Chapter 10 of 17 · Water and the Weather
Chapter 9. Measuring the weather
The quantities that are recorded, the instruments that measure them, the conditions that make a measurement comparable, and how to keep a weather notebook.
Two thermometers, two answers, and only one right one
Put a thermometer in the sun on a window ledge and another in the shade, in the same yard, at the same time. Both are accurate, neither is broken, and yet they do not announce the same thing. Which one gives the temperature of the air? The question looks trivial; it is the whole subject of this chapter.
Measuring is not merely reading a number off an instrument. It is reading a number under conditions such that the number means something to somebody else, elsewhere, later. Without shared conditions, two readings cannot be compared, and a measurement that cannot be compared is not a piece of data: it is an anecdote.
Météo-France states it plainly: "the measurements made by weather services follow standards laid down by the World Meteorological Organization". And the service adds the reason, which is the only one that counts: made under identical conditions everywhere in the world, those observations can then be exchanged and compared. This chapter therefore installs the quantities, the instruments, and above all the conditions.
The objectives of this chapter
Each one is observable. By the end of the chapter, either you do it, or you know which section to read again.
- Name the quantities recorded by a weather station and the instrument that measures each one.
- State the standard conditions for measuring the temperature of the air.
- Explain why a measurement without shared conditions compares with nothing.
- Name the direction of a wind according to the convention of the weather services.
- Keep a weather notebook for several weeks, with one dated entry a day and fixed columns.
What this chapter assumes is already known
Three prerequisites, installed by earlier chapters of this volume: the humidity of the air on the one hand, the genera of cloud and the three levels on the other, since the record calls for their names, and finally the measurement of a depth of rain.
The temperature of the air, and the screen that makes it measurable
Take up the two thermometers again. The one in the sun does not measure the temperature of the air: it measures its own temperature, which depends on the radiation it receives. A black object placed in the same spot would rise higher still. The number read is perfectly accurate for that object, and it says nothing about the air around it.
Météo-France states what the word temperature denotes: "in meteorology, temperature means the temperature of the air when no further detail is given". And it gives the exact conditions: "the measurement is made 1.5 m above the ground, in a white, louvred weather screen, allowing the air to circulate".
Every word of that sentence answers an objection. 1.5 metres above the ground, because temperature varies a great deal with height: the ground warms and cools faster than the air, and a measurement at ground level would always be more extreme. In a screen, to protect the sensor from precipitation and from radiation, from the Sun and from what the ground sends back. White, because white sends light back instead of taking it in. Louvred, because a closed screen would end up warming and would measure the air of its own box.
Remember the general principle behind that list: a condition of measurement is never a formality. Each one repairs a precise way of going wrong, and it is by knowing the way of going wrong that one remembers the condition.
Wind: two parameters, two instruments, one convention
Météo-France defines wind as "the horizontal movement of the air". It is born of differences in pressure: air goes from zones of high pressure, the anticyclones, towards zones of low pressure, the depressions. The closer together the lines of equal pressure, the stronger the wind.
Still according to Météo-France, measuring it involves two parameters, and both are needed: its direction and its speed. Speed is expressed in kilometres per hour, in metres per second or in knots, one knot being 1.852 kilometres per hour.
Two instruments share those two parameters, and this volume uses their everyday names: the wind vane, which shows the direction, and the anemometer, which gives the speed. A wind vane is nothing but a blade free to turn about a vertical axis; it sets itself along the line of the wind. An anemometer is most often a wheel of cups that the wind drives round, and whose turns are counted.
Now comes the convention, and it is the source of a permanent confusion: a wind is named by the direction it COMES FROM, never by the one it is heading towards. A west wind comes from the west and heads east. A north wind comes from the north. The vane therefore points towards the origin of the wind. This convention is the one used by the weather services, and this volume gives it as such: it is not a property of nature, it is an agreement of language, and that is exactly why it has to be learnt instead of guessed.
One last point, often overlooked, explains why two wind readings can differ without anybody being wrong: a wind measurement is always an average over a length of time. Météo-France distinguishes the instantaneous wind, measured over three seconds, from the mean wind, calculated over ten minutes. A gust and a mean wind do not describe the same thing, and comparing them makes no sense.
| Quantity | Instrument | Unit | The condition that makes the measurement comparable |
|---|---|---|---|
| Temperature of the air | Thermometer | Degree Celsius | 1.5 m above the ground, in a white, louvred screen that lets the air circulate. |
| Depth of precipitation | Rain gauge | Millimetre, equal to one litre per square metre | Straight walls, open site away from walls and trees, emptied after every reading. |
| Wind direction | Wind vane | Compass point or degree | Named by where the wind COMES from; away from any obstacle that would deflect the air. |
| Wind speed | Anemometer | Kilometre per hour, metre per second or knot | Always an average over a stated length of time: three seconds for the instantaneous, ten minutes for the mean. |
| Relative humidity | Hygrometer | Percentage | In the same screen as the thermometer, since it depends on the temperature. |
| Atmospheric pressure | Barometer | Hectopascal | Reduced to sea level in order to be comparable, pressure falling with height. |
| Cloud cover | The observer's eye | Share of sky covered, out of eight | From an open spot, at a fixed time, over the whole of the visible sky. |
The words to remember
Four words, of which two describe a pressure situation and two describe a practice of measurement.
- The anticyclone and the depression
- A zone of high pressure and a zone of low pressure. Air goes from the first towards the second, and it is that circulation which makes the wind.
- The weather screen
- The white, louvred box, placed 1.5 metres above the ground, in which temperature and humidity are measured. It protects the sensor from radiation and from precipitation while letting the air circulate.
- A standard condition
- A rule of measurement identical everywhere, laid down by the World Meteorological Organization, and which exists so that an observation made here can be compared with an observation made elsewhere.
- Cloud cover
- The share of the sky covered by cloud, estimated by eye and expressed in eighths. Zero out of eight is a completely clear sky, eight out of eight a completely overcast one.
The trap of this chapter: believing that a more precise instrument makes up for a bad location. That is false, and it is even the other way round. A thermometer reading to a tenth of a degree, placed in full sun, gives an error of several degrees, measured with fine precision; a thermometer reading to the nearest degree, correctly screened, gives the right value to within a degree. The precision of the instrument and the accuracy of the measurement are two different things, and it is the second that counts. Practical consequence for a record kept at home: look after the location and the regularity first, and only then the instrument.
Worked example. Diagnosing a record that does not fit
Question: "A notebook kept for a month gives temperatures consistently four to six degrees higher than those published for the town. The thermometer has been checked in a mixture of water and ice: it does read zero degrees." Where is the problem? Fully guided example.
- Rule out the most tempting hypothesis
The instrument is the first suspect. But the question says it has been checked in a mixture of water and ice and that it reads zero degrees: it is therefore accurate. The error is not in the instrument. - Read the sign of the difference
The difference is consistent and always in the same direction: too warm, never too cold. A random fault would give differences in both directions. A one-way difference points to a permanent cause, therefore to a condition of measurement, not to chance. - Go through the conditions one by one
The conditions are known: 1.5 metres above the ground, white screen, louvred screen. Let us take them one at a time, looking for each one at whether failing to meet it would warm the measurement. - Test each condition
Too close to the ground: the ground warms during the day, so yes, that would warm it. No screen, exposed to the Sun: yes, very strongly. A dark screen instead of a white one: yes. A closed screen instead of a louvred one: yes. Against a south-facing wall: yes, the wall radiates its heat. - Conclude without settling more than the data allow
All the possible causes point the same way, and the question does not say which one is at work. The honest conclusion is therefore: the thermometer is accurate, the location does not meet the standard conditions, and the difference comes from an input of heat by radiation, either direct or reflected. - The result, and what to do next
The record is not false, it is NOT COMPARABLE: it faithfully describes one particular spot, not the temperature of the air of the town. What has to be done is to move the measuring point, to 1.5 metres above the ground, in the shade all day, away from walls, and to start the notebook again from scratch. A badly measured series is not put right after the event.
Experiment 9. Setting up a measuring corner and keeping a notebook
What you need: a thermometer, the rain gauge built in chapter 4, a light ribbon tied to a stick to serve as a wind vane, a notebook, a ruler. SAFETY: this experiment is carried out with an adult's agreement, and the adult is present. It is the adult who chooses the location, who fixes the thermometer and who plants the stick; nobody climbs on anything, nothing is fixed up high, and nobody goes out to take a reading unless the adult says so. A reading can be missed without harm: no measurement is worth going out in dangerous weather.
- Choose the location before anything else
With the adult, look for a spot in the shade all day, away from walls, far from an air vent and from a window, and open above. This is the choice that decides the worth of the whole notebook, and it is made once and for all. - Put the thermometer at the right height
The adult fixes the thermometer about 1.5 metres above the ground, in the shade, so that it receives neither direct Sun nor heat sent back by a wall. If you can build a small white louvred screen around it, so much the better; if you cannot, write that down in the notebook rather than pretending otherwise. - Set up the rain gauge and the wind vane
The rain gauge goes on the ground, in the open, wedged with stones, as in chapter 4. The stick with the ribbon is planted next to it, in the open: the ribbon streams in the direction the wind is blowing TOWARDS, so the opposite direction has to be read in order to name the wind. - Draw the notebook up, once and for all
Make a table with always the same columns: the date, the time, the temperature, the depth of rain since the last reading, the wind direction, the cloud cover out of eight, the dominant cloud genus. A column added along the way breaks the series. - Take a reading at a fixed time, every day
Choose a time and change it no more. A series of thirty readings taken at the same time is worth more than fifty readings taken at any old time. Also note the days on which you took no reading, and why: a gap that is flagged is a piece of data, a gap that is hidden is an error. - Compare with a public source
Once a week, compare your temperature reading with the one published for your district. A difference of one or two degrees is normal. A consistent difference always in the same direction points to a condition of measurement that needs putting right, as in the worked example above. - Read the series after a month
Add up the depths of rain, look for the highest and the lowest temperature, and count the days by dominant wind direction. You will then know things nobody can guess without having measured, and that is what a piece of data is.
Judging a reading, in three moves
- Step 1
Ask first where the measurement was made, and not with which instrument: the location produces far larger errors than the instrument.
- Step 2
Then ask over what length of time it applies, above all for wind: an instantaneous value and an average cannot be compared.
- Step 3
Finally ask whether the series is complete and regular: a series with gaps, or made at varying times, allows no comparison over time.
What to remember from this chapter
- A measurement is worth something only if its conditions are shared: that is what allows observations to be exchanged and compared throughout the world.
- The temperature of the air is measured 1.5 metres above the ground, in a white, louvred screen that lets the air circulate.
- Wind is the horizontal movement of the air; it goes from anticyclones towards depressions and is measured in direction, by a wind vane, and in speed, by an anemometer.
- A wind is named by the direction it comes from, never by the one it is heading towards.
- A wind speed is always an average over a length of time: three seconds for the instantaneous, ten minutes for the mean.
- The accuracy of a measurement depends first on the location, and only then on the precision of the instrument.
Self-check for chapter 9
Answer in writing, using only the knowledge given by this volume so far. The answers are gathered at the end of the volume.
- State the standard conditions for measuring the temperature of the air, and say for each one the error it repairs. Then name the other quantities a station records, with the instrument that measures each one.
- A forecast announces an east wind. Which face of a building does the wind strike, and which one stays sheltered? Justify your answer by the convention.
- One station announces 45 km/h and another 90 km/h at the same moment, in the same place, and neither is wrong. How is that possible?
- Why is a very precise thermometer placed in full sun less useful than a rough thermometer correctly screened?
- You take the temperature every day, but sometimes at seven o'clock and sometimes at three in the afternoon. What can you still conclude from your series, and what can you no longer conclude from it?