Dr Beena Balan Sarojini is a post doctoral researcher at the National Centre for Atmospheric Science and the Geography and Environmental Sciences department at the University of Reading. Her work focuses on how climate change affects rainfall and water-related weather extremes.
To many people, the most familiar sign of climate change is the rise in surface temperature observed over many parts of the Earth since the 1950s.
Scientific evidence shows very clearly that this increase in temperature, known as global warming, is largely down to human activities, such as the emission of greenhouse gases.
But there is another consequence of climbing temperatures. As the air gets warmer it can hold more water vapour, meaning that scientists expect global warming to be accompanied by wetter conditions overall.
Indeed, evidence for a human ‘fingerprint’ on continental-scale rainfall patterns is already emerging. Recent studies show that changes in rainfall over northern mid to high latitude regions and southwest Australia can be attributed to human influence.
However, scientists are yet to confidently detect human-induced changes in local or country-level rainfall. This is because imperfections in climate models and observations are still obscurring the climate change signal in some places, even despite technical advances and a satellite record that spans over more than 30 years.
In our new study, published today in Nature Climate Change, we discuss how the scientific community is making progress towards detecting human-induced change in regional rainfall, even in the face of such challenges.
Why care about changes in rainfall?
Scientists expect rainfall extremes – such as heavy rains, floods and droughts – to happen more often in a warming world, threatening the lives and livelihoods of millions of people.
Hundreds of towns and villages across France and central Europe have been hit by extreme flooding in recent weeks. In Paris alone, the heavy and sustained rain saw the River Seine rise to over 6m, its highest since 1982, leaving thousands of properties without power, causing transport chaos, forcing the closure of the city’s most famous museums and even costing several lives.
Compared to temperature, regional rainfall is naturally far more variable. Rain is influenced both by local thermodynamic factors, such as a warmer ocean, and dynamic factors far away, such as the El Niño Southern Oscillation in the tropical Pacific.
Such is the complexity of rainfall patterns that changes can be caused both by human factors, such as greenhouse gas emissions and atmospheric pollutants, and natural factors, such as changes in the sun’s activity and explosive volcanic eruptions.
An important aspect of climate science is, therefore, to identify regions where rainfall is changing, and to determine how much of the change can be explained by human or natural factors. This is the process of detection and attribution of climate change.
Pinning down how far human activity is contributing to changing rainfall patterns is crucial for adaptation planning, as well as for strategies to reduce future emissions.
The problem of missing data
When it comes to rainfall, the success of detection and attribution depends on how well climate models can simulate the complex physical processes over a given region.
Models used in the most recent Intergovernmental Panel on Climate Change report can “see” in finer detail and are better at representing atmospheric pollutants and their interaction with clouds compared to their predecessors, for example.
But models still have difficulty simulating regional-to-local rainfall in the right location, amount and timing.
Attribution scientists compare model simulations that include both manmade and natural factors against simulations that include only natural factors. Since such analyses require that like is compared with like, rainfall simulations in climate models are often “masked” to match the area covered by available observations. This is why on-the-ground measurements collected continuously and over a long period of time are another must-have.
Our past work has shown that the lack of good observations in many parts of the world can badly distort the detected trend in model simulations, losing much of the underlying signal of anthropogenic change.
So, what can we do in the absence of good data?
Can we afford to wait?
The good news is that both climate models and observations are continuously improving with every advance in science, computing power and satellite technology. Over the longer term, climate change attribution and prediction studies should take advantage of such advances.
But simply waiting for models and observations to get better seems an insufficient response to the climate changes that are already happening.
In our study, we discuss how new ways of identifying changes in regional rainfall are beginning to show some success, even in the presence of imperfect models and measurements.
Growing scientific understanding of the physical processes at play, together with fine-detailed models, are allowing scientists to compare models and observations in a way that can disentangle the thermodynamic and dynamic aspects of the changes. This is a way forward to facilitate detection in the face of uncertainty.
Rather than waiting for models and observations to improve, or for the anthropogenic signal to emerge from the noise of natural variation, the scientific community is starting to be able to provide information to policy makers and societies on the timescales that they need it.
For example, a recent study on the severe flooding in England in the winter of 2013/14 found that as well as increasing the amount of moisture the atmosphere can hold, human-caused warming had slightly increased the number of January days with the type of wind patterns that favour extreme precipitation.Nathalie Schaller explains her research into the role of climate change on the winter floods in England in the winter of 2013/2014. Credit: Fabio Crameri and Deborah Strickland.
Another point we make in our paper is that averaging over many simulations from different models may obscure such signals of human influence, if the underlying mechanisms are not explicitly accounted for. Both the thermodynamic and dynamic aspects of a type of extreme event need to be considered to quantify the risk.
In summary, changes in the frequency of floods and droughts may impact the livelihoods of millions of people over the coming decades. While identifying the human contribution to such changes remains highly challenging – even using state-of-the-art climate models and observations, progress is still possible and, indeed, of critical importance.
Main image: If a raindrop falls on the ground. Credit: Astrid Gast/Getty Images.
Source: Balan Sarojini, B., Stott, P. A. and Black, E. (2016) Detection and attribution of human influence on regional precipitation. Nature Climate Change. doi: 10.1038/NCLIMATE2976. Dr Balan Sarojini’s co-authors on the new paper were Prof Peter Stott, Head of Climate Monitoring and Attribution at the Met Office and Professor of Detection and Attribution at the University of Exeter, and Dr Emily Black, Associate Professor and expert in water cycle variability at NCAS based at the University of Reading. Both helped in the writing of this article.