whether temperature changes originating at the surface necessarily lead to temperature changes within the troposphere.

2. Similarly, a discussion of the relation between cumulus convection and the moist adiabat provides an opportunity to use such differential trends to understand the coupling between the surface and the lifting condensation level. Indeed, in the tropics, the temperature structure consists of a surface mixed layer (up to about 500 m) and a trade wind boundary layer (up to about 2 km) above which is the free troposphere. Each of the boundary layers is topped by an inversion which tends to isolate the layer from the region above (Sarachik, 1985). Outside the tropics, the surface communicates with upper levels primarily by quasi-horizontal motions along isentropic surfaces (e.g., Hoskins, 2003). Consequently, the report and the scientific community should move beyond the naïve notion that the lapse rate is a rigid constraint operating from the surface to the tropopause. Instead the observations this report is concerned with should be exploited in order to answer important questions about climate. This objective provides meaningful motivation for ascertaining the accuracy of the temperature measurements and the resulting time series. That said, it should be emphasized that the temperature changes being considered are changes on the order of tenths of a degree (although local changes may be much greater), and current theories may prove inadequate for such small changes.

3. In general, spatial and temporal sampling is not adequately dealt with in the Temperature Trends CCSP report. Given the fact that horizontal temperature variability at the surface tends to get smoothed as one rises to the free troposphere, there may be serious issues of sampling. Horizontal smoothing over large scales occurs above the boundary layer, but that at the surface and within the boundary layer, there can be much more horizontal variation of temperature. Thus, much more data may be required at the surface to get characteristic temperatures.

4. For Chapter 1, explanations of the processes involved in determining vertical profiles of temperature should represent the current state of understanding or lack thereof. The chapter should focus less on details of the vertical profile of temperature that are not resolved by the observations that are the focus of the report. For example, the satellite data are only reported in coarse vertical layers.

5. For discussions that are felt to be too detailed for the body of the text, footnotes are a reasonable device.


1. The chapter should include more discussion of theories that provide physical constraints on the apparent differences between surface and tropospheric records.

2. The discussion on lines 69-80 should be replaced with a more accurate figure as well as a description of the differences between the tropics, the extratropics, and the polar regions. In the tropics, the temperature is hardly linear with height, given that the lapse rate associated with the moist adiabat goes from about 5 K/km near the surface to almost 9.8 K/km at the tropopause near 16 km. It should also be noted that the tropopause descends sharply to 12 km near 30 degrees latitude and to around 8 km near the poles. The existence of the near surface inversion layer at high latitudes should also be noted as well as its dependence on meteorological conditions.

The National Academies of Sciences, Engineering, and Medicine
500 Fifth St. N.W. | Washington, D.C. 20001

Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement