Climate Change ‘Fingerprints’ are Detectable From Daily Global Weather Patterns

A heavily used (but long-disproved) point of argument from climate deniers is the existence of local cold weather in a warming planet. The response of exasperated climate scientists and activists has typically been that climate is the overall trend in local weather events – weather is the immediate atmospheric state in any given area, changing by the minute, while climate is observed on a much longer timescale of years and decades. Therefore, in this view, it has been generally assumed among climate scientists that climate change can only be accurately detected through long-term trends. However, the reality may be more bleak than we expected – a study published this month in Nature Climate Change breaks new ground in this field by seeking to detect signs of climate change on a shorter timescale.

Global weather patterns have a much lower level of variability than local weather. In fact, the study shows that the level of variability in local weather covers up signs of climate change while a warming trend is strongly detected globally. So, the scientists suppose that global weather contains important information related to climate, and their study aims to establish signatures of climate change that could be seen in daily global weather patterns.

The research team, led by Sebastian Sippel at the Institute for Atmospheric and Climate Science in Zurich, applies statistical techniques and computer simulations of climate change to test their hypothesis. First, they define the variables they will work with – annual global mean temperature (AGMT) and earth’s energy imbalance (EEI). Then, they train a computer program to predict AGMT and EEI from computer models of daily surface temperature and humidity under externally forced (unnatural) climate change. Through this step, they establish ‘fingerprints’ of climate change in the model’s predictions of daily weather in a warm climate.

However, they also recognize that these ‘fingerprints’ may be influenced by a long-term global warming trend, meaning that the fingerprints could just reflect that trend, and not the existence of forced climate change. Subtracting the mean global temperature from each daily observation removes the signal of any warming trend. But even with ‘mean-removed’ temperature and humidity readings, the researchers still determine fingerprints in the computer models of forced global warming, outside of any warming trend.

In the above experiments they show that based on the models, forced climate change should exhibit characteristic ‘fingerprints’ in daily weather patterns. Next, the researchers use historical data to see if they can observe these fingerprints in real weather patterns. What they see is striking – signs of climate change are visible in over 50% of days in every single year since 2001, and in every single individual day since March 2012 (if the global trend is removed, this becomes about 96.5% of days since 2012).

This finding means that our climate has been altered past what can be considered natural variability, a reality that makes itself apparent even in day-to-day weather. This is troubling, as almost all model scenarios predict further global warming, and the global community continues to act more or less with a “business-as-usual” approach to climate change. The authors conclude that their work demonstrates confidently that global weather contains important information related to the climate, and will help to further investigate the role of climate change in extreme weather events.