Ensemble model predictions for Hurricane Matthew over the next 7-10 days. The models all predict the path up the east coast of Florida, with some very minor variation in exact track. After about 5 days, the modeled tracks start to vary, with uncertainty in the hurricane’s path increasing as you move further and further into the future.
It’s early October. Winter is coming. From a scientific perspective, we know why. Given the tilt of the earth in relationship to the sun, the Northern Hemisphere is about to receive far less incoming solar radiation than in the summer months. The obvious result…cooler temps than in summer. The depths of winter may be a few months out, but we KNOW what’s going to happen based on some very basic, easily measured scientific information.
If we KNOW it’s going to get colder several months in advance…why can’t a weatherman tell me if it’s going to rain or snow on Halloween? That’s much closer, after all. If a meteorologist can’t tell me what the weather will be like in 3 weeks, how can they possibly know that winter is going to be colder? Clearly meteorologists and climatologists have no idea what they’re talking about.
THAT is the basic argument that was making the rounds on social media over the last day or two. Climate change skeptics are trying to link uncertainty in hurricane tracks to uncertainty in climate change, stating that if we can’t perfectly predict a hurricane’s track several days in advance, how can we possibly know what the climate will do over the next several decades? Hurricane Matthew wreaked havoc in the Caribbean, and is about to strike Florida. As always, National Weather Service continues to monitor the storm, and issue forecasts on the likely future track. There are uncertainties, of course. Scientists use “ensemble modeling” to try to account for uncertainties in models. Any ONE model may or may not have biases and error, but running many different models helps a scientist to visualize overall patterns and describe the most LIKELY outcome.
A graphic circulating widely on social media, giving a false impression of hurricane forecasting.
For hurricane modeling, a common graphic is a hurricane forecast map that shows individual predictions of many different models. These graphics also typically include an “average” track, created by basically averaging all the different model runs. Typically an ensemble model graph looks like the image at the top, showing where Hurricane Matthew is likely to go over the next 7-10 days. Uncertainty is much lower closer to the present time, so model tracks tend to be close to each other at first, and then become more uncertainty as the prediction period lengthens. In the real Hurricane Matthew example above, the models are all quite consistent in predicting Matthew will hug the Florida coast They all predict Matthew will take a right turn off the coast of South Carolina, Model paths then diverge some, although in updated predictions from the graphic above, models are mostly predicting a strong clockwise turn that may bring the hurricane back to Florida for a 2nd round.
On social media over the last few days, the 2nd graphic has been circulating. It gives a very false impression of hurricane predictions, with many more modeled tracks than there are actual hurricane models, winding all over the map like a bowl of spaghetti. The “punch line” with this graphic on social media? That meteorologists have no idea where a hurricane is going in a few days, and thus they can’t possibly know that climate change is going to occur in the coming decades.
Other than the misrepresentation in the 2nd graphic of real hurricane model uncertainty, this attack on climate science makes a fundamental error in the difference between short-term weather, and longer-term climate. It’s similar to the pathetic attack on climate science by James Inhofe on the U.S. Senate floor, where he brought a snowball onto the floor and thereby declared that since it was snowing, climate change clearly wasn’t occurring. The analogy to the coming winter is quite fitting though. For seasonal change, we KNOW the physical characteristics of the earth/sun system that drive the changes between seasons. For long-term climate change, we KNOW the physical impact of increased greenhouse gases in the atmosphere. Just as seasonal change occurs because of solar radiation differences between seasons, we KNOW the climate is going to warm given greenhouse gas influences on the balance between how much solar radiation is maintained in the earth/atmosphere system, versus how much radiates back out to space.
In effect, we’re putting a blanket on the atmosphere, trapping more heat. It’s a known, physically measurable and quantifiable characteristic of the earth/sun/climate system, just as is the changing of the seasons.Just as it’s much harder to predict short-term variability in weather (including hurricane tracks over the next 7-10 days) versus long-term seasonal trends (hotter in summer, colder in winter), it’s much easier to predict long-term trends in climate, based on how we’re altering the atmosphere.
As a scientist the most frustrating thing about the 2nd graphic and the social media’s false attack on climate science is that it fits a general pattern of “science bashing” in the United States. Be it evolution, climate change, or a host of other KNOWN scientific processes, there’s an odd anti-science pushback that’s grounded more in religion and politics than actual science. It’s not a uniquely American phenomenon, but it certainly is much more amplified and prevalent in the U.S. than in most countries. The politicization of science, the blatant disregard for scientific theory and even real, measurable empirical evidence, turns even something as obvious as evolution or climate change into a faux controversy.
All for the sake of advancing a political or religious agenda.
Don’t fall for the social media bullshit. Scientists and modelers have done a wonderful job tracking and predicting Matthew’s path, giving millions in its potential path time to prepare or evacuate. Weather is weather, and modeling an exact path over a week out is still an inexact (but rapidly improving) science. That uncertainty in NO way relates to our certainty about long-term warming trends in relationship to climate change.