Rain Gauge Experiment for Kids: A Month of Real Data

A rain gauge experiment for kids looks like the simplest project on this list, and it is the one most often built wrong. Cut the top off a bottle, invert it as a funnel, tape on a ruler, done. Except the bottom of a plastic bottle is curved and the walls near the base are tapered, so the first two centimetres of your scale are lying. There is a one-minute fix for that, and with it you get a genuine instrument that produces a month of data your kid collected personally. That dataset, mostly zeros with a few interesting days, teaches more about real science than the building does.

Key Takeaways

• You cut a 2 litre bottle, weight the base, pre-fill to a marked zero line, and read rainfall in millimetres once a day for a month.
• The core idea is that rainfall is a depth, not a volume, which is why a small gauge and a big one give the same answer.
• Best for ages 6 to 12. A six-year-old can read the scale; the graphing is where older kids get the payoff.
• What goes wrong: measuring in the curved base, or siting the gauge under the eaves of a house.
• Expect most days to read zero. That is the data, not a failure.

What You’ll Need

Item	How much	Where to get it	If you don’t have it
Clear plastic bottle with straight sides	1, 2 litre	Any fizzy drink bottle; straight sides matter more than size	A tall straight-sided glass jar plus a separate funnel
Gravel, marbles or clean stones	Enough for 3 cm in the base	Garden, aquarium shop	Sand, though it clouds the water
Waterproof tape or a label strip	About 25 cm	Hardware shop, stationery drawer	Permanent marker straight onto the plastic
Ruler with millimetres	1	Home	A printed paper scale, laminated with clear tape
Scissors or a craft knife	1	Home. Adult use only	Strong kitchen scissors
Water	A cupful	Tap	Any water
Notebook or spreadsheet	1	Home	Phone notes, but a paper chart on the fridge gets used more
Stakes, bricks or a bucket of sand	To hold it steady	Garden	A heavy plant pot the bottle sits inside

Time: prep 20 min, then 2 min a day for 30 days Cost: everything is probably in your kitchen or recycling bin Mess level: low

Do the Experiment

Step 1 — Cut the bottle at the shoulder

An adult cuts cleanly around the bottle where the straight wall begins to curve inward, usually 6 to 8 cm below the cap. Keep the top piece. Run a finger along the cut edge and trim any spikes, or cover the rim with a strip of tape, because a cut PET edge is sharp enough to slice skin.

Step 2 — Weight the base and pour in the zero water

Drop in gravel until it is about 3 cm deep. Then pour water in until the surface sits a centimetre or two above the top of the gravel and above the point where the bottle wall becomes perfectly straight. This is the step every tutorial skips. The curved base and the tapered lower wall hold an unpredictable volume per millimetre of height. Starting your scale above that zone means every millimetre you read afterwards is a true millimetre of rain.

Step 3 — Mark zero and build the scale

With the bottle standing on a level surface, mark the water line and label it 0. Then tape your strip up the side and transfer millimetre marks from the ruler, starting at that zero line and going up. Label every 10 mm. Check the strip is vertical: a scale taped at a slant reads high.

Step 4 — Invert the top as a funnel

Remove the cap, turn the top piece upside down and push it into the open bottle so it sits like a funnel. Because it came from the same bottle, its mouth is the same diameter as the cylinder below, so you can read depth directly with no arithmetic. If you ever use a wider funnel, you must divide your reading by the ratio of the areas.

Step 5 — Site it properly

This decides whether your data means anything. The gauge needs open sky: no eaves, no branches, no wall nearby. The working rule used by weather services is to stand at least twice the height of the nearest obstacle away from it, so a 4 m tree needs 8 m of clearance. Set the rim 30 cm to 1 m above the ground and anchor it so wind cannot tip it.

Step 6 — Read, record, empty, repeat

Pick a time and keep it. CoCoRaHS, the volunteer precipitation network run from the Colorado Climate Center, asks its observers to read at 7 a.m. local time, which is a good habit to copy. Each day, read to the nearest half millimetre with your eye level with the water, write it down, tip the water out, and refill to the zero line.

What you should see: Over a month in most climates, 15 to 25 days of zero, a handful of days between 1 and 5 mm, and perhaps one or two days above 15 mm. A heavy thunderstorm can deliver 30 mm in an hour. The shape of that distribution, a lot of nothing and a few big events, is one of the most useful things a kid can learn about weather and about data in general.

What’s Actually Going On

Rainfall is reported as a depth because that is the physically meaningful quantity: 1 mm of rain means every square metre of ground received 1 litre of water. A gauge with a mouth twice as wide catches four times as much water, but into a cylinder with four times the floor area, so the depth comes out the same. That cancellation is why a bottle on your balcony and the official 8-inch gauge at the airport should broadly agree.

Should, not will. Rain gauges have a known systematic bias called wind undercatch. Air flowing over the gauge mouth accelerates and deflects the smallest droplets away, so a gauge typically records a few per cent less than the true rainfall, and much more in strong wind or with snow. That is why siting rules exist and why national networks use standard heights. Your bottle is also losing a little water to evaporation between readings, which matters on a hot day: a 2 cm wide water surface in dry air can lose half a millimetre in 24 hours. Both errors push in the same direction. Your readings are a slight undercount, and knowing that is better than pretending otherwise.

The other thing your month of data will reveal is that rainfall is intensely local. A convective storm cell can be 3 km across. Two gauges on opposite sides of a town routinely differ by a factor of two on the same afternoon, which is exactly why volunteer networks like CoCoRaHS exist: a professional network cannot be dense enough to catch what a thousand bottles in a thousand gardens can. The US Geological Survey’s water-cycle material makes the same point from the other side, describing precipitation as the step that connects atmospheric water back to the land surface, with rates that vary from under 25 mm a year in the driest deserts to more than 10,000 mm in the wettest tropics.

How to Explain It to Your Kid

Ages 5–8: how deep did the puddle get?

Tell them the number on the gauge is how deep the rain would sit on a flat table if none of it ran away or soaked in. Then do the comparison that makes it real: 10 mm is about the thickness of their little finger, and that much rain over the whole garden is hundreds of buckets of water. Let them pour the gauge out into a measuring jug so they feel the amount.

Ages 9–12: a depth, not a bucketful

Introduce the unit properly: 1 mm of rain equals 1 litre per square metre. Have them measure the area of the patio in square metres and calculate how much water fell on it after a rainy day. The number is always bigger than they guess, and the calculation is a genuine use of area they will meet again in maths class.

Ages 13+: bias, siting and the distribution

Give them the three error sources: wind undercatch, evaporation, and siting. Then have them design a test for one of them, for instance running two identical gauges, one in the open and one 1 m from a wall, for a fortnight. Also have them plot their month as a histogram rather than a line. Rainfall is not normally distributed, and seeing a skewed distribution in their own data is a better statistics lesson than a textbook example.

The question to ask: “If we doubled the width of the funnel but kept the same bottle underneath, would the reading go up, down, or stay the same, and why?”

Why Teaching This Early Pays Off

The value here is not learning that rain can be measured. It is a month of commitment to a boring daily reading, which is what almost all real science actually consists of. The National Research Council’s Framework for K-12 Science Education places “analyzing and interpreting data” among its eight core practices, and a dataset a child gathered themselves carries a kind of authority that a worksheet never will.

It also matters that the data is real enough to be contributed. In countries where CoCoRaHS operates, a family can submit their readings to a database used by forecasters, hydrologists and insurers. The step from “I measured something” to “someone used my measurement” is rare in school science and worth engineering when you can.

A caution on the pedagogy. Kirschner, Sweller and Clark’s well-known critique of minimally guided instruction applies directly: handing a kid a bottle and expecting them to discover meteorology does not work. The explaining you do alongside the measuring is the active ingredient. And no, a month of rain data will not change anyone’s school trajectory. It will give them one experience of owning a dataset, which is a reasonable thing to aim for in a month.

Before You Start: What Parents Should Know

Best age	6 and up for reading, 9 and up for the maths. Below 6 the daily routine falls apart by day four.
Supervision	An adult cuts the bottle. Cut PET edges are genuinely sharp and cause the only injuries this project produces.
Real hazards	The sharp cut rim, and standing water. Tip the gauge out daily; a gauge left full for weeks breeds mosquitoes.
Don’t	Do not substitute a glass bottle or jar for the outdoor gauge. It will eventually get knocked over, and broken glass in a garden is a long-lived hazard.

When It Doesn’t Work

• The gauge reads rain on a dry day. Either dew ran off a nearby surface into it, or you sited it under an eave or branch where drips concentrate. A single drip line can deliver ten times the real rainfall.
• It reads far less than the official station. Check siting first, then look at wind. A gauge in an exposed, windy corner undercatches badly. Moving it to a sheltered-but-open spot, screened by distant objects rather than near ones, usually fixes it.
• The water level drops with no rain. Evaporation. Add a thin film of mineral oil on the water surface if you are comparing over several dry days, or simply read daily and accept the small loss.
• The bottle blows over every windy night. It is too light. Bury the base in a bucket of sand or strap it to a stake at two points, not one. A single tie lets it pivot.
• The scale seems to read high. The tape is slanted, or your zero line is below the point where the wall goes straight. Re-mark zero higher up.

Take It Further With the Rain Gauge Experiment for Kids

Run two gauges and compare. One in the open, one deliberately badly sited near a wall or under a tree edge. Log both for two weeks. The difference is a measurement of a systematic error, which is a more sophisticated result than the rainfall itself.

Add the catchment calculation. Measure your roof’s footprint, then after each rainfall calculate how many litres landed on it. Compare that to a water-butt’s capacity. This is the arithmetic behind rainwater harvesting and it makes the depth-versus-volume idea permanent. Pair the daily logging habit with a homemade barometer and a full backyard weather station, or step back to the water cycle modelled in a sandwich bag.

What to Watch For Over the Next 3 Months

• Week 4: They correct a weather report. Hearing “they said 5 mm but we got 11” means they now trust their own instrument, which is the whole point.
• Month 2: The vocabulary moves. “Millimetres of rain” replaces “a lot of rain” in ordinary conversation.
• Month 3: They ask where the monthly average on a weather site comes from, and whether their garden matches it. That question leads to climate normals and long records, and it is worth following.

Frequently Asked Questions

How accurate is a bottle gauge compared to a real one?

Within about 10% if it is well sited and you measure above the tapered base. The official 8-inch gauges used by the US National Weather Service, and the 4-inch gauges used by CoCoRaHS volunteers, are more precise mainly because of their larger mouths and careful siting, not because of exotic engineering.

Can we measure snow with it?

Badly. Snow blows across a narrow opening rather than falling into it, and gauges undercatch snow severely. The standard method is to collect snow in a wide cylinder, melt it, and report the water equivalent. A bottle is the wrong shape for that.

What time should we read it?

Any consistent time. Picking the same hour daily matters more than which hour. Volunteer networks commonly use early morning so that a reading covers the previous calendar day, which also means you measure before the day’s evaporation.

Our area had no rain for three weeks. Is the project dead?

No, that is a result. A run of zeros is data about your climate, and it sets up a genuinely good question about why some places get rain in sharp bursts and others get drizzle for days. Keep the log running and let the first storm land in a dataset that already has context.

---

About the author

Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.

---

Sources

1. CoCoRaHS. “About Us.” Community Collaborative Rain, Hail and Snow Network, Colorado Climate Center. https://www.cocorahs.org/Content.aspx?page=aboutus
2. US Geological Survey Water Science School. “Precipitation and the Water Cycle.” https://www.usgs.gov/special-topics/water-science-school/science/precipitation-and-water-cycle
3. US Geological Survey Water Science School. “Evapotranspiration and the Water Cycle.” https://www.usgs.gov/special-topics/water-science-school/science/evapotranspiration-and-water-cycle
4. National Research Council. (2012). A Framework for K-12 Science Education: Practices, Crosscutting Concepts, and Core Ideas. National Academies Press. https://doi.org/10.17226/13165
5. Kirschner, P. A., Sweller, J., & Clark, R. E. (2006). “Why Minimal Guidance During Instruction Does Not Work.” Educational Psychologist, 41(2), 75–86. https://doi.org/10.1207/s15326985ep4102_1
6. Chi, M. T. H., & Wylie, R. (2014). “The ICAP Framework: Linking Cognitive Engagement to Active Learning Outcomes.” Educational Psychologist, 49(4), 219–243. https://doi.org/10.1080/00461520.2014.965823