Latest News (hidethedecline)
Discussion of the DMI arctic 80-90N data |
| Posted by Frank Lansner (frank) on 5th February, 2011 |
Klaus Flemløse has challenged my findings on the DMI Arctic 80-90N, and his viewpoint deserves a little extra debate on the subject.
Klaus Flemløses writing:
Personally I was surpriced to see that other estimates of the DMI data suggested that I had presented the data from DMI wrong, in fact Klaus shows that there is some trend difference in my readings vs. 2 other. So, i decided to re-analyse data from DMI, this time using software based pixel counting:
The difference between my manual pixel readings and the software reading are minimal, better than i imagined, in fact. So, differences between my readings and other readings must be due difference in what is read, for example a slightly different date interval or the like.
Non the less, all data readings shows rather similar trends.
Klaus is challenging my article:
Here im comparing the decline 1991-2009 in DMI Arctic data with the GISS 80-90N June-july data that shows a significant temperature increase:
Klaus mentions the shift in DMI-data in 2002 where DMI ends using ERA-40 data and then starts using T511 data. Im not sure why these 2 different data sets should lead to a significant different dataset since both measure the celsius temperature using rather similar input type.
Non the less, it is a problem stitching datasets, (i need just to mention how IPCC stitches tree ring proxy data with the mostly city based temperature data etc...).
DMI uses both dataset ERA-40 and T511 for year 2002. As mentioned, both datasets are the actual temperatures in Celsius, so im not sure DMI could just make a adjustment to one of the datasets when stitching. Heres how the 2 datasets resembles each other over the melt period, summer:
And allthough the datasets appears similar, the ERA-40 is 0,121 Celsius warmer than T511 in this particular period of 2002: What if this difference where general for all years? - I believe perhaps this is what Klaus suspects. Then we would have a step change that explains 0,12 Celisus of the cooling trend:
The years after 2002 should then be:
However, according to Klaus Flemløse, the 2 datasets due to the overlap in 2002 should be viewed as 2 totally independent series, both showing no trend:
However I find this to be against the DMI presentation of data: DMI shows 1991-94 data to be far above their green average line while recent years are shown to be far under average. If this should have any meaning, DMI should be of the opinion that temperatures 1991-2010 has indeed declined for 80N-90N melt season. If data presented as far below average today is NOT far below average etc, then the DMI presentation has no meaning.
Further, my article compared the DMI decline 1991-2009 with GISS data showing a 0,5-1 Celisus increase in temperatures. When Klaus in stead focus on the whole 1958-2002 flat trend I dont believe he is really adressing why I was actually pointing out.
Finally, the idea that all show a flat trend has "consequences".
If the entire flow of temperatures 1958-2010 should have had a flat trend in 80-90N temperatures, how much warmer should T511 data then be?? Answer: 0,5 Celsius:
(Compare with the suppsed 0,7 Celsius global warming over a century).
So termometers used to create T511 should be showing temperatures 0,5 Celsius too cold compared to ERA40 to claim a flat trend. And EVEN if this should be true, dont forget, that GISS data shows a 0,5 Celsius or more temperature increase. So... to make GISS in compliance with ERA40/T511 we should not "just" add 0,5 Celsius to T511 but rather more than twice as much.
I dont find this likely, and thus I dont find that GISS projected data has any value over the Arctic 80N-90N.
Further: The year 2002 with a direct compare of T511 vs ERA40 does not at all support that there should be a significant difference besides the small 0,12 Celsius difference. So on what ground can one claim a difference of 0,5 Celsius or 1 Celsius should be true??
In general I am not really convinced that there has been no decline in Arctic temperatures 80N-90N in the melt seasons 1991 - 2010, at least, I need to see more support of this opinion.
However, its truly odd how temperatures over the ice covered 80-90N melt season appears to have been cooling while the Arctic ice extend has been reduce in the period, quite odd, I certainly admit that!
This would suggest, that ice is at least partly lost due to currents and winds rather than warmer temperatures over the ice.
Last changed: 5th February, 2011 at 20:46:30
BackComments
| Rev 1. Sea Ice Emissivity in the Arctic | By Unknown on 30th July, 2011 at 20:34:11 |
| Nope, I'm wrong. Emissivity of sea ice, salt water, and snow are all very nearly equal across almost all wavelengths of interst: 0.98 to 0.99 Makes all three flat surfaces (if they are at the same 1.0 Deg C temperature) very nearly perfect black body radiators into the sky - regardless of whether the sky is cloudy, clear, or partial. All three will act (nearly) the same. But, the sea ice will prevent the surface water from evaporating (even though that amount of evaporation will be very low at 1 degree C, compared to a tropical 30 degree C seawater); and it will tend to "insulate" the surface of the sea water from the direct transfer of heat energy into the air. Complex problem. Robt RACookPE1978 |
|
| Net Heat Radiation in the High Arctic: Above 80 N, Does Open Ocean "Heat" or "Cool" the Planet? | By Unknown on 30th July, 2011 at 18:55:18 |
| Check my calculations and my data sources: I'm finding that in high latitudes (above 75 north), at 1 degrees C (mid-summer through Sept), at sea level into a clear sky, at the equinox (12 hour sunshine, 12 hours of darkness), 1 m^2 of "black" open ocean water (albedo = 6) radiates more energy than 1^ meter of sea ice (albedo = 30) or 1 m^ of snow-covered sea ice (albedo = 90) during every hour of darkness. Now, solar radiation received during the hours of daylight is a different question, but this is only for the hours of darkness on a clear night. RACookPE1978 |
|
| Thanks to Frank | By Unknown on 23rd February, 2011 at 20:31:04 |
| Firstly, I am pleased that Frank has given a fair presentation of my point of views in respect of the quality of his pixel readings. Secondly, Frank is now aware of the problem of data quality. I will not go into the discussion about Frank assertions that the GISS temperatures can’t be used for 80N-90N. I will refer to Skeptical Science. I have no further comment so far. Regards Klaus Flemløse |
|
| Superbe inut, Mr Duffin :-) | By Frank Lansner on 21st February, 2011 at 22:21:02 |
| Its the kind of info that i will have to speculate a little about (was that English??) Arctic warming mostly in winter... hmmm. Well, that might make some sence. Another thing, one talks about Arctic amplification of temperature variation. I just find it "odd" that we see pretty much the same arctic temperature peak around 1930-40 as today. The princible of Arctic amplification does not explain/support that r.o.w. should be warmer today than 1930-40. If the amplification works the same way today than in 1930-40 we should to some degree could read R.O.W. tempeature variations from Arctic data? And if so: Hardly any warming. |
|
| Air temp vs melt | By Unknown on 21st February, 2011 at 19:20:29 |
| Frank, both theory and data suggest that arctic warming and cooling are felt mainly at night and in the winter. In fact the Danish data shows no change in the warmest temperature since the record started. In winter, less cold will not cause melting and ice loss. I have spent considerable time on glaciers in the spring and summer, and find that the surface melts under bright sun even when the air near the surface is still below zero celsius, and conversely, the ice will not melt under overcast, even when the air near the surface is a little above zero celsius. My vote for ice loss would be ocean currents plus insolation, aided by wind circulation pattern, not air temperature. In 2007 there was a stationary high over the west arctic and Siberia that kept clouds away for most of August, and there was a wind circulation pattern driving ice eastwards, and out into the Atlantic. |
|
Add Comment