John Christy’s Comment On “If You Want To Roll The Climate Dice, You Should Know The Odds”

In response to the post yesterday

Debate On The “Climate Dice” Issue

with respect to the post on The Conversation

If you want to roll the climate dice, you should know the odds

John R. Christy, Distinguished Professor of Atmospheric Science and Director of the Earth System Science Center of the Univeristy of Alabama in Huntsville, has provided us with his perspective on the Conversation post.

Following is John’s insightful comment.

To make an apples to apples comparison between the 1981 paper by Hansen and observations since 1979, a couple of adjustments need to be made. Without these adjustments, the comparisons are apples to oranges and conclusions misleading.

Eyeballing the forecast curve gives a model projected surface trend of about +0.15 C/decade since 1979.  I suspect the rise in CO2 was actually faster, and thus this curve underestimates what the model would have shown had it used better emissions numbers.

This +0.15 C/decade is a surface trend, so to compare with GISS model troposphere, the scaling factor is 1.25, giving a 1981 model trend of +0.19 C/decade.  That’s one adjustment.

The second adjustment is to account for the volcanic cooling in the first part of the 1979-2012 observations which tilts the observed trend to be more positive than otherwise by about +0.04 C/decade.  So had the 1981 model included real volcanoes which cooled the early portion, its tropospheric trend would be tilted upward to about +0.23 C/decade.  This compares with observations of UAH and RSS of +0.13 C/decade.  So, the apples to apples comparison with tropospheric temperatures since 1979 would be, model: +0.23 C/decade, observations +0.13 C/decade.

Now if we start after the effects of Mt. Pinatubo (say around 1996) to avoid having to calculate the volcanic impact, we have the following. The model goes from +0.2 to +0.5 in 16 years (i.e. trend of about +0.19 at the surface or +0.23 C/decade in the troposphere) while observations show UAH +0.11 C/decade and RSS +0.04 C/decade.  Quite a difference between models and observations!

As an interesting footnote, the average climate model for the most recent CMIP-5 RCP4.5 result has a calculated tropospheric trend of +0.22 C/decade for 1979-2012 and +0.29 C/decade for 1996-2012.  So, not much has changed in 30 years as far as model skill in replicated tropospheric trends it seems.

John C.

source of image