The Challenge of Measuring the Past

Michael Schlesinger, of the University of Illinois at Urbana-Champaign.

Michael Schlesinger is a professor in the Department of Atmospheric Sciences at the University of Illinois at Urbana-Champaign, and he has devoted almost four decades to studying global climate. He has worked extensively with global temperature records, and he recognizes their limitations. “The observations of temperature over Earth were not made for the purpose of climate, they were made for the purpose of weather,” he says. People naturally built weather stations where weather conditions would affect human activity, which excluded sparsely populated or rarely traveled areas. The addition of new weather stations over the years has enhanced recent temperature records, but not historical records. Meanwhile, some older weather stations have been moved, for instance, from city centers to airports. More importantly, the weather stations have focused on land even though oceans cover 70 percent of our planet.

“Earth is unique in our solar system in that it has an ocean,” Schlesinger explains. “The atmosphere stirs the ocean by wind, and as warm surface waters mix with cooler waters below, they take heat away from the surface. This reduces the amount of surface warming that would occur if Earth were all land.” This complicates the task of calculating a representative average temperature for the entire globe, and different research groups choose different methods to account for the complexities.

“If you had the ability to uniformly distribute thermometers over the planet, and gather measurements from this regular grid, then I think everybody would analyze the data in the same way,” Schlesinger says. On our real world, however, climatologists must work with “a hodgepodge of measurements over the continents.” Not surprisingly, climatologists have developed more than one approach to understanding historical temperature records based on weather station measurements.

Global temperature anomalies estimated by NASA’s Goddard Institute for Space Studies (GISTEMP) and the UK Met Office and the University of East Anglia’s Climatic Research Unit (HadCRUT3). The consistently higher temperature indicated by GISTEMP is the result of the two groups measuring anomalies from different base periods: GISTEMP values show how much the world has warmed compared to the average from 1951 – 1980 while HadCRUT3 values show the change from the 1961 – 1990 average. Differences in how each group calculates global temperature result in differences in each year’s ranking. HadCRUT3 lists 1998 as the warmest year, followed by 2005 and 2003. GISTEMP shows 2005 as the warmest year, followed by 2007 and 1998.

Two of the world’s foremost global temperature data sets are from the Hadley Centre in the United Kingdom, and the NASA Goddard Institute for Space Sciences (GISS) in the United States. Both groups use a combination of land-surface air temperature and sea-surface temperature records, incorporated from satellite observations and historical records, to calculate values for global temperature. Though the groups use different equations in processing the data, they do so consistently so their successive values indicate how global temperature has changed over time.

One area that has experienced some of the most dramatic warming in recent years—but lacks permanent weather stations—is the Arctic, and the Hadley Centre and GISS handle the lack of observations in the region differently. The Hadley Centre offers a data set that works only with existing temperature measurements, essentially excluding the Arctic from its global temperature calculations. GISS offers a data set that accounts for Arctic temperatures by interpolating from the surrounding region. Like the Arctic, the high-latitude region around it has seen greater rates of warming than the global average over the last decade, including substantial warming in 2005. As a result, the Hadley data set identifies 1998 as the warmest year on record whereas the GISS data set identifies 2005 as the warmest year on record.

“It’s a question of whether you interpolate or not, and different people handle it differently,” observes Easterling. “There’s no exactly right or wrong way.” And his overriding conclusion remains: Recent years rank among the warmest on record.