Water’s climate risks show high temperature sensitivity

Like having a fleet of miniature research vessels, the global flotilla of more than 3,000 robotic profiling floats provides crucial information on upper layers of the world’s ocean currents. Credit: CSIRO/Alicia Navidad

Climate change is strengthening the cycle of rain falling and evaporating by twice as much as models predict. That’s what data collected by Susan Wijffels from the Commonwealth Scientific and Industrial Research Organisation (CSIRO) in Hobart, Australia, and her co-workers suggest. They’ve studied how rainfall has changed the sea’s surface salinity – or ‘saltiness’ – over the past 50 years, and compared that against current climate models. Wet areas have become wetter and dry areas drier as the water cycle strengthened by 4 per cent over this time. The scientists’ findings suggest the cycle will get 8 per cent stronger for every 1°C warming at the world’s surface, with 2-3°C warming expected by the end of the century. “As a mother I’m hoping that we’re wrong, but if this high sensitivity holds up then it could potentially be quite a significant impact to the future,” Susan told Simple Climate.

Climate change strengthens the water cycle because a warmer atmosphere can
hold and transport more moisture as water vapour. Scientists had expected this to enhance worldwide patterns of rainfall and evaporation. But understanding how rainfall is changing across the world is hard, as around four-fifths of it falls on oceans, where we have little monitoring equipment. So together with Paul Durack, who is currently at the Lawrence Livermore National Laboratory in Livermore, California, Susan had previously used salinity to get round this. As oceans lose water to the atmosphere, the sea gets saltier, and as it then returns to the sea as rain or snow it freshens again. Using salinity measurements reaching back 60 years, including from the Argo network of floating data recorders since 1995, Susan and Paul showed that the water cycle had strengthened. But they didn’t show by exactly how much, even though they wanted to. One other group of scientists had tried, but they had assumed they knew how salinity and the water cycle are linked, something Susan and Paul wanted to avoid. “We actually asked the question: How is salinity and the water cycle related and how is it related to the observed increase in temperature?”

Tell-tale fingerprints

Surface salinity changes for 1950 to 2000. Red indicates regions becoming saltier, and blue regions becoming fresher. Credit: Paul Durack

To test what the link between salinity, the water cycle and temperature was, they teamed up with fellow CSIRO researcher Richard Matear. Together, they looked at how well the fingerprint of salinity changes for 1950-2000 produced by a large group of climate models matched real-life measurements. In particular, they wanted to know whether they matched how much the salinity patterns had strengthened, or amplified. The first experiments simulated this period by running the models without any of the kind of “radiative forcing” that greenhouse gases produce to drive global warming. This gave the researchers a baseline, or control experiment, much like the world was before humans begun burning fossil fuels to compare against. “The first thing that we looked for was: Did they simulate the changes we see in the observations?” Susan explained. “You don’t obviously see that.”

That story began to change when they looked at models set up to reproduce what’s happened to the world’s climate through the 20^th century. Fingerprint matches started to emerge in those results, published in top research journal Science last week – but only by the models that showed significant warming. And there is a clear link between how strong the warming is and how well the fingerprints match. But the match between model and real salinity fingerprints was poor overall, Susan said. “While the models accurately capture the large-scale patterns of salinity changes, the amplification of these changes is underestimated,” she underlined.

But there is a strong link between how much the water cycle and the salinity patterns had strengthened, supporting the idea of using salinity to track rainfall changes. “The surface salinity pattern is a sensitive fingerprint, an indicator, of what the water cycle’s doing,” Susan commented. In fact, the models’ shortcomings meant that in the ones with the right kind of fingerprint pattern salinity suggested even greater strengthening of the water cycle than their rain and evaporation predictions did. After considering how well the models’ fingerprints matched the real data, the best-fit of salinity pattern amplifications said that the water cycle is speeding up 8 per cent for each 1°C warming. That compares with just 2-5 per cent when considering the models’ evaporation and rainfall figures. “The implication is that our climate models might be underprojecting how the hydrological cycle is going to move in the future,” Susan said.

“More important than warming alone”

A stronger global water cycle will affect rainfall and in turn food availability, stability and access, a “more important risk to human societies and ecosystems than warming alone”. The power of the local water cycle is often obvious, for example in the UK last month, which saw the highest monthly rainfall on record for April while much of the country was still technically in a drought. Events like this will be more likely with a stronger cycle, with the potential to become even more severe. “It’s associated with the fact that a warmer atmosphere can shuttle and carry around more water and energy,” Susan said. “Flood periods will be more intense and drought periods will be hotter, drier, leading to more water deprivation during those periods.”

For us to prepare for this, models must accurately tell us how much drier dry places, and how much wetter wet places, will become. Consequently, Susan and her colleagues will continue looking at how well models simulate salinity patterns. In particular, they’ll be looking at the models to be used in the upcoming Intergovernmental Panel on Climate Change (IPCC) fifth assessment report, due out in 2014. She hopes that her team’s work will trigger people to check their findings. “I’m hoping that our paper will spur people to look harder at the model data,” she said. “Maybe some people will repeat our calculations, they’ll go and look at the models in a different way, to confirm that this high sensitivity is correct.”