Recent ocean warming: are satellites right?

Recent ocean warming: Are satellites right?

Satellites have detected a significant warming in Earth's oceans between 1982 and mid-1988 that conventional methods have underestimated, reports Alan E. Strong from the National Environmental Satellite Data Information Service in Suitland, Md. But another researcher who works with the same measurement sees the purported warming as a largely artificial one created by biases in the satellite information and by the brevity of the record.

According to Strong, who presents his analysis in the April 20 NATURE, "The global ocean is undergoing a gradual but significant warming of [approximately] 0.1 [deg.] C per year, whereas the trend obtained for the same period from conventional data sources (ships and buoys) is about half that magnitude."

Satellite measurements for ocean temperature go back no farther than 1982. While this relatively short observation period makes it premature to use satellite data to detect long-term trends, such as a greenhouse warming, "we may just be beginning to witness the onset of this warming through satellite surveillance of ocean-surface temperature," says Strong.

Yet Richard W. Reynolds of the Climate Analysis Center in Camp Springs, Md., says he is "flabbergasted" by the reported warming in the satellite data. "I think this whole thing is an error."

The controversy resolves around data taken by thermal sensors aboard several satellites run by the National Oceanic and Atmospheric Administration. These instruments measure infrared radiation emitted by the ocean, which can give an indication of sea-surface temperature once researchers perform difficult corrections for water vapor and clouds in the atmosphere.

By checking the satellite observations against measurements taken by drifting buoys, Strong says he has corrected for the important biases in the satellite data. (Bias is a consistent tendency to overestimate or underestimate.)

But Reynolds says significant biases remain in the satellite data. The 1982 eruption of the Mexican volcano El Chichon skewed the record by flooding the stratosphere with dust particles that absorbed infrared radiation. For the first two years of the observation period, these particles made the oceans seem cooler than they were. Other biases can affect certain regions of the globe. Right now, satellite instruments consistently indicate the Western Pacific is 0.5 [deg.] C colder during day than during night, which poses a clear problem, Reynolds says.

In his paper, Strong reports that eliminating the el Chichon years of 1982 and 1983 only slightly reduces the observed warming. But Reynolds contends that these biases introduce a substantial artificial warming into the record. In his own analysis of global ocean temperatures, Reynolds mathematically blends satellite measurements with those taken by both ships and buoys, a technique he says removes the satellite biases. The blended record shows almost no temperature rise between 1982 and 1988.

Reynolds notes that ending the record in mid-1988 exaggerated the satellite errors because the record includes the hot El Nino of 1987 but misses much of the cool La Nina of 1988. Because of such natural fluctuations, he and other researchers caution against using a six-year-long record to talk about temperature trends.