Skip to main content
U.S. flag

An official website of the United States government

Publication Date
29 June 2021

Satellites may Underestimate Warming in the Troposphere

Complementary observations may help improve understanding of the rate of satellite-era atmospheric warming.
Print / PDF
Powerpoint Slide

Scientists at Lawrence Livermore National Laboratory in collaboration with colleagues from several institutions analyzed the relationship between different climate variables in state-of-the-art climate models and observations. These inter-relationships expose possible observational biases in the rate of atmospheric warming and/or moistening.


Past studies have shown that the observed rate of tropospheric warming is significant and cannot be explained by natural climate variability. Over the last four decades, however, models tend to exhibit greater warming than satellite observations. This study indicates that part of the model-observational difference over the satellite era may be due to biases in the observed rate of tropospheric warming.


The team studied the ratio between trends in pairs of several different “complementary” variables. Complementary variables – like tropical temperature and moisture – are expected to show correlated behavior that is governed by basic, well-understood physical processes. The researchers showed that distinct pairings between tropical sea surface temperature, tropospheric temperature, and water vapor content are tightly constrained in climate model simulations, despite model differences in climate sensitivity, external forcings, and natural variability. In contrast, each ratio exhibits a large range when calculated with observational datasets. Model trend ratios between WV and temperature were closest to observed ratios when the latter are calculated with data sets exhibiting larger tropical warming of the ocean surface and troposphere. It is possible that observed water vapor trends are systematically too large or that trends in certain tropospheric and sea surface temperature data sets are too small. This research indicates that further research using multiple complementary geophysical measurements may help to understand observational biases and reduce observational uncertainties.

Point of Contact
Benjamin Santer
Lawrence Livermore National Laboratory (LLNL)
Funding Program Area(s)