One of the oldest bits of weather lore is "Red sky at night, sailor’s delight. Red sky in morning, sailors take warning." The sky is often red, or reddish orange, at sunrise and sunset. Reds, pinks, and oranges dominate at these times of day because the distance from your eye to the sun is greater at these times than when the sun is directly overhead. When sunlight travels through the atmosphere—and especially when it travels that extra distance from the horizon—an effect known as Rayleigh scattering takes place.
The British physicist John William Strutt (the third Baron Rayleigh) explained in 1871 that air molecules are so small they interfere with and scatter photons of light. The various molecules that make up the Earth’s natural atmosphere—mostly nitrogen and oxygen, with trace amounts of other gases—scatter blue light (shorter wavelength light) more efficiently than red, orange, or yellow light (longer wavelengths). So when you look up, you see more scattered blue light than red or orange light. This is why our sky appears blue during most of the daytime (Figure 6-1).
When the sun is on the horizon at dawn or dusk, however, its light must travel through more of the atmosphere before we see it than when it is overhead (Figure 6-2). This leads to more and more of the blue light being scattered away, leaving behind the red-orange-yellow light. As well, dust and aerosols in the atmosphere more ably scatter long wavelengths of light—leading to lurid red sunsets.
Figure 6-1. As sunlight hits the Earth’s atmosphere, air molecules scatter the shorter, blue wavelengths of light, but let the longer wavelengths through. Credit: NOAA ERSL.
As atmospheric particles get larger, they tend to scatter all wavelengths of light equally and indiscriminately. This explains why clouds, which are made of relatively large water droplets, are white.
Figure 6-2. This chart describes the absorption of sunlight by various atmospheric molecules. Look carefully in the 700 nm range of wavelengths: the dips in the "Direct Solar Irradiance at Sea Level" line indicate where molecules like water and oxygen are blocking out certain wavelengths of sunlight.
Of course, molecules don’t just scatter wavelengths of light. Certain molecules, such as water vapor, carbon dioxide, and ozone, also absorb specific wavelengths of light. To a measuring instrument on the ground, there is less light at those specific wavelengths than there is at other nearby wavelengths, because some of that light has been absorbed. By measuring these "holes" where wavelengths of light are absent, you can get a very good estimate of how much water vapor, ozone, and some other substances are in the atmosphere above your head.
Atmospheric aerosols are tiny particles of matter suspended in the atmosphere. These minute particles scatter and absorb sunlight, reducing the clarity and color of what we see. This is the effect we’re referring to when we talk about "haze" in the sky.
The LED photometer measures the difference between the voltage produced by a red LED, which absorbs light around the 580 nm range, and the voltage produced by a green LED, which absorbs light around the 500 nm range. The readings appear on the LCD, and are then stored on the Arduino’s EEPROM (and/or SD card, if you choose to use one).
You can chart this data by hand, extract the data, and graph it in a spreadsheet program, or upload it to Cosm (formerly Pachube) to share it with the world. More on this shortly.
On a clear day with very little haze, the voltages produced by the green LED and red LED will be somewhat similar. As atmospheric aerosols increase, however—whether over the course of a day, several days, or from season to season—the difference between the red LED’s voltage and the green LED’s voltage will increase. The green LED will produce less current when there is more atmospheric haze, because less green light than usual is reaching it.
Note
The voltages will almost never be exactly the same; Rayleigh scattering guarantees that more green light will be scattered by our atmosphere than red light. Also, white light reflected by clouds can sometimes cause the green LED to spike. The best results come from days without clouds (or with a uniform cloud cover).
Photosynthesis, the process by which green plants turn sunlight into carbohydrates, does not use green light.
Plants appear green precisely because they reflect green light; they have no use for those wavelengths, and therefore can afford to throw them away. Green leaves absorb a great deal of the light at red and blue wavelengths. You can use LEDs to measure how much of those wavelengths—how much photosynthetically active radiation—is reaching your detector.
You do this by summing the outputs from two LEDs: the same red LED as in the aerosol detector, and a blue LED. As with the aerosol detector, Arduino reads the voltages produced by the red and blue LEDs, displays them, and stores them where they can be graphed, extracted into a spreadsheet, or uploaded to Cosm.
Note
Forrest M. Mims, the man who discovered the Mims effect, has shown that using red and blue LEDs to measure photosynthetically active radiation agrees very well with more "professional" PAR sensors.
Atmospheric water vapor absorbs EM wavelengths in the range of infrared light, at around 940 nm. By using an infrared LED specifically designed to absorb light at 940 nm, you can get a very accurate representation of the amount of water vapor in the atmosphere. Simply compare the outputs of the 940 nm LED with a similar infrared LED sensitive to around 880 nm.
One thing to keep in mind, however, is that this type of measurement is incredibly dependent upon local weather conditions. The rapid passage of a storm front over your observational area can play havoc with your readings: obviously there’s going to be more water vapor in the air when it is raining, snowing, or about to do either. Thus, this gadget is better suited to measuring upper atmosphere water vapor—water vapor high above what we might think of as "the weather." Because of this, the best water vapor measurements are those made on clear days.
Okay, so you’ve collected a bunch of data with your LED photometer. What do you do now?
Once you’ve processed the data in a spreadsheet, you can graph it. Your results might end up looking like these charts by Forrest M. Mims, who has been collecting atmospheric data with an LED photometer in Texas since the early 1990s (Figure 6-3).
Figure 6-3. Forrest Mims III has gathered this data on total water vapor in the atmosphere using an LED-based photometer. Image used with permission.
Notice the patterns in Mims’s decades of data: nearly every reading shows some kind of seasonal variation. It might take a year or more before you can see a full cycle in your data.
You can upload your data for all the Internet to see thanks to a service called Cosm. Our previous book, Environmental Monitoring With Arduino (O’Reilly), contains a whole tutorial on getting started with Cosm; for now, let’s just say that Cosm makes it very easy to upload a CSV data file. Just cut and paste the serial output of the EEPROM reader program, or use the file DATALOG.TXT from the SD card, and place it on your website. Then go to your Cosm account and tell it where to find that file. Before you know it, your LED photometer data will be on the internet, neatly graphed, for all the world to see.