Extensive Antarctic campaign finds cold bias in satellite records

May 07, 2013

Advances in satellite sensing have now made it possible to track changes in the world's most remote locations. Over the Antarctic continental interior, the vast majority of profiles of atmospheric temperature are provided by satellite remote sensing, making proper calibration of the satellite observational equipment and analysis algorithms particularly important. However, the hostile environmental conditions in regions such as the Antarctic make it difficult or even dangerous to conduct the in-the-field observations needed to calibrate and validate the satellite observations.

From September through December 2010, Wang et al. carried out an extensive program to directly measure atmospheric temperature profiles over coastal and interior Antarctica and over the surrounding ocean. The researchers launched 13 Driftsondes, gondola-laden stratospheric balloons that carry up to 56 individual dropsondes-sensors designed to measure the pressure, temperature, humidity, and wind speed and direction as they fall from the stratosphere to the surface. Using the dropsondes, the authors collected 639 profiles of the Antarctic atmosphere. The campaign produced a collection of observations that spanned the southern continent and provided an opportunity to assess the skill of various remote-sensing satellites.

Comparing their observations against coincident measurements made with 10 different satellite sensors, the authors find that 9 of the 10 sensors consistently underestimated Antarctic atmospheric temperatures and that errors are generally larger over the continental interior than along the coast. The findings reveal that the cold biases can be from 0 to 4 degrees Celsius (0 to 7 degrees Fahrenheit), depending on the satellite sensor and the altitude of the atmosphere being observed. Despite the cold bias, however, the satellites do tend to properly represent the profile's shape.

Explore further: Lava from Hawaii volcano picks up speed

More information: Unprecedented upper air dropsonde observations over Antarctica from the 2010 Concordiasi Experiment: Validation of satellite-retrieved temperature profiles, Geophysical Research Letters, doi:10.1002/grl.50246, 2013 http://onlinelibrary.wiley.com/doi/10.1002/grl.50246/abstract

add to favorites email to friend print save as pdf

Related Stories

First satellite detection of volcanogenic carbon monoxide

Dec 17, 2012

Measuring and tracking the gases that vent from an erupting volcano is a project wrought with potential dangers and difficulties. On the ground measurements place researchers in harm's way, as do airborne sampling surveys. ...

Metop-B delivers first data from polar orbit

Oct 01, 2012

(Phys.org)—Four of the instruments on the Metop-B weather satellite (AMSU-A, ASCAT, MHS, GRAS) have been activated this week and are delivering data.   This demonstrates that Metop-B, launched on 17 Septe ...

Reanalyses find rising humidity in the Arctic

Jul 06, 2012

(Phys.org) -- Direct, reliable observations of atmospheric conditions stretch as far back as the mid seventeenth century, with otherwise consistent records being punctuated by periodic updates in methods, practitioners, and ...

Recommended for you

Icelandic volcano sits on massive magma hot spot

15 hours ago

Spectacular eruptions at Bárðarbunga volcano in central Iceland have been spewing lava continuously since Aug. 31. Massive amounts of erupting lava are connected to the destruction of supercontinents and ...

NASA sees Tropical Storm Ana still vigorous

17 hours ago

NASA's TRMM satellite saw that Tropical Storm Ana was still generating moderate rainfall is it pulled away from Hawaii. The next day, NASA's Aqua satellite saw that wind shear was having an effect on the ...

User comments : 2

Adjust slider to filter visible comments by rank

Display comments: newest first

The Alchemist
1.8 / 5 (8) May 08, 2013
Underestimated temperatures, that means the Antarctic is warmer! Amazing.
LindaSerena
1 / 5 (1) May 18, 2013
It should have no effect on temperature trends, if this supposed bias is consistent over time.

The big issues are still the warm biases on surface trends, such as UHI, siting issues, extrapolation of GISS warm land trends on Arctic and the overwriting (!) of sea surface meta data in the 1940s (inleads instead of bucket).

The largest data error appears to be a massive increase in ocean heat content data during switching to ARGO floats in the early 2000s. In 3 years 0-2000, heat content increased by the same amount as in the other 37 years between 1970-2100 combined, what is totally implausible.