Yes. The point I was making was that the amplitude of the temperature peaks depend on the filtering of high frequency signals by either running averages or, in the case of data before the historical record, by averaging inherent in the proxy technique. Does anyone really know how much?
and 1877-1878 was a period with a big El Nino.
Sure. And El Nino, Solar activity (and probably CO2) fluctuate naturally. So spikes happen naturally.
But when you look at all the reconstructions in the figure I offered -- from ClimateAudit, created by a colleague of Loehle -- the basic hockey stick picture is still there
How much of this reconstruction relies on tree-ring data?
What we need to know scientifically is how much natural variability is contributing to the current temperature rise.
Before we do that we should first determine how much of the temperature rise is real or caused by:
a) Local artifacts like Heat Island Effect
b) Improper siting of temperature stations
c) Errors from aging instruments
d) Mathematical errors or outright fraud
Good science allows transparency so that other scientists can see your data and your methods. Unfortunately that is difficult with much of government science and getting harder all the time. Data sets have been moved to pay-per-view sites or given only grudgingly. NCDC weather site data has been withdrawn entirely citing "privacy" concerns. The formula and algorithms used by computer models are shielded from public scrutiny. This is not good science. This is a medieval priesthood.
I am completely not an expert, but I can intuitively state that it is probably not good practice to evaluate trends based on maximum-to-minimum points when the data has considerable variability. Filtering is useful, but I can't say much more about it.
And El Nino, Solar activity (and probably CO2) fluctuate naturally. So spikes happen naturally.
That's why spikes shouldn't be used to evaluate trends.
How much of this reconstruction relies on tree-ring data?
Can't tell from the graph. Certainly some does. Ask McIntyre! But we're discussing Loehle, which is a tree-ring-free analysis -- and the same basic picture emerges. Most of the question is about the global significance of the MWP temperatures.
a) Local artifacts like Heat Island Effect
b) Improper siting of temperature stations
c) Errors from aging instruments
d) Mathematical errors or outright fraud
Yes, it's always good to evaluate the quality of the data. I've been following the discussions of data quality, and I believe this: if you had "perfect" data, you'd see maybe a 10-25% difference in the trends observed now. Here's a couple of things to think about when you are considering data quality:
Global warming brings earlier spring thaw to Great Lakes
Warming Trend Seen In Late Freeze, Early Thaw Of Northern Waterways, Say Science Researchers
Rising Height of Atmospheric Boundary Points to Human Impact on Climate (I think you'll definitely find this one interesting)
The reason I provide these -- and I could provide others -- is that these "natural" indicators demonstrate clearly that a temperature increase is happening, regardless of cause. What would be confounding would be to have all these natural indicators, yet the instrumental data did not show an increasing trend.