A more intense global water cycle
Been meaning to post on Paul Durack, Susan Wijfells and Richard Matear’s work on the intensification of the global water cycle using changing ocean salinity, but Paul has written a great article for The Conversation (reproduced below). Paul and I used to give each other grief when we were both at CSIRO, so for light relief he went and did a Ph D, doing this great work in the process.
The work cracked Science magazine (full article behind paywall) and has been featured on Real Climate. It has also attracted a rejoinder in correspondence by Roderick and colleagues who maintain that the evidence of an intensified rainfall response on land is not there (all of which is behind a paywall). I reckon they’re wrong and there is growing evidence that the models are understating hydrological sensitivity. This means that droughts and floods are changing faster than projected by the models. Furthermore, I think these changes are strongly non linear as has been observed in south-eastern Australia – something that Paul is a bit dubious about (for the moment!). Anyhow, from the man himself, read on …
Are the world’s wet regions becoming wetter and dry regions becoming drier?
By Paul J Durack, Lawrence Livermore National Laboratory
Surprising evidence from the oceans suggests they are responding to warming at a faster rate than we previously thought. These changes are expressed by patterns of freshening and enhanced salinity in the ocean surface layer.
How does an ocean get fresher or saltier? If there is more rain than evaporation in an area, it’ll get fresher (less saline); less rain and more evaporation leads to a saltier ocean. These estimates of long-term ocean surface salinity suggest that the global water cycle (which is comprised of all rainfall and evaporation fluxes) has intensified over the 1950-2000 period by 4% – around twice the rate that global climate models have suggested.
These might not sound like large numbers, but it turns out that, over time, they represent significant changes to the predominant rainfall patterns in any given area. A difference of a few percentage points in the intensity of the regional water cycle can mean the difference between being classified as one climatological zone or another.
The spatial patterns of change are telling. Fresh ocean regions are getting fresher (wet regions are getting wetter), and salty ocean regions are getting saltier (dry regions in the subtropics, the same climatological zone of southern Australia, are becoming drier). These new observations agree well with our understanding of how the climate system will respond to a warming Earth. Over land, we measure change with rainfall gauges. Over 71% of the globe (which is ocean), it’s a little more difficult. This is where long-term salinity measurements become useful.
Why consider the ocean?
So why should we care about changes to the global oceans anyway? It’s not as if we live there, are dependent on rainfall to grow crops there (some island nations excluded) or really care how salty or fresh parts of the ocean are. Well the answer to that question is pretty simple.
The oceans are the flywheel of climate, with the top “skin” of the global ocean – down to a little over 3m – able to store the same amount of heat as the entire atmosphere. The depth of the ocean actively involved in climate is far deeper than this. As a consequence, what happens in the ocean, and the way that the ocean changes over time, matters to climate. Indeed the heat that the ocean “stores” and then “releases” at a later stage will markedly influence the weather that we experience day-to-day, and the climate that we experience decade-to-decade and over even longer timescales.
The ocean covers 71% of the Earth’s surface and holds 97% of the Earth’s free water (it’s where the water is..). It has absorbed 90% of the excess heat over the last 50 or so years in response to warming and is currently soaking up 50% of our anthropogenic carbon dioxide emissions.
Multiple assessments of long-term ocean temperatures have independently come to the conclusion that the oceans, both at the surface and subsurface, have been constantly warming throughout the period of observational coverage. As these numerous studies have checked and re-checked the numbers, inconsistencies have uncovered biases with the data. But once these biases have been corrected (using many different techniques, generated from many independent groups), yep, it’s still been warming at an even more steady and faster pace than the previous estimates suggested.
Salinity is a less-sampled ocean variable. There are fewer observational data points, as historically salinity has only been measured by oceanographic research cruises. Since around 2000 however, it has also been measured by the new fleet of automated Argo profiling floats.
These robots are cool! They’re carried around by currents in the global ocean and profile from the near-surface to 2000m depth over a 10-day cycle. They measure temperature and salinity over this cycle, and when they reach the surface, they upload this data to satellites, which then relay on to data centres around the globe. Thanks to these research-quality observing platforms, we’re confident of salinity measurements and are unaware of platform biases to date.
Salinity is linked to the global water cycle by the surface ocean fluxes of evaporation and rainfall. Australians are very water-aware; the recent Australian flooding and our memories of long-term droughts are a testament to that. The development and success of global communities is intimately related to water availability, with water scarcity limiting growth and development. So how can we use long-term salinity measurements, and estimates of its long-term change to assess changes to the global water cycle over the same time period?
Importantly, ocean salinity provides an independent line of evidence we can use to test our understanding of long-term climate and its change. The ocean salinity field responds to changes in both evaporation and rainfall at the ocean surface, not one or the other independently. So it is a good indicator of long-term, whole system water cycle change. Ocean salinity is a fairly stable, reliable way to measure how much water goes up and comes down; and provides a low-pass filter to the noisy spatial and temporal patterns of rainfall.
So what do these new long-term estimates of salinity change tell us? The global water cycle and consequently our climate is changing. The pattern of “fresh-getting-fresher, salty-getting-saltier” is quite striking, is replicated in independent analyses, and is also captured in the subsurface patterns of salinity change. This pattern of change is even there when we sample each global ocean basin – the Pacific, the Atlantic or the Indian Ocean – independently. Changing ocean salinity is really trying to tell us something.
So how does changing salinity link back to rainfall and evaporation? The best tools scientists currently have to investigate relationships between observed changes in rainfall, evaporation and ocean salinity are coupled global climate models. These systems simulate both the atmosphere and the ocean over time.
The good news: when we investigated the response of global climate models to enhanced carbon dioxide forcing, we discovered that the models were capturing the processes of change, responding in the same way, with the same modelled broad-scale patterns of salinity intensification as the observational estimates. They are performing well at replicating our observed Earth system.
The bad news: the rates of change of the modelled salinity fields are around half that of the observed estimates for 1950-2000 – a cause for concern.
So the rate of change suggested by the models was worrying enough. If ocean salinity (and the Earth’s water cycle) is changing faster, and is more sensitive than we simulate in models, it isn’t really good news for projections of future 21st century climate.
Which data source is right?
Whenever our observed world and climate models disagree, the tendency among non-experts is to blame the models. Surely there is something missing in the physics that represents the modelled climate system?
However among scientists, the approach is more cautious. Sometimes the models tell us that there is a problem with the observations. Could an error in the analysis of the observational data mean that we are systematically overestimating the observed water cycle changes?
Patterns of salinity change agree with independent satellite estimates of both rainfall and evaporation changes over tropical regions since 1980. These studies also say that the rates of change captured by models are conservative, when compared to observations. Even if we exclude the great new ocean observational coverage of the Argo robotic fleet of profiling floats, those patterns of salty getting saltier and fresh getting fresher are still there. However, the patterns are even more clear in the updated analyses.
If we take the new estimate of water cycle amplification – 4% over 1950 to 2000 where we have experienced a 0.5°C Earth surface warming – it is possible that a big change of around 20% could occur in response to a 2° to 3°C warmer world.
That could mean really big changes in rainfall and drought by the end of the 21st century. Scientists hope that this is not the case, but it shows that uncertainty is a double-edged sword when it comes to climate change, and not the refuge from risk that some would want us to believe.
Paul J Durack does not work for, consult to, own shares in or receive funding from any company or organisation that would benefit from this article, and has no relevant affiliations.