What was done
Hoping to detect "changes in the earth's greenhouse effect," the authors analyzed the difference between the spectra of outgoing longwave radiation obtained by two orbiting spacecraft that looked down upon the earth at periods of time separated by a span of 27 years: NASA's Nimbus 4 spacecraft, which carried an Infrared Interferometric Spectrometer (1970), and Japan's ADEOS satellite, which carried an Interferometric Monitor of Greenhouse Gases instrument (1997).  The data utilized were obtained over a specific area in the central Pacific (10°N-10°S, 130°W-180°W) and a "near-global" area of the planet (60°N-60°S).  The data were further constrained by masking out land/island areas and areas believed to contain clouds.

What was learned
A number of differences were found in the land-masked and cloud-cleared data, which the authors attributed to changes in atmospheric concentrations of CH4, CO2, O3, CFC-11 and CFC-12 that occurred over the 27-year period separating the times of their two sets of measurements.  Hence, they concluded their results provided "direct experimental evidence for a significant increase in the earth's greenhouse effect" over the 27-year time interval.

What it means
The conclusion as stated is somewhat misleading.  It does not provide direct experimental evidence for a significant increase in earth's greenhouse effect.  It does so only for the cloud-free part of the atmosphere located over a portion of the planet's oceans.  And for such circumstances, the authors' results are exactly what we would expect: over the 27 years in question, measured increases in the atmospheric concentrations of the gases mentioned should indeed have increased the greenhouse effect of the pristine cloudless atmosphere.

This finding, however, tells us nothing about earth's climatic response to the inferred increase in radiative forcing, which is what the climate change debate is all about, i.e., trying to evaluate the competing effectiveness of various positive and negative feedbacks that come into play when there is a small change in the radiative properties of the cloudless atmosphere.  In fact, the authors' finding is so rudimentary as to be essentially meaningless.  Assuming, for example, that their handling of their data is correct - and this is a huge assumption they spend over half their paper discussing - they have simply verified the definition of the greenhouse effect!

Clearly, the outcome of a battle is a function of both the intensity of the offence and the response of the defense; and earth's climate system has an unbelievable arsenal of superb negative-feedback defenses that can be called upon to repulse the attack of a minor intensification of the greenhouse effect of the cloudless atmosphere (see, for example, Feedback Factors in our Subject Index).  So don't freak out just yet.  Harries et al. have merely demonstrated what everyone has known since the term "greenhouse gas" was invented: CO2 is one.