Marcel

The reanalysis data no doubt has problems, but are they sufficiently severe that they should be completely disregarded? When I look at that data for RH in particular, its consistently falling at all levels of the atmosphere. It doesn’t look like random data to me. So at the least one should conclude that there are some contradictory indications within the field.

Science questions are hard. What I am asking is easier.
Is this a high quality study and future research is unlikely to change the result, OR
Is this a study with severe limitations and further research could
well overturn it?
Is this really proof positive of EGE, or is there another body buried up there in the upper troposphere?

The reason we are talking about the RH is because IPCC makes claims about it, and it has impacts, e.g. on model assumptions as you mentioned. Even if global RH has no physical reality, it has a statistical validity, same as global temperature.

Don’t get me wrong, I am not in favor of IPCC’s “science by consensus”, to say the least.

What IPCC refers to – correct me if I am wrong – is the upper tropospheric humidity as measured by satellites. This appears to be a more useful product, i.e. it is a more consistent dataset because it is measured by just one instrument at a time. Obviously there is the issue of how to combine the observations from various platforms that have subsequently been observing but that is a whole other discussion; see for example discussions w.r.t. the determining long time series of microwave temperatures vs. long time series of surface or radiosonde temperatures.

Measuring RH from sondes is very tricky, especially at higher altitudes where temperatures go below the freezing level of water and you also have to consider relative humidity over ice, which behaves differently from relative humidity over water. Furthermore, the temperature changes with altitude are such that absolute humidity decreases by several orders of magnitude going from the surface the the upper troposphere.

A “google” search on “relative humidity bias” gives you some idea about the problems encountered in observing relative humidity from sondes.

Nevertheless, I still think that averaging RH fields – even if they are of high quality – is questionable. I would have to look in some publications whether or not this is discussed and addressed in their papers, but I have not run across any discussion for now. What does it mean, a global average RH? Mix two air masses with different temperatures and the resulting RH is definitely not the average of the RH of both samples.

Furthermore, I simply don’t know why the reanlysis humidity should be better than those from satellites, let alone that I even don’t know if I should trust the reanalysis w.r.t. humidity. (how to interpret the satellite observations is yet another question …)

Finally, the statement about constant RH and increasing (probably) refers to a paper by Soden en colleagues in Science (4 November 2005) about those satellite observations of upper air humidity, where they do not see any change in RH, which, by default, means that if temperatures change, specific humidity changes a well. Since temperatures have changed slightly, humidity has changed slightly as well. By the way, that paper claims a good correlation between observed and climate model behavior, but in their climate model they prescribe historical sea surface temperatures, hence it would be surprising if there would be a difference. But that doesn’t tell you much, other than that the model keeps RH constant for good or bad reasons.

jenne.

Thats OK. Perhaps you would like to answer some of the questions.

1. “I would not recommend averaging RH values”

But the IPCC does, in many places, and concludes: “To summarise, the available data do not indicate a detectable trend in upper-tropospheric relative humidity.”

There is data available that shows a decreasing RH, however flawed.

2. All of the problems with SH measurements could be applied to global temperatures: ” i.e. “several deficiencies and spurious trends”. Where are the confidence intervals that allow you to write them off?

3. “To summarise, the available data do not indicate a detectable trend in upper-tropospheric relative humidity. However, there is now evidence for global increases in upper-tropospheric specific humidity …” How could this be an accurate statement? Where is (a) the data to show no trend in relative humidity, (b) acknowledgment of the variance in the data and results around specific humidity?

… As Trenberth and Smith (2005) showed, such checks indicate considerable problems prior to 1979 in reanalyses, but results are in better agreement thereafter for ERA-40. Evaluations of column integrated water vapour from the NASA Water Vapor Project (NVAP; Randel et al., 1996), and reanalysis data sets from NRA, NCEP-2 and ERA-15/ERA-40 (see Appendix 3.B.5.4) reveal several deficiencies and spurious trends, which limit their utility for climate monitoring (Zveryaev and Chu, 2003; Trenberth et al., 2005a; Uppala et al., 2005). The spatial distributions, trends and interannual variability of water vapour over the tropical oceans are not always well reproduced by reanalyses, even after the 1970s (Allan et al., 2002, 2004; Trenberth et al., 2005a). …

So, in other words, don’t trust the reanalyses w.r.t. water vapor, especially its long term variations! This is not surprising, given the known problems with for example water vapor observations made by radiosondes.

Anyway, since relative humidity is field – a statistic of an air mass directly related to the temperature of that air mass – and not a physical quantity, I would not recommend averaging RH values – also because of the non-linear relation between RH and temperature (Clausius-Clapeyron). I actually have no idea how to interpret an average RH value…

[and yes, I realize that RH is key in climate model cloud parameterizations, but at least there it is used as a local variable, although keeping it constant is definitely an ad-hoc assumption]