contribute substantially to the trends observed during a period of record as short as 20 years. The presence of these sampling variations, together with the remaining uncertainties inherent in the temperature measurements themselves, preclude the possibility of drawing more definitive conclusions concerning the cause of the observed disparity in the trends.
It is clear from the foregoing that reconciling the discrepancy between the global-mean trends in temperature is not simply a matter of deciding which one of them is correct, or determining the ideal "compromise" between them. In the long term, it will require major advances in the ability to interpret and model the subtle variations in the vertical temperature profile of the lower atmosphere that occur in association with the internal variability of the climate system in response to volcanic eruptions and solar forcing, and in connection with changes in atmospheric composition due to human activities. It will also require more precise and extensive satellite- and ground-based observations for monitoring climate change, and changes in the way these observations are implemented and processed. A detailed consideration of these issues is beyond the scope of the panel's charge (see Preface). However, the panel does offer a number of recommendations (see chapter 4) for short-term actions that it views as steps in the right direction.break