Toward a standard understanding of aquarium lighting

You know what sticks in my craw? "Watts per gallon," that's what!

For the purposes of growing plants, what matters really are two things: intensity of the light reaching the plants and the proportion of wavelengths which fall into the useful range for photosynthesis.

Let's look at intensity. I propose that we standardize on something like the intensity of light at the bottom of a 21" high, 55 gallon tank with 55 watts lighting with even dispersal throughout. We give it a name like wpg-unit and we deal with tank lighting in terms of intensity as measured in wpg-units. We could use lumens, lux, irradiance, and other photometric terms instead but it is so confusing that as a community, the aquarium hobby has rejected technical light measurement. The idea here is to have a unit which we are all familiar with, watts per gallon, more clearly defined and then have a simple look-up table to multiply by to arrive at the "real" intensity of lighting in the tank.

Now, I've got a tank that's 10 gallons, with 10 watts of fluorescent lighting at 15" height above the floor of the tank. According to the inverse square law, which is the rule used to calculate the intensity of light from a given source at a given distance which states that when the distance from the source doubles, the strength of light from it decreases to 1/4 the intensity of light at the original distance. So if the distance quintuples (5 x) the intensity will diminish to 1/25th the strength. The square of 5 is 5x5 or 25. The inverse square is 1/25. In the case of my hypothetical 15" tall tank, with the 1 wpg unit light (remember, that's a light which delivers the same intensity to the floor of the tank as a 21" tall 55 gallon with 55 watts lighting [or actually any tank with gallons equal the wattage of lighting and 21" tall]), the lighting is actually 1.96wpg units.

So:
The intensity of 1 watt light per gallon of tank capacity at the bottom, from 21" height = 1wpg-unit.

To find the intensity of your tank floor's light, adjust as follows, for height, assuming a light of the same strength as above
8" height, multiply a standard 1wpg-unit by 6.89
10 1/2" height, multiply a standard 1wpg-unit by 4
12" height, multiply by 3.09
14" height, 2.25
16" height, 1.72
18", 1.36
20", 1.1
22", 0.91
24", 0.76
26", 0.65
28", 0.56
30", 0.49

Of course, as plants grow closer to the surface, the intensity of light their upper leaves receive becomes stronger by the same rules, such that a plant 13" tall in a 21" tank, 55 gallons capacity with 55 watts lighting will receive 6.89 times more light at top than at the bottom.


There are issues with how we deal with colors and the spectrums of our lights.

I've noticed a trend toward higher and higher color temperatures and have heard people saying that a 10000K light is "whiter" than a 5500K, and that higher color temperatures are whiter in general than lower color temperatures. Color temperature is actually the closest match of a real light emitting object's spectrum to an ideal spectrum emitted by a perfectly black (at room temperature) object at a specific temperature. In the real world, spectrums are not the same, perfectly smooth distribution of wavelengths and intensities. In fact, a real spectrum may be totally missing entire ranges of wavelength (wavelength = color when of visible light) but match the same ideal spectrum as the sun at noon, in which case things will appear to be different colors under that source than they do in real sunlight. We are adapted to seeing things in the varying spectrum of sunlight as it reaches earth (~4900K to 6700K) so something that's white will reflect evenly back light from sources with said spectrums. At 10000K, the ideal spectrum has much more intense light in the blue frequencies so as source emitting light that most closely matches a 10000K color temperature will have a spectrum with much less intensity in the red end and towards the blue end, although it may be in the form of gobs of greenish light, some blue, and very little red at all or any other mix so long as it best matches a 10000K ideal spectrum. So, a white object, which reflects accurately the wavelengths which strike it at the same intensity, will look quite blue under 10000K light when contrasted with the same white under 5500K light will look quite blue. Nothing seen under 10000K lights will show the same colors as under lights in the 5000K - 6700K. Our brains compensate a great deal for varying spectrums so a white object will eventually appear white even when the soul source of light is 10000K color temperature but the colors we see will render differently than we are accustomed to.

Aphotic Phoenix said:

Actually there is a perfectly good reason why measuring light in "lumens" can be very misleading. "Lumens" is a measurement of human perception of brightness, but is measured by the color wavelength least useful for plants. Since the dominant wavelengths emitted by different bulbs / manufacturers can vary considerably, using lumens as a measurement can be very misleading. The reason that WPG in conjunction with color temperature is most often used is because on average it tends to be approximately correct. No one is trying to be lazy here, there are just too many variables.

PAR (Photosynthetically Active Radiation) is arguably the best measurement, but almost no manufacturer actually bothers to include a PAR rating in their light analysis, and most people don't have the money to sink into a PAR meter.

Click to expand...

No argument that "lumens" can be a misleading measure but actually since it is a measurement of the brightness of a thing as perceived by the human eye, it actually measures the output of an object along the entire visible spectrum including those frequencies useful to plants. The light source's visible output is then weighted according to the eye's sensitivity to the frequencies of that spectrum - which happens to be greatest in the greens and very weak in the photosynthetically useful blues and reds. A spectrally pure yellow or green light might have plenty of lumens strength but only shine in that particular color's frequency, and be pretty useless for growing plants.

PAR as a great absolute measurement of the intensity of light at all frequencies between 400 and 700 nanometers wavelength. Unfortunately, it isn't weighted according to the actual photosynthetic response to frequency, and a light which emitted mostly the wavelengths with the least photosynthetic response but emitted strongly could have a nice, high PAR but also be a poor grow lamp.

There are arguments for a Modified PAR (MPAR), which would be a weighted measurement reflecting the photosynthetic response to the wavelengths between 400 and 700 nanometers. Such a measurement for the output of a light source would be a great start. Since there are two peaks each for the response of the two varieties of chlorophyll found in plants and their respective high and low peaks are pretty close to one another's, perhaps two numbers, one for the output in the spectral range of the low peaks and one for the range of the high peaks. Admittedly, two numbers may be more confusing than one, but at least it would really mean something, consistently from lamp to lamp and manufacturer to manufacturer. Alternately, a single number reflecting the overall output in the region of the peaks could be almost as useful.

Now that we have a hypothetically useful measurement for the photosynthetically significant light emitted from a light source, which in no way guarantees decent color rendering but doesn't by any means rule it out, let's consider lighting a planted tank. The spectrum of the light source does matter, but that isn't the complete issue when it comes to planted aquaria.

The issue is that what actually counts for plant growth is intensity of the photosynthetically significant light which reaches them. This brings us to the inverse square law. Let us take lamp x without a reflector and call the MPAR measurement of intensity from 20 inches distance 1 unit or 1 MPAR. Well, at 10 inches distance, that same lamp's MPAR would have an intensity of 4 units. At 30 inches its MPAR would be 0.444 and at 40 inches its MPAR would be .25. So you see distance plays a major role. Using reflectors complicates this but the manufacturer could simply provide an easy to read lookup table.

But let us return to color temperature and wpg. Yeah, you can assume that pretty many lamps will have alright output along the most powerfully useful wavelengths for photosynthesis. But if one is limited for some reason in how much lighting or how much power one can employ, it may be a good idea to get lights which really emphasize those helpful frequencies over the rest and for this purpose, color temperature is useless. And as for wpg, that ignores the two factors which tell us how much light is getting where we want it.. Watts per gallon tells us nothing about the distance from the light to the bottom of the tank and the amount of wattage converted to visible light is quite different from one light source to another. So we don't really know if we have a high, medium, or low light tank - and this isn't taking into account the proportion of photosynthetically useful frequencies in that light.

Naturally, we want a simple rule of thumb, but like the 1" of fish per gallon rule, wpg is limiting. By questioning it, even proposing alternatives to it, we can perhaps encourage a change. Even a small one in the direction of a more realistic rule of thumb would be helpful.

Of course no one is trying to be lazy and yeah, there are lots of variables. But I think by using a single variable like MPAR with labeling giving its value from say 8", 10", 15", 20", and 30" we could make better decisions about lighting our aquariums. It'd be nice to be able to hang lights above the tank and know how much light was reaching the bottom or to light a nano-tank without concern for the common-sense saying that the wpg rule breaks down for small tanks.

Another point I'd like to make is that not everyone is equally involved in and invested in the hobby. Some people rather have more technical information than others and that information should be available - its up to interested hobbyists to work for it, to push it. Other people rather not be bothered and that's perfectly fine. A better, simple system could still benefit them and it still wouldn't prevent using the wpg rule just as consideration for the mass and activity level of fish when stocking a tank doesn't interfere with other people's using the inch per gallon rule.

Aphotic Phoenix said:

I couldn't find this article earlier while I was at home, but explains things rather well, and compares some various brands of lights to one another. It's about 5 years old, so finding the specific brands might be a bit difficult, but still useful for learning about the complexities of comparing light sources.

http://www.aquabotanic.com/lightcompare.htm

Click to expand...

WOW! Now thats the sort of information I like to see. Sometimes it isn't even the immediate practical application of information as much as how interesting it is. This PUR figure is promising. I'd like to have a simple figure like that.

Yes, chlorophyll A and chlorophyll B. Their absorption peaks in the red and blue respectively are pretty close to one another. My thought was that either a curve could be plotted which reflected the total or average absorption of both pigments by wavelength.

Or that PUR figure or an MPAR which would reflect the output of light weighted by photosynthetic action response to frequency would obviate the need to actually consider pigments; it just reflects the response of plants to different frequencies, which actually varies from plant to plant and is modified by the presence or absence of carotenoid pigments which help gather light.

Yeah, to consider all the variables, we'd need a computer program, data on the individual spectral response of different species of plants, the individual spectra of light sources, oh - it could go on forever, eh?

We haven't touched on any aesthetic issues yet, which would make forever go on forever forever. And that's something I'm unfortunate enough to be interested in.

If we consider color rendering, we might have to look at lumens again, no? Or RGB colorspaces?