Counting every drop: Measuring evaporation in dams
Despite days of rain and flood warnings earlier this year, authorities asked people in Brisbane to conserve water. Wivenhoe Dam, the main water supply for Brisbane, largely missed out on the rainfall that saturated surrounding catchments. It shows the patchy nature of rainfall, even in a La Niña season, and even with nearby flooding.
Decision makers have to make difficult choices about town water supplies. Rainfall is relatively easy to measure. But how about evaporation rates? How do you measure that, for a dam?
Australia is one of the sunniest places on the planet. That means we lose large volumes of the water stored in reservoirs through evaporation, especially in summer.
Dr David McJannet is a research scientist and hydrologist at CSIRO. Since 2012, his team has helped decision makers to understand evaporation rates. And they’ve created a very nifty piece of technology to do just that.

The CSIRO-designed floating evaporation pan measures evaporation rates with expert precision, helping to better manage water resources.
Floating a clever idea: How it all works
It may look like an elaborate pool toy, but the floating evaporation pan is a serious piece of equipment.
At a whopping 6 x 3 metres, the evaporation pan is made up of a PVC pipe frame. The frame encases the equipment and it helps to break up any waves (it’s known as the ‘wave splitter’). The whole structure is anchored to the middle of a dam or lake. Like a dinghy, the structure spins and aligns into the wind.
In the middle of the raft, protected by the frame, is a circular metal pan that floats on the water. It’s free-floating, moving independently, but it’s chained to the frame.
“The evaporation pan uses state-of-the-art instruments, ultrasonic algae control systems, and fully automated filling and emptying routines,” explains David.
Each night, at the stroke of midnight, a small automatic water pump refills the pan to a predetermined level. Then, during the day, the water level drops as the water evaporates.
Inside the pan, magnetic level sensors give a high resolution, highly accurate reading of the water level. Changes in water level show how much water has evaporated each day.
“The floating pan is then left in the water body to provide ongoing daily evaporation numbers. We typically run the system for six to 12 months.”
The floating pan has many advantages over traditional land-based evaporation pans (such as those deployed by the Bureau of Meteorology). A floating pan is immersed in the waterbody, so it can correctly represent the unique wind and energy transfer processes that occur over water. In addition, all of the measurement processes and data reporting are automated.

The floating evaporation pan measures wind speed, direction, temperature, humidity, atmospheric pressure and rainfall.
Not just evaporation
The evaporation pan developed by CSIRO measures wind speed, direction, temperature, humidity, atmospheric pressure and rainfall. Two sensors provide non-contact measurement of water temperature in the pan and in the lake. This enables the scientists to correct for any differences in those temperatures (and the effect that will have on evaporation rates).
“Over the six to 12 months, we develop a computer model that relates lake evaporation to the meteorological variables measured by the floating pan or nearby weather station. The code we develop enables ongoing estimates of daily evaporation.”
Automatic photos are taken every day, and all of this information is sent to CSIRO servers in near real-time by mobile phone telemtry.
But arguably the coolest part of the entire setup is the “ultrasonic algae killer’: it’s a superhero, according to David. It bursts algae cells growing in the pan, allowing the system to stay out on the water for months.

Since 2012, Dr David McJannet and his team have helped decision makers to understand evaporate, with the floating evaporation pan.
Testing the waters outside Darwin
David’s team started doing this work in 2012. Back then, the purpose of the system was to understand evaporation losses from mine pit lakes, water storages and tailings dams for the mining industry.
“We needed the ability to measure evaporation in situ, on the lake’s surface, which is also simpler than other techniques out there. This is a nice, robust piece of equipment that can be deployed for long periods in remote locations.”
Since then, David’s team has expanded this technology to understand evaporation for town water supply and the agricultural industry.
In February 2021, David and a six-strong team set up a floating pan in the Darwin River Dam. This dam provides 85 per cent of Darwin’s water supply, but it’s been at record low levels recently.
“Dam owners like Power and Water Corporation are really interested to see how evaporation rates behave in tropical environments. Evaporation rates in the tropics are very poorly understood; most of the work has been done in southern Australia and Murray Darling Basin.”
The drier the air, the higher the evaporation. Think of the high evaporation somewhere like Mt Isa and Alice Springs. But as you go further north in Australia, the humidity picks up, so evaporation is not quite as high.
“When we’re estimating volumes of dams or lakes, we currently use models from other parts of the world. But we need to know if they’re accurate for Australia. We’re keen to test these numbers in the field, and to see if they’re true.”
Next up, David and his team will install a floating evaporation pan on a large farm irrigation storage dam. This is part of the Roper River Water Resource Assessment, a CSIRO-led initiative to investigate opportunities for water and agriculture development in the Roper River catchment, in the Northern Territory.
Through this project, and many others, David and his team are helping to improve water security in Australia.
May 10, 2021 at 4:21 pm
I can remember the days of a class A evap pan (and even the old “sunken tank”) – automation looks like a good step forward in science to me. Trust your project goes well Dr McJannet.
May 19, 2021 at 10:51 am
How do you correct for the depth variation between the lake and the pan? Also there are edge effects from the pan wall compared to the same space on the lake, ie no walls around a similar space on the lake. I don’t criticize, just want to know. I did a study on small fluvial lakes, where I put the class A pan in the reeds right at the edge of the lake. In that case, the pan walls were shorter than the reeds that surrounded the lake.
May 19, 2021 at 1:08 pm
Hi Natalie, thank you for your interest – great question. Here is David’s response: “If the first part of your question is referring to the surface level variation between water in the lake and the pan, it is only small (~20mm), this does change as the pan evaporates but typical evaporation is ~5mm/d so this is not a big deal. If the first part of your question is more with respect to heat storage differences between the water in the pan and the lake, then we have a correction for this based on concurrent water surface temperature observations (details here: https://doi.org/10.1016/j.jhydrol.2017.10.064). The water surface temperature is a key driver of the evaporation rate at the air-water interface. There will to be edge effects from the pan wall but they will be small. Wind speed at the lake surface is at its lowest in the profile and the wall height is only 80 to 100mm. This is a necessary “observer effect” (disturbance caused by observation) which we unfortunately cannot eliminate.”