Scientists from the United States concluded that satellites had incorrectly measured the effects of global warming for many years. The temperature on the planet could rise even more.

The researchers said global warming could be even worse than the researchers expected. They came to this conclusion after studying methods of satellite temperature measurement. In their opinion, the devices do not take into account the warming of the lower atmosphere over the past 40 years.

Their devices build a correlation between temperature and humidity. However, many measurements of temperature and humidity used in climate models diverge from this relationship. This means that either satellite measurements of the troposphere have underestimated its temperature or overestimated its humidity.

“It is difficult to determine which interpretation is more plausible right now,” said study leader Ben Santer, a climate scientist at Lawrence Livermore National Laboratory (LLNL) in California. “But our analysis shows that several sets of observations, especially those with the smallest ocean and tropospheric warming values, are inconsistent with other independently measured complementary variables.”

Sunter and his team compared four different ratios of climatic properties. These are the ratio of the surface temperature of the tropical sea to tropical water vapor, the ratio of the temperature of the lower troposphere to the tropical water vapor, the ratio of the temperature of the middle and upper troposphere to the tropical water vapor, and the ratio of the temperature of the middle and upper troposphere to the temperature of the surface of the tropical sea.

In the models, these relationships are strictly defined based on the physical laws governing moisture and heat. It takes more energy to heat humid air because the water absorbs heat efficiently. Warmer air can also hold more moisture than colder air, as seen in the morning dew – when the air cools, it loses water.

However, the researchers found that satellite observations did not adhere to these well-defined rules. Instead, they varied widely, depending on which dataset the researchers used. This may mean that some datasets are less accurate.