Your School Rooftop Is an Observatory: A Student's Guide to Measuring Light Pollution

From a Bengaluru terrace to a rooftop in Varanasi — your school holds a citizen-science station nobody has switched on yet, and this guide shows exactly how to change that

It starts, almost always, with a single question in a science class. A student in Pune raises a hand and asks why the sky in their grandmother's village in Satara looks nothing like the sky above their school. The teacher pauses. The textbook has a diagram of the solar system but no answer for this. The student is not wrong to be confused — they have noticed, without knowing the word for it, the precise thing that SkyQI exists to measure.

The good news is that you do not need to travel to Satara to begin answering the question. Your school's rooftop — flat, elevated above street-level clutter, accessible on a Friday evening with permission — is already a functional measurement platform. All that is missing is a method.

This guide gives you that method. It is written for secondary and higher-secondary students in Indian cities, though it will work equally well for a science teacher, an astronomy club, or a curious family standing on any terrace from Shillong to Thiruvananthapuram. You will learn what to measure, how to measure it using nothing more than a smartphone and your own eyes, how to record it consistently enough for the data to be scientifically useful, and how to interpret what you find.

By the end of your first session, you will have contributed a real measurement to a real dataset. By the end of a year of measurements, you will understand your local sky better than most professional urban planners do.

Why the Rooftop, and Why It Matters

The school rooftop is not merely a convenient vantage point. It is, in most Indian cities, one of the most strategically interesting measurement locations available. Here is why.

Most light-pollution studies use readings taken in parks, fields, or designated dark-sky sites. These tell us where the sky is good. What they tell us less cleanly is what the sky looks like where most people actually live and where most children actually spend most of their time. A reading from a school rooftop in the middle of a residential neighbourhood fills a gap that researchers genuinely need filled.

Indian schools are also distributed in a way that no network of professional instruments can match. There are roughly 1.5 million schools in India. A fraction of those — even one in a thousand — taking regular sky-quality measurements would produce a spatial resolution of light-pollution data that no satellite currently delivers. Satellites measure integrated light leaving the atmosphere; a rooftop measurement from inside the atmosphere, looking up, is a fundamentally different and complementary data type.

There is also something less statistical and more immediate. The rooftop, elevated above the neem trees and the compound wall, above the streetlights at eye level, gives a clean view of the sky dome in a way that a courtyard does not. If the school is in a city, the horizon will be bright with skyglow; the zenith will be somewhat darker. Learning to read that gradient — which way the glow comes from, how high it reaches, whether it comes from one direction or all of them — is the beginning of understanding light pollution as a spatial phenomenon rather than an abstract statistic.

Equipment You Already Have

One of the persistent myths about citizen-science astronomy is that it requires expensive instruments. For basic sky-quality measurement, the list is short, and most of it fits in a school bag.

Your smartphone. Any phone with a camera made after 2018 will work. The key variable is whether you can manually set the ISO and shutter speed, either through the native camera app or through a free app like ProCamera (iOS) or Open Camera (Android). If you cannot — if your phone only has an automatic mode — you can still contribute visual estimates, and a standard photo taken in automatic mode still contains useful data for SkyQI's algorithm.

A free compass app or the phone's built-in compass. You need to record which direction you are photographing. Magnetic north is fine; you do not need GPS-calibrated bearings.

A dark-adapted torch. Use red light if you can — a red cellophane sheet over a small torch works. Red light does not collapse your pupils the way white light does, so you can take notes without ruining your night vision.

A simple notebook or a notes app. Paper with a clipboard is easier in the dark than a bright phone screen, which will wreck your night adaptation. Pre-print a data sheet (see the table in the Recording section below) and fill it in by feel and torchlight.

Optional but excellent: a Sky Quality Meter (SQM-L). These handheld devices cost roughly ₹8,000–12,000 new; your school's physics department or astronomy club may already own one. They give a direct numerical reading in magnitudes per square arcsecond, the same unit SkyQI reports. If you have access to one, use it. If you do not, your smartphone and naked eye will still produce scientifically usable data.

Preparing Your First Session

Good data comes from preparation. A session that begins with everyone crowding the rooftop at 8 PM with phone torches blazing will produce poor measurements. One that follows a brief protocol will produce data that holds up across months and across different observers.

Choose your date deliberately. The single most important variable you cannot control is the moon. A full moon brightens even a genuinely dark sky by two to three Bortle classes — a Bortle 2 site near Hanle effectively becomes a Bortle 5 under full moonlight. Plan your first measurement session within four days of a new moon. The new moons in the second half of 2026 fall on 13 August, 12 September, 12 October, 10 November, and 10 December in Indian Standard Time. Mark those windows now.

Get permission and assign roles. Three or four students per session is enough. Assign: one measurer (takes photos, reads SQM if available), one recorder (notebook), one compass holder (calls out cardinal directions), and one timekeeper (notes exact IST for each reading). Rotate roles across sessions.

Arrive at least thirty minutes early. Let your eyes adapt to darkness. Switch off all phone screens. Do not use white torches. The visual estimates you take — can you see the Milky Way? How many stars are visible in Orion? — require dark-adapted eyes, and dark adaptation in humans takes a genuine 20–30 minutes to complete. A flood of white light from a phone resets the process almost instantly.

Agree on a fixed set of measurement points. You will take multiple readings per session, each in a different direction. The four cardinal directions (north, south, east, west) plus straight up (zenith) is a good starting set. Mark these permanently if you can — a small chalk mark on the parapet wall, a piece of tape on the floor. Consistency across sessions is what makes your data comparable over time.

How to Take a Sky-Quality Photo

For SkyQI, the measurement lives in the photograph. The algorithm analyses the background sky brightness, the number and magnitude of visible stars, the colour and uniformity of skyglow, and several other parameters. Feeding it a good photo is important.

Manual settings, if possible. For a first session, try these starting values and adjust if needed:

• ISO: 1600–3200
• Shutter speed: 15–25 seconds
• Focal length: widest available (lowest focal length number)
• White balance: fixed at 4000K (not auto — auto white balance changes the colour in ways that confuse the algorithm)
• Focus: set to infinity manually; do not let the camera hunt for a focus point in the dark sky

Point toward the zenith or 45° above the horizon. Do not photograph directly at a streetlight or at the horizon below 20°. The algorithm needs enough actual sky in the frame to work with.

Keep the phone still. A 20-second exposure with a moving phone produces star trails and blurred skyglow — neither of which gives reliable data. Use a small tripod if your school owns one. If not, prop the phone against a parapet wall or lay it flat for the zenith shot (place it on a book for a slight angle). Even a rubber band holding the phone to a railing works.

Take three photos in each direction. Exposure-to-exposure variation exists. Three photos let the platform average the readings and raise confidence.

Record your settings. Write down the ISO, shutter speed, and direction for every set of photos. This metadata is what separates useful science from pretty pictures.

Recording Your Observations: A Data Sheet

Consistency is everything. Below is a simple table structure you can reproduce in your notebook or print as a data sheet. Fill in one row per photo set per direction per session.

A few column notes:

Altitude is measured from the horizon. Zenith is 90°. A fist held at arm's length covers roughly 10° of sky; a fully spread hand covers about 20°.

Cloud cover can be estimated on a 0–3 scale: 0 = clear, 1 = a few clouds (less than 25%), 2 = partly cloudy, 3 = mostly overcast. Do not take sky-quality measurements when cloud cover is 2 or higher — scattered cloud reflects streetlight back down and gives readings that look like a brighter sky than it truly is.

Moon phase should be recorded as "New", "Crescent", "Quarter", "Gibbous", or "Full", plus how many days from new moon (e.g. "New+3d"). This lets anyone reading your data reproduce your sky conditions.

Milky Way visible? is a simple yes/no. If you are observing in summer (June–September IST), look toward the south-southeast; in winter (November–February), look toward the southwest before midnight. This naked-eye check is a quick calibration of your eyes against your instruments.

Interpreting Your Results

You have uploaded your photos to SkyQI, your data sheet is filled in, and the platform has returned a Bortle class and an SQM value for each measurement. Now what?

Understand what the numbers mean. The SQM value is expressed in magnitudes per square arcsecond (mag/arcsec²). A higher number means a darker sky — this is counterintuitive but follows astronomy's magnitude convention, where larger numbers mean fainter objects. A reading of 21.5 is extremely dark (Class 2, comparable to Spiti Valley on a good night). A reading of 18.0 is heavily light-polluted (Class 6, comparable to suburban Bengaluru or the outer rings of Hyderabad). Most Indian city schools will return readings between 16.5 and 19.5.

Compare your zenith reading with your directional readings. Almost every school in an Indian city will show an asymmetric pattern: the zenith is darker than any horizon direction, and some directions are significantly brighter than others. The bright directions point toward the denser commercial or industrial areas nearby. The relatively darker directions often point toward a river, a large park, or an industrial zone that is not lit at night. This directional mapping is genuinely useful data — it lets urban planners and researchers understand where light is escaping into the sky, not just that some total amount of light is present.

Track the pattern over time. A single session produces a snapshot. A session per month for a school year produces something far more valuable: a time series. You will see seasonal variation (post-monsoon October skies in most of India are dramatically cleaner than pre-monsoon May skies, because the rains wash particulates from the atmosphere), variation tied to festivals (Diwali in late October or November produces a measurable spike in many city readings that takes several nights to subside), and sometimes a long-term trend — gradual brightening as a new commercial complex opens nearby, or an unexpected improvement if a factory closes or a major road loses traffic.

Three Naked-Eye Tests to Do Every Session

Before you take a single photo, spend five minutes doing these tests. They take no equipment, train your observational instincts, and produce the kind of qualitative data that complements the quantitative SQM reading.

Test 1: The Orion Belt Count. In winter sessions (October–March), find Orion — the three belt stars in a row are unmissable even from a Class 7 sky. Once you have the belt, try to count how many stars you can see in the entire constellation: the belt (3), the two shoulder stars Betelgeuse and Bellatrix, the two foot stars Rigel and Saiph, and the three fainter stars of the sword below the belt. That is a potential 10 stars, of which the sword stars are the faintest (around magnitude 4.5). If you can see all 10 cleanly, you are in at least a Class 5 sky. If you can only see the belt and the four bright corner stars, you are in a Class 7 or worse. Write down your count every session.

Test 2: The Zenith Star Count. Pick a circular region of sky roughly the size you can cover with your closed fist at arm's length (about 10° diameter). Count every star you can see inside that circle, near the zenith, over 60 seconds. This is easier with a second person calling time. More than 15 stars: Class 4 or better. Eight to fifteen: Class 5–6. Fewer than eight: Class 7 or worse.

Test 3: The Skyglow Direction Test. Stand at the centre of the rooftop and slowly turn 360°, noting how bright the horizon glow is in each of the four cardinal directions. Pick the brightest and the darkest. The ratio between them — even estimated qualitatively as "twice as bright" or "ten times brighter" — tells you about the spatial structure of your local light pollution. Record your estimate. After several sessions, you will notice the pattern is remarkably stable unless something changes in the local environment.

What This Means for SkyQI Readings

Every measurement your school contributes to SkyQI does two things simultaneously: it improves the map of Indian light pollution, and it generates a longitudinal record of your specific location that no satellite or fixed station can produce.

The platform's value scales with density. A single reading from a school in Nashik tells us that Nashik has a particular sky quality on that particular night. Ten readings from the same school across ten months tells us how Nashik's sky quality varies with season, humidity, moon phase, and local events. Readings from twenty schools across Nashik tells us the internal spatial structure of the city's light pollution — which wards are brighter, where the industrial zones show up in the data, whether the river corridor is measurably darker than the commercial strips.

When you upload, use the SkyQI location tag consistently. If your school's rooftop is your measurement point, tag it exactly the same way every session — the same coordinates, the same label. The platform can then link your readings into a coherent time series automatically.

There is a specific contribution only schools can make. Most casual SkyQI users take a measurement once or twice — on a trip to Vagamon or a visit to Pushkar during a festival — and then not again for months. That is useful data but it is spatially sparse and temporally isolated. A school that commits to one measurement session per month, year-round, from a fixed rooftop in the middle of a city, produces exactly the kind of long-term baseline that researchers need to detect trends. You are not just measuring the sky; you are building the archive.

A Note on Diwali, Festivals, and Special Measurements

Some of the most scientifically interesting measurements your school can take are the ones tied to specific events. Indian festivals produce measurable spikes in sky brightness that persist for several nights.

During Diwali — which typically falls in October or November, depending on the lunar calendar — the combination of fireworks (aerosols and smoke) and increased decorative lighting reliably brightens skies across northern and central India. A reading taken on Diwali night, compared with a reading from the same location a week earlier and a week later, gives you a direct measure of the festival's sky-quality impact. This is publishable-quality citizen science.

Similarly, the inauguration of a new commercial mall, a cricket stadium event, or a large outdoor mela near your school can produce localised brightness spikes. The methodology is straightforward: establish a baseline (at least three readings across a week of ordinary nights), take readings during the event, and take post-event readings until the sky returns to baseline. The difference is the event's contribution to local light pollution.

Other events worth measuring around: Republic Day and Independence Day celebrations (floodlighting of public buildings), construction of new elevated expressways with lighting (a permanent change), and even something as simple as a neighbouring building switching from warm-white to cool-white LED exterior lighting (cool-white LEDs scatter more blue light and produce higher SQM penalties).

Staying Safe and Getting Permission

A practical note, because it matters.

Permission from school administration should be obtained in writing and should specify the dates and approximate times of sessions. Most school authorities are receptive once the educational purpose is clear — "we are measuring light pollution for a citizen-science project" is a more persuasive framing than "we want to do astronomy on the roof". If your school has a science club or participates in any olympiad or science fair structure, framing the rooftop sessions as part of that programme tends to ease the approval process.

Safety on the rooftop at night requires: at minimum two students plus one adult supervisor at all times, stable footwear (not slippers), awareness of the parapet edge, and a mobile phone with a charged battery for emergencies. Do not lean over parapets to photograph lower horizons — a horizon reading is not worth a safety risk.

Younger students (Class 6 and below) should always have an adult present who is specifically designated as the safety supervisor, not the science lead. The science lead and the safety supervisor should be two different people.

What To Do Tonight

If there is a clear sky above your school tonight, and the moon is thin or absent, here is a complete first session in six steps.

First, check the moon phase. If the illumination is above 50%, wait for a darker night — not because tonight's measurement is useless, but because your first experience of your own school's sky should be the darkest version of it, not the moonglow-brightened version.

Second, arrive on the rooftop 30 minutes before you plan to measure. Turn off all unnecessary lights. Let your eyes adapt. Do not look at a bright phone screen during this time.

Third, do the three naked-eye tests: the Orion belt count (if it is winter) or an equivalent bright constellation count (if it is summer — try counting stars in Scorpius, which is a brilliant southern constellation from Indian latitudes in June through September). Record every count in your notebook.

Fourth, set up your phone with the manual settings described above. Take three photos toward the zenith, three toward the north at 45° altitude, three toward the south at 45° altitude. Write down every setting.

Fifth, upload to SkyQI immediately or the following morning. Use the school's consistent location tag.

Sixth, compare your result against the table in the Interpreting Your Results section. Write one sentence in your notebook summarising what you found: "Our school rooftop is Bortle 7, SQM approximately 17.8. The brightest direction is south, toward the new shopping complex on the main road."

That sentence is the beginning of a scientific record. A year from now, you will have twelve of them, and the story they tell together will be more interesting than any single number.

The Vedic astronomers who wrote the Surya Siddhanta measured the sky systematically because they knew that one night's observation meant nothing and a year's observations meant everything. The methodology is more sophisticated now, the instruments more precise, and the platform that receives your data rather more distributed than a manuscript copied by hand. But the principle is exactly the same: the sky does not reveal itself all at once. It reveals itself, slowly and honestly, to those who keep showing up with a notebook and an open eye.

Your rooftop is ready. The sky is already there. What it is waiting for is you, consistently, with your phone and your data sheet, month after month — turning the ordinary act of looking up into something that actually gets recorded, shared, and counted.