The globally-averaged surface temperature trend has been highlighted as an icon of climate change. For example, a meeting was held In Exeter, United Kingdom from Feb 1-3, 2005 entitled “Avoiding Dangerous Climate Change.” The focus on a globally-averaged temperature trend was clear in the emphasis at the meeting. The Hadley Centre brochure relevant to this meeting stated “Once a tolerable (i.e., non-dangerous) change has been determined – say in terms of a global temperature rise – we then have to calculate what this corresponds to in terms of tolerable greenhouse concentrations in the atmosphere.” The message is that a clear global surface temperature threshold exists over which there are dangerous effects on the climate system.
This perspective however, avoids discussing the real issue associated with long-term variability and changes in climate.
First, in the context of atmospheric circulation changes (which is, after all what produces our weather), it is the regional tropospheric temperature and humidity trends that are important, not a global average surface temperature A change in the globally-averaged surface, or even globally-averaged tropospheric, temperature are important primarily in the context of how this results in circulation changes. The globally-averaged surface temperature is a very poor metric to use to assess these circulation changes. The 2005 NRC report recognized this limitation in using globally-averaged surface temperatures. Secondly, with respect to even “global warming” the ocean heat content changes, rather than the surface temperature anomaly provides a more robust metric (see R-247).
With respect to the surface temperature itself, there are several issues with respect to the spatial representativeness of the trends that have been incompletely (or not at all) investigated. These are:
- 1. Poor microclimate exposure:
- This is a land issue. The use of photographs to exclude questionable stations is obvious (and we are quite puzzled why anyone would not make this a high priority). The effect of poor exposure (which results in different site exposure depending on the wind direction) and changes in the site conditions over time have not been quantified. Our qualitative assessment based on the photographs that we have seen is that this it is likely to insert a warm bias for most sites.
- 2. Moist enthalpy:
- This is both a land and an ocean issue. The use of the terms “warming” and “cooling” are being incompletely used when there is significant water vapor in the surface air (tropics and mid-latitude warm seasons, in particular). This will produce a warm bias when the air actually became drier over time, and a cool bias when the air becomes more humid over time. This effect has not been quantified with respect to how it influences regional and global surface temperature trends. It has been shown to be significant for individual sites.
- 3. Vertical lapse rate issues (paper in preparation):
- The influence of different lapse rates, heights of observations and surface roughness have not been quantified. For example, windy and light wind nights should not have the same trends at most levels in the surface layer, even if the surface-layer averaged temperature trend was the same.
- 4. Uncertainty in homogeneity adjustments:
- Time of observation, instrument changes, and urban effects have been recognized as important adjustments (see R-234) that are required to revise temperature trend information in order to produce improved temporal and spatial homogeneity. However, these adjustments do not report in the final homogenized temperature anomalies, the statistical uncertainty that is associated with each step in the homogenization process.
Thus even if the globally-averaged surface temperature was a particularly appropriate metric to assess climate change, there are issues on the robustness of this data set which have been overlooked. Our recommendation, however, is to deemphasize the globally-averaged surface temperature as a climate change metric and assess instead circulation changes as defined by tropospheric temperature and water vapor (and for the ocean, temperature and salinity) variability and trends.