Measuring
Rainfall by Chris Skinner (@cloudskinner)
Before I embarked upon my PhD research I had not paid
much attention to how we recorded rainfall. My previous experience, probably like
many people, came from my Primary School that had a small weather station in
the grounds, that consisted of a weather vane for measuring wind direction, and
a raingauge for recording the rainfall. It was nothing more than a small
bucket, which collected the rainfall and you recorded the level from the side
each day. If it was up to 4mm, you would record it as 4mm of rain having fallen
in the last day.
That was it, as far as my knowledge went, and as far as I
assumed it went in regards to recording rainfall for the weather forecasts. I
wasn’t wrong, the Met Office here in the UK do still make extensive use of raingauges
to observe rainfall. I will let Ralph James explain them to you –
However, as I soon learnt, raingauges only measure
rainfall at one stationary point. The little bucket I used at my Primary School
could tell me how much rain fell at the school, but it could not tell me how
much rain fell at my house, or how extensive that rainfall was. To fill in the
gaps, meteorologists use weather radars. Over to Biz Kyte –
Brilliant! There we have it then, measuring rainfall,
easy peasy. You just need a network of thousands of raingauges, enough radar
stations to cover your country and enough highly qualified engineers and
scientists to operate and maintain it all.
You won’t be surprised to hear that these conditions do
not extend to many areas of the world. Sub-Saharan Africa for example has not
had the resources and/or the political will to establish the infrastructure
required for timely, accurate rainfall observations, and this has implications
when trying to forecast floods, crop yields, droughts or water resources.
Obviously, being able to observe rainfall in realtime in this region would be
greatly beneficial, but the installation and maintenance of raingauge and radar
networks is just not currently feasible.
One way is to turn to satellite observations. Satellite
platforms carrying Passive Microwave (PM) sensors are the most accurate for
this role, with the instruments measuring the amount of microwave backscatter
from the Earth’s surface. As droplets of water scatter the microwave signal in
a distinctive way it is possible to directly observe where it is raining and
its relative intensity. But (there’s always a but), PM sensors have to be
placed in Low Earth Orbits (LEO) to operate, and therefore travel over the
planet’s surface, recording snap shots of the rainfall as it goes. To add to
problem, sandy ground scatters microwaves in a similar way to water, making
observations by PM satellites more difficult in arid regions, such as much of
sub-Saharan Africa.
Another way, such as that adopted by the TAMSAT team at the University of
Reading, is to use Thermal InfraRed (TIR) instruments mounted on geostationary
platforms. These satellites orbit at a distance that allows them to orbit at
the same speed as the Earth’s rotation, meaning they always observe the same
area of the planet’s surface - this is known as a geo-stationary orbit. TAMSAT
use a relationship called Cold Cloud Duration (CCD), where it is assumed that
if a cloud is cold enough, it will be raining, and the amount of time a cloud
is below that temperature will let the team calculate the rate of rainfall. It
is an indirect relationship, so it does not directly record the rainfall, but
it does provide an estimate that is accurate enough, and timely enough, to be
useful in forecasting seasonal crop yields or droughts.
Again, there is a but. TAMSAT produces ten-day
observations, useful for the above applications, but not very useful for flood
forecasting, for example, that requires realtime observations at atleast a
daily timestep. It is possible to use the CCD method for this, but the
observations are highly uncertain so require some complex statistics to be
properly used. This has led meteorologists to get creative.
Telecommunications are taking off in sub-Saharan Africa,
with mobile phones spreading fast. Professor Hagit Messer, of the University of Tel Aviv, suggested that
interference of signals sent between antennas by rainfall could be used to
measure the rainfall rate between the antennas. Over a whole network of
telecommunication antennae the spatial spread of rainfall and its intensity
could be built up, evolving over time. This form of rainfall observation could
be used to dramatically improve the spatial and temporal coverage over
sub-Saharan Africa, with little need for additional investment.
And again, there is a but. Whether it is observation by
radar, satellite or telecommunication networks, the instruments can only record
where it is raining, when it is raining, and the relative intensity of the
rainfall. That relative intensity needs calibrating, bringing scientists full
circle back to the humble raingauge. There are raingauges in sub-Saharan
Africa, but not a lot of them. The study area I researched had one gauge per
7,000km^2, enough to cover the whole of the UK with just 27 raingauges, and of
course these weren’t evenly spread, concentrated along rivers and in towns,
leaving large areas relatively uncovered.
They can also be poorly maintained and not all raingauges record all of
the time.
There are some good stories about raingauges in Africa. A
couple I have heard from the TAMSAT team are of one gauge that recorded no
rainfall at night, even during the wet season. When investigated it was found
the locals looking after the gauge were storing inside so it wouldn’t be
stolen. Another gauge was consistently recording a light drizzle – this was
caused by people hanging wet clothes on it to dry. We have similar issues in
the UK, with one organisation who should know better placing a gauge on their
roof next to an air conditioning vent that blew rainfall away from it.
One project that I am excited about is TAHMO. The project team
have the highly ambitious objective of dramatically increasing the raingauge
coverage (as well as coverage of other meteorological instruments), for
sub-Saharan Africa by mass producing a cheap, self-contained weather station
and distributing them to schools. One of the most significant outputs to date was
the creation of a low cost acoustic disdrometer, that uses the vibrations of falling
raindrops to measure rainfall rates and reports the readings automatically
using mobile phone technology. For me, this is the great hope of rainfall
observation for poorly gauged regions and really hope they can pull it off. For
now, I’ll leave you with Rolf Hut discussing TAHMO, acoustic disdrometers and tinkering.