Course Description¶

This is the seed document for the learning graph. It defines the audience, prerequisites, topics, and outcomes that learning-graph-generator uses to enumerate concepts and their dependencies.

Title¶

Raspberry Pi Environmental Monitoring: Grades 6–12

Audience¶

Students in grades 6–12 building a working environmental monitoring station, in a science classroom, an after-school STEM club, or an independent science fair project. The book also serves the teachers and club advisors who guide them, who may be introducing this hardware for the first time themselves.

Readers are assumed to be curious about how measurements are made but new to nearly all of the technology involved. No prior experience with Linux, the command line, electronics, or programming is expected. Everything specific to the Raspberry Pi, the sensors, and Python is taught within the book.

Because the grade band is wide, the material is written to a middle-school reading level and introduces every technical term before using it. Depth comes from the extension activities and the data-analysis chapters, which give older students somewhere to go without leaving younger ones behind.

Prerequisites¶

Readers are assumed to have mastered:

Arithmetic with decimals, ratios, percentages, and unit conversion
Reading and constructing line graphs and scatter plots from tabular data
Introductory physical science: matter, energy, temperature, and states of matter
Basic computer literacy: files and folders, typing, connecting to a Wi-Fi network
Following a written multi-step procedure and recording results in a notebook

Readers are not assumed to know:

Any programming language, including Python
The Linux command line, or what an operating system does
Circuits, voltage, current, soldering, or how to read a datasheet
Statistics beyond the mean and the range
Trigonometry or calculus

Topics¶

What environmental monitoring is, and the questions monitoring data answers
Measurement fundamentals: units, resolution, accuracy versus precision, uncertainty, and calibration
A short history of each measurement: who first measured it, with what instrument, and on what scale
The atmosphere and weather: temperature, barometric pressure, relative humidity, and dew point
Solar radiation: irradiance, the solar constant, insolation, and the ultraviolet index
Wind: speed, direction, gusts, the Beaufort scale, and apparent temperature
Sensing the ground: accelerometers, seismic waves, magnitude, and intensity
Time and place: GPS coordinates, elevation, and why stations record UTC
How sensors work: transduction, and what happens inside a BME280
What each measurement affects: weather and severe-storm forecasting, agriculture and irrigation, energy demand and solar generation, air travel and altimetry, human health and comfort, building and infrastructure design, and hazard warning systems
The Raspberry Pi as a computer, and what Ubuntu Server does
Working at the command line: files, permissions, editors, and services
Connecting hardware safely: GPIO pins, the I2C bus, wiring, and static discipline
Reading a sensor from Python and turning raw values into physical units
Data logging: sampling rate, timestamps, CSV files, and why records need units
Storing and retrieving a growing dataset
Charting time-series data and reading it critically
Data quality: drift, outliers, gaps, and validating a suspicious reading
Getting data off the device: networks, cellular telemetry, and intermittent links
Powering a remote station: batteries, solar panels, and building a power budget
Siting and weatherproofing: enclosures, sensor exposure, and why placement changes the data
Publishing and sharing data, including citizen-science networks
Documenting the build and presenting findings to a non-technical audience

Topics Not Covered¶

Setting boundaries keeps the book at a length students can finish:

Circuit design, soldering, or PCB layout; all connections are made with jumper wires
Analog electronics theory: transistor operation, op-amps, filter design
Chemistry of the atmosphere: ozone depletion, carbon cycles, aerosol chemistry
Numerical weather prediction, atmospheric modeling, or climate model construction
Inferential statistics: hypothesis testing, confidence intervals, regression theory
Machine learning applied to sensor data
Network security, cryptography, or server administration beyond running one service
Seismic wave inversion, earthquake prediction, or fault mechanics
Manufacturing a sensor; the book explains how sensors work but does not build one

A Short History of Each Measurement¶

Every number this station records has a story behind it. Somebody had to decide that a quantity was worth measuring, invent an instrument that responded to it, and agree on a scale so that two people in different places could compare readings. Those three steps — quantity, instrument, scale — took centuries for some measurements and only decades for others.

These histories are seed material for the learning graph and for the opening section of each measurement chapter. They are written at the same reading level as the book itself. The barometric pressure history follows the model set by the original human-written lesson plan, Under Pressure: About One Atmosphere.

Barometric Pressure¶

In ancient Greece, philosophy and science were nearly the same subject. Aristotle, writing in the 4th century BC, concluded that "nature abhors a vacuum" — that empty space could not exist — and that the atmosphere reached up into the heavens. He had no instrument to test the idea, only logic, and the idea stood for almost two thousand years.

In 1643 Evangelista Torricelli, a student of Galileo, filled a sealed glass tube with mercury and upended it in a dish. The mercury fell part way and stopped, leaving a gap at the closed top. The column always settled near 29.9 inches. Torricelli argued that the weight of the air pressing on the dish was holding the column up, and that the gap above it was the vacuum Aristotle said could not exist. Because mercury is about thirteen times denser than water, a column of water doing the same job would have to be more than thirty feet tall — which is why the experiment had waited so long for the right liquid.

Blaise Pascal saw the test that would settle the argument. If air has weight, then less air overhead means a shorter column. In 1648 he asked his brother-in-law, Florin Périer, to carry a barometer up the Puy de Dôme mountain in France and record the height of the mercury along the way. The column dropped steadily with altitude and rose again on the descent. The atmosphere had been weighed. The measurement took the name barometric pressure, from the Greek baros, meaning weight.

Instruments got smaller after that. Lucien Vidi built the first practical aneroid barometer in 1844, using a sealed metal capsule that flexed instead of a tube of mercury, which finally made a barometer portable enough to carry aboard ship. Admiral Robert FitzRoy — the captain of Darwin's Beagle — used networks of these instruments to issue the first public storm warnings in the 1860s, the beginning of weather forecasting as a service.

The unit has moved as well. The United States still reports inches of mercury (inHg), with 29.92 inHg at sea level. Most of the world reports hectopascals (hPa), where sea level is 1013.25 hPa; a hectopascal is the same size as the older millibar, and the pascal itself was named for Blaise Pascal in 1971. All three are descriptions of the same standard pressure, called one atmosphere.

Today the sensing element is a piece of silicon. In the 1960s engineers found that a thin silicon diaphragm changes its electrical resistance when it flexes — the piezoresistive effect — and by the 1980s an entire pressure sensor could be etched onto a chip. The BME280 in this project descends directly from Torricelli's tube: same quantity, same units, four centuries of shrinking.

What it affects. Pressure changes drive the wind and signal approaching weather, so forecasts depend on a dense network of barometers. Aircraft altimeters are barometers, and pilots reset them to the local pressure before landing. Building ventilation, weather-triggered evacuation orders, and the timing of hurricane landfall warnings all rest on pressure readings.

Temperature¶

Warmth was something people could feel long before it was something they could count. Around 1592 Galileo built a thermoscope: a glass bulb with a stem standing in water, in which the water rose and fell as the trapped air expanded and contracted. It showed that something was changing, but it had no numbers on it, and it also responded to air pressure. A thermoscope tells you different. A thermometer tells you how much.

Numbers arrived in stages. Santorio Santorio put a scale on a thermoscope around 1612 and used it on patients. In 1654 Ferdinand II of Tuscany sealed the liquid inside the glass, which removed the pressure error. The problem left was that no two makers agreed on the scale, so no two readings could be compared.

Daniel Gabriel Fahrenheit fixed that in 1724. Working with mercury in glass, he produced the first scale that different instruments could reproduce, fixing 0 °F on a stable ice-water-salt mixture and 32 °F on the freezing point of water, with 180 degrees eventually separating freezing from boiling. Anders Celsius proposed a hundred-step scale in 1742 — originally with 0 at boiling and 100 at freezing, inverted to today's arrangement shortly after his death — and it was later renamed in his honor.

In 1848 William Thomson, Lord Kelvin, argued that there is a coldest possible temperature, the point at which atomic motion stops, and built a scale starting there. That reframed what temperature is: not a substance called heat, but a measure of how fast atoms are moving. The kelvin is the SI unit, and it is why scientists can speak of an absolute amount of thermal energy rather than a position on someone's arbitrary scale.

Electrical thermometers followed. Thomas Seebeck discovered in 1821 that a junction of two metals produces a voltage that depends on temperature, giving us the thermocouple. Platinum resistance thermometers arrived in the 1870s. The sensor inside a BME280 uses a simpler trick still: a silicon diode whose voltage changes predictably with temperature, small enough to sit beside the pressure and humidity elements on the same chip.

One more piece of the history is not an instrument at all. In 1864 Thomas Stevenson designed the louvered white box — the Stevenson screen — that shields a thermometer from sunlight and rain while letting air pass through. Without an agreement about how to expose the thermometer, a worldwide temperature record is not comparable from station to station. Siting turned out to matter as much as the sensor.

What it affects. Temperature sets growing seasons and planting dates, drives heating and cooling demand on the electric grid, determines how much water vapor the air can hold, and expands the bridges and rails of the built environment enough that engineers must design gaps for it. The instrumental record that began in the mid-1800s is the backbone of what we know about a changing climate.

Humidity¶

Water vapor is invisible, so measuring it began with materials that quietly respond to it. Around 1450 Nicholas of Cusa described balancing a ball of wool against stones on a scale: as the air grew damp, the wool absorbed moisture and tipped the balance. Leonardo da Vinci sketched a version of the same idea about thirty years later. Both were measuring a change, not an amount.

In 1783 Horace-Bénédict de Saussure built the hair hygrometer, using the fact that a human hair lengthens by a few percent as it takes on moisture. It was the first humidity instrument good enough to be read as a number. Around 1801 John Dalton supplied the physics behind it, showing that water vapor exerts its own partial pressure in the air independent of the other gases.

That insight separated two ideas students still mix up. Absolute humidity is how much water vapor is actually in the air. Relative humidity is that amount compared to the most the air could hold at its current temperature, written as a percent. Because warm air can hold far more vapor than cold air, the same absolute humidity reads as a low relative humidity on a hot afternoon and a high one at dawn — which is why fog forms overnight without any new water arriving.

The dew point is the third way to say it, and the most useful outdoors: the temperature to which air must cool before its vapor condenses into liquid. It is an absolute measure, it never exceeds the air temperature, and it predicts how muggy a day will feel far better than relative humidity does. An 80 °F day at 50% humidity is more oppressive than a 30 °F day at 100%, because the dew point is far higher. Condensation is also how clouds form, so the dew point is the link between a number on a screen and rain on the ground.

Practical instruments followed the theory. The wet-and-dry-bulb psychrometer — two thermometers, one wrapped in a wet wick, with evaporation cooling the wet one — became standard after 1825 and was refined into Richard Assmann's aspirated psychrometer in 1887. Chilled-mirror hygrometers measure the dew point directly by cooling a mirror until it fogs. The modern breakthrough was the thin-film capacitive sensor, introduced by Vaisala in 1973: a polymer film between two electrodes absorbs and releases water vapor, changing the capacitance in proportion to relative humidity. That is the mechanism inside the BME280.

What it affects. Humidity governs how quickly sweat evaporates, so it sets heat-stress limits for athletes and outdoor workers; it drives mold, corrosion, and wood movement in buildings; it determines irrigation need in agriculture; and the dew point is what forecasters watch for fog, frost, and thunderstorm fuel.

Solar Radiation¶

The measurement begins with an accident. In 1800 William Herschel held a thermometer just past the red end of a prism's spectrum, expecting nothing, and found the temperature rising — invisible infrared radiation. Sunlight, it turned out, carries energy the eye cannot see, and energy is something a thermometer can count.

Claude Pouillet built the first pyrheliometer in 1838 and used it to make the first real estimate of the solar constant: the power arriving on a square meter facing the Sun at the top of the atmosphere. Samuel Langley invented the bolometer in 1880, sensitive enough to map the spectrum in fine detail, and hauled instruments up Mount Whitney to measure how much the atmosphere absorbs on the way down. Anders Ångström's pyrheliometer of 1893 and the Moll-Gorczynski thermopile pyranometer of the 1920s turned solar measurement into routine station work.

Ground measurements could never fully settle the number, because the atmosphere always takes a cut. Satellites did: instruments beginning with Nimbus-7 in 1978 measured total solar irradiance from orbit and pinned the solar constant near 1361 watts per square meter. They also showed it is not quite constant, varying by about 0.1% across the eleven-year sunspot cycle.

A parallel line of work made the Sun measurable with electricity instead of heat. Edmond Becquerel discovered the photovoltaic effect in 1839; Charles Fritts built a selenium photocell in 1883; Bell Labs produced the first practical silicon solar cell in 1954. A photodiode or a small solar cell produces a current proportional to the light falling on it, which is how inexpensive irradiance sensors work today — cheap and fast, though less accurate across the full spectrum than a thermopile pyranometer.

Two derived measures matter for daily life. Insolation is irradiance added up over a day, in watt-hours or kilowatt-hours per square meter, and it is the number that sizes a solar panel. The UV index, developed in Canada in 1992 and now used worldwide, converts ultraviolet irradiance into a sunburn-risk scale.

What it affects. Solar radiation is the engine of nearly everything else this station measures: it heats the ground, drives evaporation, and creates the pressure differences that become wind. It sets crop yields and evapotranspiration, sizes the solar panel and battery that keep this very station alive, determines the output of utility-scale solar farms, drives skin-cancer risk through UV, and forces architects to design shading, glazing, and cooling loads around it.

Wind Speed¶

Wind direction was measured long before wind speed. The Tower of the Winds in Athens, built around 50 BC, carried a bronze weather vane on top and personified the eight winds on its faces. Speed had no instrument at all — only adjectives.

Leon Battista Alberti built the first anemometer around 1450: a flat plate hung so that the wind pushed it up an arc, with the angle indicating the force. Robert Hooke reinvented the swinging-plate design in the 1600s. Both measured pressure against a surface, which is a real physical effect but a hard one to calibrate.

The lasting fix for the calibration problem was social rather than mechanical. In 1805 Francis Beaufort, a Royal Navy officer, wrote a scale that described wind by what it does — how much sail a ship could carry, later how trees and smoke behave — from 0 (calm) to 12 (hurricane). The Royal Navy adopted it in 1838. The Beaufort scale let observers around the world file comparable reports using no instrument at all, and it is still in use.

John Thomas Romney Robinson gave the world the instrument in 1846: the cup anemometer, whose spinning cups turn at a rate proportional to wind speed regardless of direction. It is still the standard on weather stations, usually paired with a vane for direction. The propeller anemometer, the hot-wire anemometer, and — since the mid-20th century — the sonic anemometer, which times ultrasonic pulses between fixed heads and has no moving parts at all, fill in the rest of the field.

Units stayed divided by profession: knots at sea and in aviation, meters per second in science, kilometers or miles per hour on the news. Wind also required its own derived measures. Gust is the peak over a short window against a sustained average, and the difference matters more to a structure than the average does. Wind chill, first published by Paul Siple and Charles Passel in 1945 and rebuilt on modern heat-transfer models in 2001, converts wind and temperature into what the air feels like on skin. The Saffir-Simpson hurricane scale (1971) and the Enhanced Fujita tornado scale (2007) rate storms by wind speed and the damage it does.

What it affects. Wind carries heat, moisture, dust, smoke, and pollen, so it governs air quality, wildfire spread, and pollination. It sets structural loads on buildings, bridges, cranes, and towers — the built environment is designed against a wind speed a region is expected to see. It determines whether a wind farm generates power, whether aircraft can land, and whether a small boat should leave harbor.

Motion: Sensing Earthquakes¶

The oldest instrument in this book is Chinese. Around 138 AD Zhang Heng built a seismoscope: a bronze vessel ringed with dragon heads, each holding a ball over the mouth of a bronze toad. A distant tremor released one ball with an audible clang, reportedly indicating the direction of a quake several hundred kilometers away. It detected and it pointed, but it did not record, and nothing comparable appeared for seventeen centuries.

Luigi Palmieri built an electromagnetic seismograph on the slopes of Vesuvius in 1856 that could record the time of a shock. The modern instrument came from a group of British scientists working in Japan: John Milne, James Ewing, and Thomas Gray built the first true seismograph around 1880, using a heavy mass suspended so that it stays still by inertia while the ground and the frame move beneath it. A pen attached to the mass draws the ground's motion on a moving drum. Every seismometer since, including the chip in a phone, works on that principle — measure the ground against something that resists being moved.

Milne went on to organize the first worldwide network of stations, which made a new kind of science possible: comparing arrival times at many places to locate an earthquake, and reading the waves to infer the structure of the Earth's interior. Seismology began measuring the planet, not just the shaking.

Scales came next, and there are two kinds. Intensity describes shaking at a place from its observed effects — Giuseppe Mercalli's 1902 scale, revised as the Modified Mercalli scale in 1931 — and it differs from town to town for a single quake. Magnitude describes the size of the event itself. Charles Richter, with Beno Gutenberg, published the first magnitude scale in 1935 for Southern California; because it is logarithmic, each whole step is about ten times the ground motion and roughly thirty-two times the energy. Richter's scale saturates for very large events, so since 1979 seismologists have used the moment magnitude scale of Hiroo Kanamori and Thomas Hanks, which is computed from the physical size of the fault slip.

Small, cheap sensing arrived with MEMS accelerometers — silicon chips with a microscopic proof mass on springs, whose displacement is read as a change in capacitance. Developed for automotive airbags around 1991, they now sit in every phone. They are far less sensitive than a research seismometer, but there are billions of them, and dense networks of ordinary devices — the Quake-Catcher Network from 2008, MyShake from 2016, and the ShakeAlert public warning system on the US West Coast — can detect a quake's fast P wave and send an alert seconds before the damaging S wave arrives. That is the class of sensor a student station can afford.

What it affects. Seismic measurements set the building codes that decide how schools, hospitals, and bridges are designed in a given region; they trigger automatic shutdowns of trains, gas lines, and industrial plants; they feed tsunami warnings; they are how nuclear weapons tests are detected worldwide; and they are still the primary evidence for the structure of the Earth's deep interior.

Location: Latitude, Longitude, and GPS¶

A measurement without a place attached is nearly useless — a pressure reading only means something if you know where and how high it was taken. Latitude was the easier half: measure the angle of the Sun at noon or of the pole star, and the answer follows from geometry. Sailors managed it with a quadrant, and later a sextant, for centuries.

Longitude was a clock problem in disguise. The Earth turns 15 degrees of longitude every hour, so knowing the time at a reference place and the local time tells you how far east or west you are — but no clock could keep reference time through years of ship motion, damp, and temperature swings. Britain's Longitude Act of 1714 put a prize on it. John Harrison, a carpenter and self-taught clockmaker, spent decades on the problem and produced H4 in 1761, a marine chronometer that lost only a few seconds on a voyage to Jamaica. Even then, everyone had to agree on which reference place to use: the International Meridian Conference settled on Greenwich in 1884, and the same conference laid the groundwork for standard time zones and a universal day.

The satellite era began with an observation, not a plan. When Sputnik launched in 1957, two researchers at Johns Hopkins, William Guier and George Weiffenbach, tracked it by the Doppler shift of its radio signal and realized they could determine its orbit from a known ground position. Their colleague Frank McClure asked the reverse question: if you know the satellite's orbit, can you find your position on the ground? That became Transit, the first satellite navigation system, operational for the US Navy in 1964.

GPS — the Global Positioning System — was approved in 1973, launched its first satellite in 1978, and reached full operation in 1995. It works by timing: each satellite broadcasts its position and an extremely precise time from an atomic clock, and a receiver that hears four or more of them solves for its own latitude, longitude, altitude, and clock error. Until May 2, 2000, the US degraded the civilian signal deliberately — a policy called Selective Availability — and switching it off improved civilian accuracy overnight from about 100 meters to under 10. GPS is now one of several systems, collectively GNSS, alongside Russia's GLONASS, Europe's Galileo, and China's BeiDou. Positions are reported against the WGS 84 model of the Earth's shape.

A GPS receiver returns three things this station needs. Latitude and longitude say where the readings were taken. Altitude — less accurate than the horizontal fix, because of the satellite geometry — matters because barometric pressure falls with height, and a pressure reading cannot be compared to another station's without it. And time: GPS delivers a clock accurate to well under a microsecond, which is why remote stations timestamp in UTC rather than local time. UTC has no time zones and no daylight-saving jump, so records from stations in different countries can be laid on the same axis without a single ambiguous or duplicated hour.

What it affects. Location data is what turns isolated readings into a map, and maps are what reveal a heat island, a pollution plume, or a storm track. Precise GPS timing synchronizes the electric grid, cell networks, and financial systems; GPS-derived positions guide aircraft, shipping, farm equipment, and emergency response; and continuously recorded GPS positions are themselves a measurement of the ground, tracking the millimeter-per-year drift of tectonic plates and the swelling of volcanoes.

Learning Outcomes¶

The primary goal of this book is for students to understand what is being measured in the natural environment and how those measurements affect the natural and man-made environments. The Raspberry Pi station is the means, not the end: it exists so that students take their own readings, of their own air, in their own place, and come to trust numbers they produced themselves. Building and programming skills are taught in service of that understanding.

After this course, students will be able to:

Remember: Name the seven quantities this station measures — temperature, barometric pressure, relative humidity, solar radiation, wind speed, ground motion, and location — along with the standard unit and typical range of each; recall who first measured each quantity and with what instrument (Torricelli and the mercury barometer, Fahrenheit and Celsius and their scales, de Saussure and the hair hygrometer, Herschel and infrared, Beaufort and his wind scale, Milne and the seismograph, Harrison and the marine chronometer); state what one atmosphere, the solar constant, the dew point, and moment magnitude mean; identify the major components of the station, including the single-board computer, the BME280, the power subsystem, and the telemetry module; state what the I2C bus does.
Understand: Explain in their own words what each quantity physically is — that temperature is the motion of atoms, that pressure is the weight of the air overhead, that relative humidity is a comparison rather than an amount; describe how a sensor converts a physical property into a number a computer can store, using the piezoresistive, capacitive, and diode mechanisms inside the BME280; explain why warm air holds more water vapor and what that means for dew and fog; explain how one measurement drives another, such as solar radiation heating the ground, which raises temperature, which changes pressure, which produces wind; distinguish accuracy from precision; explain why two thermometers a meter apart can legitimately disagree.
Apply: Wire a BME280 to a Raspberry Pi over I2C without damaging either; write and modify a Python program that reads the sensors and appends timestamped UTC readings to a CSV file on a fixed interval; convert between °C, °F, and K, and between inHg, hPa, and atmospheres; calculate dew point from temperature and relative humidity, and apparent temperature from temperature, humidity, and wind; produce a labeled chart of any measured quantity over time.
Analyze: Examine several days of their own readings and separate the daily solar-driven cycle from a passing weather system and from sensor drift; correlate two channels — pressure falling as wind rises, temperature dropping as a cloud crosses the solar sensor — and explain the physical mechanism connecting them; locate outliers and missing intervals in a log file and determine which are instrument faults rather than real events; compare their station's readings against a nearby official station and account for the differences.
Evaluate: Judge which environmental measurement matters most for a given real-world decision — whether to issue a frost warning, size a solar array, schedule an irrigation cycle, close a bridge to high-profile vehicles, or evacuate ahead of a storm — and defend the choice with data; assess whether a proposed solar panel and battery will keep a station running through a week of overcast weather; critique a siting decision by predicting how it biases the measurements; weigh the trade-off between a cheap MEMS accelerometer and a research seismometer for a given purpose; judge the credibility of an environmental claim made in a news article by asking what was measured, how, and where.
Create: Design, build, deploy, and document a working environmental monitoring station that records data unattended and makes it available to others; formulate an investigable question about the local environment, collect the data needed to answer it, and report the finding; write up the design decisions, the data collected, the effects the measurements imply for the surrounding natural and built environment, and the limitations of the results, for a general audience.

Why This Book Matters¶

Environmental data is now the evidence behind decisions students live inside every day: when their school closes for heat, where a solar farm gets built, how high a new bridge deck sits above the river, whether a wildfire warning is issued, how a building is braced against an earthquake. Those decisions are made from numbers that came out of instruments, and the instruments were built by people who had to argue about what was worth measuring in the first place.

A student who has wired a sensor, watched it drift, found a bad reading, and corrected it holds environmental data differently than a student who has only read about it. They know what a measurement costs, what it can honestly support, and where it can mislead. That is the difference between consuming data and understanding it — and it is transferable to any field a student later enters.