Evaporation and its place in the water cycle
When we can’t ‘see’ it, but need to understand it!
When we think of the water cycle, we usually think of rainfall and surface runoff, since these are the components we can see and experience directly. However, there are two other, less conspicuous components and both are important: infiltration (soaking into the soil to reach groundwater reserves) and evaporation (conversion to water vapour in the atmosphere). In most Australian catchments, evaporation represents a large portion of the total outgoing water. This means that understanding evaporation rates is vital for water management, and this is an important area of research in hydrology.
Evaporation from catchments is more correctly referred to as evapotranspiration, which includes both direct vapourisation of water from wet surfaces and transpiration from plants. When plants take in carbon dioxide for photosynthesis, incidental water loss from inside the plant occurs, which contributes to the overall evapotranspiration rates (interestingly, when carbon dioxide concentrations increase, plants can start to use water more efficiently, causing the transpiration rate to go down slightly — but this is a complex topic for another time).
When researching evapotranspiration rates, we define two types of catchments: water-limited (evapotranspiration is limited by the amount of water available to vapourise — this catchment will be arid) and energy-limited (there’s plenty of water and not enough energy available to vapourise it all— this catchment will be quite wet). Of course, a catchment that is normally water-limited will sometimes be energy-limited (e.g. after a large rainfall event), and vice versa. When a catchment is in a water-limited state, the evapotranspiration rate will be closely related to the amount of rainfall, since most of the rainfall will vapourise. However, when a catchment is in an energy-limited state, we need to understand how much water can be vapourised based on the local atmospheric conditions. This is called the ‘evaporative demand’.
Evaporative demand depends mainly on four variables: temperature, radiation, wind speed and vapour pressure deficit (this is how much more water vapour the air can theoretically hold). Because we often don’t have all of these measurements, it is useful to measure evaporative demand directly. We use ‘pans’ for this purpose: in Australia, we typically have round pans of about 1.2m diameter and 10cm depth. Each morning, the pans are topped up and we see how much water evaporates during the day (obviously we need to account for any rainfall, so there will be a rainfall gauge nearby). The rate of water loss from the pan is called ‘pan evaporation’ and is generally accepted to be the most useful indication we have of evaporative demand.
This is where things get interesting. We have a fairly large network of pans in Australia, many of which have been in place for several decades. This gives us a great dataset to study how pan evaporation rates might be changing in a warming world. In the mid 2000s, this was a hot topic and several studies were released. Interestingly, it seemed that although temperatures in Australia had increased since the 1970s, pan evaporation had decreased! The same was the case in many other regions, including Asia, North America and Africa. This is not an intuitive result because warmer air can hold more water — this means the vapour pressure deficit (remember that from earlier? It’s the amount of additional water vapour the air can hold) should have increased, driving up evaporation rates. There was obviously something strange going on…
Remember the four variables that control evaporative demand? Temperature, radiation, wind speed and vapour pressure deficit are all key drivers. If temperature was going up, but pan evaporation was going down, then radiation, wind speed and/or vapour pressure deficit had to be going down(and exerting a substantial influence!).
It turns out that there are credible theories to support each of these possibilities. Incoming short-wave radiation (from the sun) did decrease from the 1970s through to the 1990s because of a phenomenon called ‘global dimming’. In a nutshell, this is because all the air pollution we generate causes more clouds to form, and these clouds reflect incoming short-wave radiation. Wind speed has also been shown to be decreasing around the world, which we call ‘global stilling’ (can you see the pattern in the nicknames?). We don’t really know why this is happening, but we’ve observed it in wind speed measurements. There is also a theory that increased evapotranspiration from the land surface could increase the water vapour present in the air (called ‘global moistening’… Just kidding, no nickname this time), which could decrease the vapour pressure deficit. This means that the air can hold less additional moisture, e.g. from a pan. Also, a lot of energy could be used for evaporation from the land surface, leaving less energy available to specifically evaporate pan water. The overall effect is that pan evaporation decreases as land surface evapotranspiration increases (paradoxical I know). A good way to think about this is that, in an arid area with dry air and very little evaporation from the surrounding landscape, you would expect a lot of evaporation from a pan. If this area received more rainfall and more evaporation occurred from the landscape, it would be cooler and the air would be wetter, so pan evaporation would slow down.
So we have three good theories for decreasing pan evaporation, but how do we know which is relevant for Australia between the 1970s and mid-2000s?
Canberra has led the way in solving this problem (see references at the end of this article). Researchers from the Australian National University and CSIRO took an existing formula that uses temperature, radiation, wind speed and vapour pressure deficit to calculate evaporative demand, and adjusted it to calculate pan evaporation by accounting for the extra radiation that you get through the sides of the pan. In a later study, they used this formula to separate out the effects of the different climatic drivers of pan evaporation.
This showed that, while global dimming seemed to have had an impact in some locations, the main reason that Australian pan evaporation was decreasing was reductions in wind speed.
Understanding these trends is really important for us as water managers because it means that, if we can work out why wind speed was decreasing and whether it’s likely to continue, we can make educated guesses about what might happen to pan evaporation rates in the future. This information, along with results from complex environmental models, can help us to manage urban water supply, irrigation demand and environmental water needs. Also, if you’ve read my first and second articles, you might remember that the wetness/dryness of a catchment (antecedent conditions) preceding a rainfall event has a big impact on whether we get a flood. The evaporation rate from a catchment will affect how wet the conditions are, so this is important for flood risk management too.
Key references:
Roderick, M. L., Rotstayn, L. D., Farquhar, G. D. & Hobbins, M. T. (2007) On the attribution of changing pan evaporation. Geophysical Research Letters, 34(17), n/a-n/a.
Rotstayn, L. D., Roderick, M. L. & Farquhar, G. D. (2006) A simple pan-evaporation model for analysis of climate simulations: Evaluation over Australia. Geophysical Research Letters, 33(17), n/a-n/a.
