The Conservation Conversation: Understanding Noisy Data

In my previous series of columns, I described a 15-year effort to reduce nitrogen discharges from sewage treatment plants into the Long Island Sound. 

The purpose of the multi-billion dollar program is to reduce or eliminate disastrous summertime "dead zones" in the Sound. I ended thecolumn with the question: Is it working? Are the dead zones getting smaller and shorter? 

I wrote that the answer is most likely 'yes,' but that it's notoriously difficult to say for certain. An ecosystem like the Long Island Sound has enormous natural variation from year-to-year and day-to-day in conditions of varying temperature and wind. Weather plays an important role in determining when and where dead zones appear and how long they persist. 

In any given year this natural variation in weather conditions can easily overshadow the beneficial impact of the expensive nitrogen reductions that have been implemented during the past 15 years. For us to be able to say with confidence that the nitrogen reductions are working and are, therefore, worth the billions they cost, will take more time - perhaps, another decade or more. 

The problem is a phenomenon called "noisy data" - that is, data from the natural world where short-term variability is so great that it masks long-term trends. 

The noisy data problem is frequently encountered when trying to assess environmental improvements or declines. You generally have to wait two or three or more decades for a true trend to become apparent amidst the year-to-year fluctuations produced by annual variability. 

It is for this reason that no responsible scientist will say that a given hot day, week, month or year is due to global warming. We have had incredibly hot days and years sprinkled across the century and a half for which good records exist, so any given hot day or year may be just another similar event. 

But virtually every responsible scientist around the world agrees that the earth is in the midst of a significant warming trend. There is no longer any real doubt that the long-term trend is towards a much warmer global climate. This is readily apparent when you plot annual average temperatures on a graph: the points jump up and down from year to year. But if you connect the dots, the overall direction of the zig-zag line is clearly and strongly upward and the upward trend is most evident over the past 30 years. 

By the way, it's now official: 2012 was earth's warmest year on record and seven of the top 10 warmest years in the US have occurred since 1998 (the second warmest year on record).

This may be hard to believe when we've just come through the coldest week of the past two years with daytime temperatures not making it out of the teens. But that's the point about noisy data: there's always lots of short term variability, making it difficult to see the long-term trend. Global warming doesn't mean we'll never have any cold weather. It means there will be more frequent and longer heat waves, winters will on average be warmer than in the past and summer weather will arrive earlier and stay later.

Global warming also means we'll have more frequent severe weather events such as hurricanes and droughts. Because of natural variability, no responsible scientist can say that any given big storm like Sandy or Katrina is attributable to global warming. 

But almost all responsible scientists agree that global warming will contribute to increased frequency and intensity of storms. That's why events that we consider to be "100-year storms" - that is, a storm so big you would statistically predict it would only happen about once every century - now are occurring about once every 20 or 25 years. 

Skeptics like to point out that there have been global warming and cooling trends throughout time - trends with which human beings had nothing to do. This is true. For most of the past 110,000 years, the earth was in one of its periodic ice ages with warmer and colder periods interspersed along the way.

A mere 15,000 years ago, mile-thick ice sheets extended from northern Canada as far south as New York. More recently, a several-century long warmer period in the Middle Ages (approximately AD 950-1250) encouraged the Vikings to settle Iceland, Greenland and North America. The ensuing "Little Ice Age" (about 1350-1850) didn't result in continental glaciers, but the significantly colder climate caused the collapse of the Viking settlements in North America and Greenland.

That period also gave us the terrible winter at Valley Forge through which George Washington's Revolutionary War army suffered. And it gave us the cold, snowy winters in London memorialized by Charles Dickens in "A Christmas Carol."

But by now, almost all scientists agree that the current global warming trend is not just due to natural variation. This time it is caused or significantly exacerbated by anthropogenic (human) emissions of greenhouse gases such as carbon dioxide and methane. 

Most of those emissions come from burning fossil fuels - coal, oil and gas in our homes, factories, cars, power plants, etc. What we've been able to learn from many different lines of evidence is that comparatively small changes in global atmospheric carbon dioxide levels have triggered relatively large and surprisingly fast changes in global climate. Over the past century, human beings have pumped enough carbon dioxide into the air to dwarf the natural fluctuations that the earth has experienced for many hundreds of thousands of years.  

In short, the true trend is now clearly visible through the fog of noisy data. And all of us, especially our children and grandchildren, are in for a lot of heartache.