Time for the Annual Weather Station Clean — When 0.4 mm of Rain Clearly Wasn't Right

A day or so ago, we had some very heavy rain.

Looking outside, there was no question about it. Water was running from roofs, the ground was thoroughly wet and everything suggested that a substantial amount of rain had fallen.

My Davis weather station, however, had a rather different opinion.

It recorded just 0.4 mm.

That immediately raised a question.

Was the rainfall really much more localised than I thought? Was there a problem with the electronics? Had the battery begun to fail? Or was something much simpler going on?

The answer turned out to be wonderfully ordinary.

The rain gauge was blocked with debris.

A quick clean and rinse later, and the tipping rain gauge was operating normally again.

It was also a useful reminder that even a very good automatic weather station still needs somebody to go outside occasionally and look at it.

The Data Was Telling Me Something Was Wrong

One of the advantages of having your own weather station is that you gradually develop a feel for the numbers it produces.

You begin to know roughly what certain weather conditions look like in the data.

A brief shower might produce a fraction of a millimetre.

A period of steady rain might produce several millimetres.

A prolonged heavy downpour should certainly produce rather more than 0.4 mm.

That discrepancy was therefore useful information in itself.

The problem was not initially obvious from looking at the weather station display. The system was still communicating. Temperature measurements looked sensible. Humidity appeared normal. Wind measurements were being reported.

Nothing was flashing an enormous warning saying:

RAIN GAUGE BLOCKED.

Instead, the clue was simply that one measurement did not agree with what I could see happening outside.

That is an important principle in science:

Never look only at whether an instrument produces a number. Ask whether the number is reasonable.

An instrument can be working electrically and still produce misleading data.

How a Davis Rain Gauge Measures Rainfall

The rain gauge in a Davis weather station is actually a beautifully simple device.

Rain falls into a collector at the top of the station. It is then channelled down into a small measuring mechanism.

Depending on the particular Davis design, the mechanism uses a tipping bucket or tipping spoon arrangement.

A small quantity of water collects on one side.

When enough water has accumulated, the mechanism tips.

That tip is detected electronically.

The other side then begins collecting water.

Each tip corresponds to a known amount of rainfall.

The station therefore does not need to weigh the rain or continuously measure the depth of water.

It effectively counts small, accurately defined quantities of water.

Add all those small quantities together and you obtain the total rainfall.

It is simple, reliable and capable of operating automatically for long periods.

But there is one obvious weakness.

The water has to reach the measuring mechanism.

If the entrance becomes blocked, the electronics may be perfectly healthy while the recorded rainfall becomes completely wrong.

Leaves, Seeds, Dust and Insects Can All Cause Problems

Weather stations live outside.

That sounds obvious, but it means they have to deal with considerably more than weather.

Over the course of a year, the rain collector can accumulate:

• small leaves;

• fragments of vegetation;

• blossom;

• seeds;

• dust;

• dirt;

• moss;

• pollen;

• spider webs;

• dead insects;

• insect nests;

• bird debris.

Some of this material may simply sit harmlessly in the collector.

Eventually, however, enough material can gather around the outlet to restrict the flow of water.

That appears to be what had happened with mine.

Heavy rain was falling into the collector, but it was not reaching the measuring mechanism correctly.

The result was an apparently precise but completely misleading measurement.

0.4 mm.

That little decimal place can make the figure look very scientific.

Precision, however, is not the same thing as accuracy.

A Quick Clean Solved the Problem

Fortunately, this was not a complicated repair.

I removed the accumulated debris and gave the rain collector and mechanism a gentle rinse and clean.

Once the obstruction was removed, water could again flow properly through the gauge.

I also checked that the tipping mechanism moved freely.

Everything was then reassembled and tested.

The rain gauge was working normally again.

No replacement sensor.

No new electronics.

No expensive repair.

Just a few minutes of maintenance.

It is precisely the sort of job that is easy to forget because modern weather stations normally operate so reliably in the background.

While I Was There, I Checked the Rest of the Station

Once you have climbed up or otherwise gained access to a weather station, it makes sense to check more than just the fault that brought you there.

In my case, the battery was still fine and the rest of the system appeared to be operating correctly.

That is reassuring, but it is also why I think an occasional physical inspection is worthwhile.

A typical annual weather-station check might include the following.

Check the rain collector

Remove leaves, insects and other material.

Make sure that water can enter the funnel freely and drain into the measuring mechanism.

Check the tipping mechanism

Make sure it can move freely.

There should not be dirt, cobwebs or debris physically preventing it from tipping.

Inspect the solar panel

Many automatic weather stations use a small solar panel to maintain their power system.

A layer of dirt, bird droppings or algae reduces the amount of light reaching it.

A gentle clean may therefore be worthwhile.

Check the backup battery

The battery may last for years, but it should not simply be assumed to be good forever.

If the station reports battery condition, check it.

If the battery is approaching the end of its expected life, replacing it before winter may be easier than waiting for the station to fail during the worst weather of the year.

Inspect the temperature and humidity sensor housing

Temperature sensors are normally protected from direct sunlight by a radiation shield.

Check that it has not become excessively dirty or blocked.

Air still needs to circulate properly.

Look at the anemometer

Make sure the wind cups rotate freely.

Check that the wind vane moves properly.

A spider web wrapped around a rotating component can sometimes have a surprisingly large effect.

Check the mounting

Weather stations experience wind, rain, frost, heat and constant vibration.

Check brackets, poles and fixings.

The station should still be level and securely mounted.

This is particularly important for a tipping rain gauge because the measuring mechanism assumes that the station is correctly positioned.

Calibration Is Not Much Use If the Funnel Is Blocked

Weather-station enthusiasts quite rightly discuss calibration.

We might compare temperature sensors.

We might check atmospheric pressure against a nearby reference station.

We might question whether the rain gauge calibration is correct.

But there is a stage before calibration that is even more fundamental.

Is the instrument physically capable of making the measurement?

A perfectly calibrated rain gauge with a blocked inlet is still a useless rain gauge.

The same principle applies to many scientific instruments.

A laboratory balance may be extremely accurate, but not if something is touching the weighing pan.

A thermometer may be well calibrated, but not if it is sitting in direct sunlight when you are trying to measure air temperature.

A light sensor may be excellent, but not if its window is covered in dust.

Good measurements require good instruments, but they also require good experimental practice.

Why Home Weather Data Is More Interesting Than Simply Looking at an App

Somebody might reasonably ask why anyone needs a personal weather station when weather information is available instantly on a phone.

For me, the answer is that they are doing slightly different jobs.

A weather app tells me the weather for an area.

My weather station tells me what is happening here.

That difference can become surprisingly interesting.

It allows me to observe:

• the actual temperature around my home;

• the highest and lowest temperature during the day;

• wind speed and gusts;

• humidity;

• atmospheric pressure;

• rainfall;

• rainfall rate;

• changes over time.

Over months and years, that becomes a local environmental record.

You can compare storms.

You can look at heatwaves.

You can see how quickly pressure fell before bad weather arrived.

You can compare wet and dry months.

You can even relate the data to gardening, solar energy production, heating demand or local environmental conditions.

But that long-term dataset becomes valuable only if we can trust the measurements.

A blocked rain gauge demonstrates how easily a simple mechanical problem can suddenly introduce bad data into an otherwise excellent record.

The Best Fault Detector May Be Common Sense

There is another lesson here that is particularly relevant when teaching science.

Instrumentation is becoming increasingly automated.

Sensors take readings.

Computers store them.

Software draws graphs.

Cloud services analyse trends.

All of this is extremely useful.

But automation should not replace judgement.

If the computer says only 0.4 mm of rain fell while you have just watched torrential rain bouncing off the patio, perhaps the correct response is not:

"The computer must be right."

Perhaps the correct response is:

"Why doesn't the measurement agree with reality?"

That question is at the heart of experimental science.

Unexpected results are not simply inconvenient.

Sometimes they are telling you that your hypothesis is wrong.

Sometimes they reveal something interesting.

And sometimes they are telling you to clean the rain gauge.

An Annual Five-Minute Weather Station Inspection

I will certainly be adding the rain gauge inspection to the list of jobs worth doing regularly.

Ideally, I would not wait for an obviously incorrect rainfall reading before checking it.

A useful routine might be:

Spring: remove winter debris and check the moving parts.

Summer: inspect for insects, spiders and dry vegetation.

Autumn: watch particularly carefully for falling leaves and seeds.

Winter: check the battery condition and make sure the station remains securely mounted.

The frequency will depend very much on where the station is installed.

A station surrounded by trees may need considerably more frequent attention than one mounted in a more open location.

The important point is simply not to forget that it is there.

A Tiny Maintenance Job That Protects Years of Data

Modern automatic weather stations are remarkably capable devices.

They sit outside through heat, frost, wind and rain, quietly measuring the atmosphere day after day.

Most of the time, they require remarkably little attention.

That reliability can almost become a disadvantage because it encourages us to forget about them.

My suspicious 0.4 mm rainfall reading was a useful reminder.

The battery was fine.

The sensors were fine.

The electronics were fine.

The rain simply could not get through the debris in the collector.

A quick rinse and clean restored everything to normal.

So, if you own a weather station, perhaps today is a good time to go outside and have a look at it.

You may discover nothing wrong at all.

And that is an excellent result.

But if the next thunderstorm arrives, it is rather nice to know that the rainfall entering the top of the gauge is actually going to reach the instrument underneath.

Sometimes maintaining a sophisticated scientific instrument really does come down to removing a few leaves.