Looking at the spectral power distribution from these LED, one notices that we have something much closer to the McCree action curve. This does make a light like this more suited for growing many different types of plants, but despite this it does not mean that the colors of the plants are to the viewers liking. Another example is the article “Shaping and Tuning Lighting Conditions in Controlled Environment Agriculture: A Review “ ((PDF) Shaping and Tuning Lighting Conditions in Controlled Environment Agriculture: A Review).The challenge in horticulture is to maximize the output while minimizing the input, but that is most of the time only something that works if you are handling plants that have very similar light requirements. Unfortunately most aquariums does not contain monocultures and often the plants come from different environments, some are from predominantly shaded areas, and others grow in bogs where almost only the roots are submerged but the part of the plant that is above water is exposed to full sunlight. Only a few of the plants are truly freshwater plants, but with the water depths that are found in most aquariums we can disregard the attenuation of light in water (link to an article on this at the bottom).
So which is better? You could argue that it depends on your goals, but as a minimum you should provide the plants with the spectral power distribution they need for a natural development. This means that if you like the look of RGB lights then make sure that there are enough photons outside the RGB peaks to ensure this, in other words WRGB with the majority of the light coming from the white .
Measurements of PAR:
In reality there are only a few solutions for getting a proper PAR reading from all light sources in the PAR range, all of them involve getting a scientific grade instrument. I will link to the two manufactures I know to make good instruments. The first of them is LI-COR. The important things to look at in the link here are the two figures showing the Normalized response in photon units and the Response in energy units (LI-190R Quantum Sensor | LI-COR Environmental).
The other is Apogee, a cheaper alternative to LI-COR but still very good, where the link provided only show the equivalent of the LI-CORs photon units (Quantum Sensors - PAR Meters | Apogee Instruments) . As can be seen, the two curves are very much alike.
There are many other instruments on the market, but many of them does not make their quantum yield public and given the fairly small price difference I would suggest getting a sensor from Apogee. LI-COR are a bit harder to come by, but the tech support that you get from LI-COR is outstanding. Tech support at Apogee is very good. I would in the end recommend Apogee based on price and, in my mind, slightly better quantum yield, as well as the fact that it is less of a hassle to get the sensor from Apogee.
What about using your mobile/tablet for taking PAR readings?
In order to understand how you can use an app for getting PAR readings we first need to look at how the camera sensor in mobile devices are constructed. As we saw in the article in the above section ‘Quality of light’, the human eye have only three types of cones that detect colors, and those colors are Red, Green and Blue, and this is what is being used in camera sensors. There is no need to record pictures in anything but those three colors as the brain gets “tricked” into seeing upwards to 100000000 colors just by using those 3 colors. Take a look at the following article “Camera Sensors: What Are They and How Do They Work?“ Camera Sensors: What Are They and How Do They Work? | FUJIFILM Exposure Center – USA for a quick walkthrough of how light is collected and detected by camera sensors. Granted this is for camera sensors, but to my knowledge most, if not all, mobile devices that have some type of camera in them, use the same principals for their sensors.
What does that mean for PAR readings that are made using a camera sensor? Well it means that the camera sensor will miss the photons that lie outside the cutoff range for the filters on the pixels. Remember the spectral power distribution from “Spectral-Temporal LED Lighting Modules for Reproducing Daily and Seasonal Solar Circadian Rhythmicities” in the Quality of light section above and then compare it to the sensitivity of the rods in our eyes from the article “Tutorial: Theoretical Considerations When Planning Research on Human Factors in Lighting “.
This means that right from the start you are missing photons that a proper PAR sensor would register. Obviously this is not a problem if the light you are measuring is a RGB light where the spectral output from the diodes exactly match the cutoff filters on the pixels, and that the software is compensating for the actual area of the sensor that the particular parts of the spectrum hits. But everything outside will not be counted and in order to get an estimate for those photons that software will need to make some assumptions about the light source. And here we get to the problem, because what assumptions have the programmer made about the light you are trying to get a PAR value from…
In order to get meaningful values the programmer will have to know the spectral output of a very wide range of lights and then have those in the software. You will then have to input the exact light source in the software (app) to get the numbers. But why not just look up the numbers from the database that was used for making the software?
The solution is not to get better apps, the solution is getting the manufactures to actually measure their lights and make the data public.
A few more things about light:
This section is dedicated to touch on a few other properties of light that are of some important when we look at lights for aquariums and the demands of the plants that will be using that light. So a quick intro to CRI, CCT and how big a difference there is with regard to the spectral power distribution that plants at different locations in the wild receives.
CRI is short for Color Rendering Index,
Is telling us how well the light is able the reproduce the colors of an object compared to what we see under natural sunlight. It should be said that is only true for light with a CCT (see below) at, or greater than 5000K. If the CCT of the light is lower than 5000K it is compared to the light from a source that is actually radiating at the given temperature (halogen or incandescent bulbs).
With CRI it can be said that the higher the CRI is (on a scale [0;100]) the better the light is at reproducing the natural colors. One should remember that we are moving into the realm of esthetics so just because a light have a high CRI value it might not be what you are looking for looks wise. There is a good chance though that the light with a high CRI, is fairly close the natural sunlight. I have seen examples of lights with a high CRI that were not that close to natural sunlight, but can not remember where I saw that, so this is anecdotal.
Some lights also lists Chromaticity group and diagram, but this is not something I have looked into, so I will refrain from talking about that here.
CCT is short for Correlated Color Temperature,
It is a way of trying to classify the apparent color of a light source, by comparing it to the radiation from a black body radiating at a given temperature (a black body is a theoretical object used in physics). All it is tell us is that the ‘typical’ photon in the light source we are looking at are is emitted at the same wavelength as that of a black body with a given temperature.
Spectral power distribution as a function of latitude and atmospherically composition.
In general it can be difficult to find sources on how the spectral power distribution varies throughout the day and with respect to latitude. I have one where they look at the diurnal variations in two different locations in the USA. Variation of outdoor illumination as a function of solar elevation and light pollution - Scientific Reports The interesting figure here is figure 1. The changes seen in the twilight regime is due to Rayleigh scattering. Nothing in this figure should come as a surprise, except the power distribution during night time. If any lesson should be taken from this, it should be that if people insist on having night light on their aquariums it should be green, not blue as most night lights are. And a possible explanation why the moon is made from green cheese…
I have a couple of extra articles in the further reading section about the above topic.
Further reading:
Flowgrow (Flowgrow), a German website where I have found that the plant database is fairly good, and far better than the one found at Tropicas webpage. Here you can see the geographical localities of where the plant can be found as well as other useful information. It is possible to view most of the content in English, but not all.
“LED Light Sources and Their Complex Set-Up for Visually and Biologically Effective Illumination for Ornamental Indoor Plants” LED Light Sources and Their Complex Set-Up for Visually and Biologically Effective Illumination for Ornamental Indoor Plants Further reading on human eye sensitivity and plant growth
“BASICS IN SOLAR RADIATION AT EARTH SURFACE-REVISED VERSION #2” https://hal.science/hal-02175988/document the practical chapter in this document is chapter 6 where they go through, among other things, the spectral power distribution as a function of latitude. The entire document is a very practical document when it comes to get a better understanding of how seasonal differences are produced.
“Patterns in the spectral composition of sunlight and biologically meaningful spectral photon ratios as affected by atmospheric factors” https://www.sciencedirect.com/science/article/pii/S016819232030143X variance in the solar radiation tropical zone vs temperate zone, Discussion on how the changes in the spectral power distribution drive different growth cycles, as well as an indirect way of comparing the spectral power distribution
“Absorption and attenuation of visible and near-infrared light in water: dependence on temperature and salinity” https://ir.library.oregonstate.edu/downloads/mc87ps04g - This is a fairly practical paper, and using their results you can check for yourself that in our typical setups, we do not have to worry about attenuation.
