Are my measurements right?
Once you've decided what you're going to measure as part of the
Weather Labs, you need to make sure your measurements are
How to get valid measurements with your weather
Experiments do not always give valid results. Errors in
measuring weather conditions might be caused by local winds and
draughts, shade, inaccurate readings, faulty devices and much more.
Get it wrong and no-one will trust your weather forecast again! So
you need to be sure that your weather readings are as good as you
can make them.
Weather scientists at Manchester University have recorded two
special podcasts for Weather Labs all about this. Fiddly
Forecasts discusses how weather forecasts work and what can
make them very wrong. In Murky
Measurements, the weather scientists give you their top tips
for making accurate measurements.
There is a detailed guide to all the issues and problems with
measuring weather here: http://metlink.org/pdf/resources/simweameasurements.pdf
Make sure your measurements are
Here are four key Weather Labs questions for using instruments
to get the most accurate data readings possible, for whatever
device you are using.
- What size of team should you use to make, place and read your
instrument and record its data?Think about why scientists work in
teams and why it is beneficial.
- What instruments will you use? Would it help to make several
versions of the same instrument? Would the instrument be more
useful if each one was identical or slightly different? How can you
be sure that readings from different instruments are valid?
- How many times should you carry out the experiment? Would it
help to repeat the experiment several times? What should be
different or similar between each run of the experiment? Think
about the location, timing and size of team.
- If you collect data from several instruments, or from several
repeats of the experiment, how can you turn all the different data
into one final figure? Should you consider every result even if one
seems wildly different from the rest? How helpful is an average?
Would a mode be more useful? Ask a maths or science teacher for
Have a look at Get the method straight for
more suggestions on how to get accurate measurements from your