Guest post by Steven Goddard
Photo above from: NY Daily News: Record Snowfall in New York
Now that we have reached the end of the meteorological winter (December-February,) Rutgers University Global Snow Lab numbers (1967-2010) show that the just completed decade (2001-2010) had the snowiest Northern Hemisphere winters on record. The just completed winter was also the second snowiest on record, exceeded only by 1978. Average winter snow extent during the past decade was greater than 45,500,000 km2, beating out the 1960s by about 70,000 km2, and beating out the 1990s by nearly 1,000,000 km2. The bar chart below shows average winter snow extent for each decade going back to the late 1960s.
Here are a few interesting facts.
- Average winter snow extent has increased since the 1990s, by nearly the area of Texas and California combined.
- Three of the four snowiest winters in the Rutgers record occurred during the last decade – the top four winters are (in order) 1978, 2010, 2008, 2003
- The third week of February, 2010 had the second highest weekly extent (52,170,000 m2) out of the 2,229 week record
The bar graph below shows winter data for each year in the Rutgers database, color coded by decade. The yellow line shows the mean winter snow extent through the period. Note that the past decade only had two winters below 45 million km2. The 1990s had seven winters below the 45 million km2, the 1980s had five winters below 45 million km2, and the 1970s had four winters below 45 million km2. This indicates that the past decade not only had the most snowfall, but it also had the most consistently high snowfall, year over year.
It appears that AGW claims of the demise of snowfall have been exaggerated. And so far things are not looking very good for the climate model predictions of declining snowfall in the 21st century.
Many regions of the Northern Hemisphere have seen record snowfall this winter, including Washington D.C, Moscow, China, and Korea. Dr. Hansen’s office at Columbia University has seen record snowfall, and Al Gore has ineptly described the record snow :
“Just as it’s important not to miss the forest for the trees, neither should we miss the climate for the snowstorm,”
A decade long record across the entire Northern Hemisphere is not appropriately described as a “snowstorm.”
Sponsored IT training links:
If want to improve TK0-201 score then go through 646-230 exam dumps and self test RH302 exams and get guaranteed success in first attempt.
kadaka (15:06:54) :
From what I can see, the meteoric dating yields remarkably close figures, which can indicate a single original source. Did they somewhere find so much variation in so many that multiple original sources became the best answer?
The Sun is a second, or fifth, or whatever high number, generation star. It was born with about 1% ‘metals’ [which is everything heavier than He]. And that 1% is debris from ~ a million of supernovae exploding over time, dusting the Galaxy with.
Since this involves dating by examining radioactive decay, I do not see how the theoretical collision forming the current Earth and Moon could have affected it.
Because the result is now the mixture of the protoearth and the colliding body, which likely formed at different times and at different places. Perhaps the protoearth first, because it is larger. Another point is that the dating methods assume that the parent and daughter products have been together since the ‘beginning’, but if the material was vaporized or melted and then recrystallized that assumption does not hold.
I have a witty and erudite response to Mr Goddard unfortunately it requires a graph ias a .png for it to properly make sense, and I can’t see a way of including it in the comment. Any advice..?
Leif Svalgaard (18:41:00) :
The Sun is a second, or fifth, or whatever high number, generation star. It was born with about 1% ‘metals’ [which is everything heavier than He]. And that 1% is debris from ~ a million of supernovae exploding over time, dusting the Galaxy with.
Yup, a Population I star, metal-rich. But first generation in the sense that there was no predecessor star at this location (in terms of relative position). By the predominant theory at the time that I knew, we had a single predecessor star, Population II or perhaps even III, and possibly some additional stuff from the interstellar medium got mixed in during the formation of the current star and planets. Currently the predominant theory is it was all gathered up from the interstellar medium, no single predecessor. I am reading about the finding of short-lived isotopes in old meteorites, which suggested there were nearby supernovae while our system formed. But I have found nothing which says the original molecular cloud could not have primarily come from a single predecessor star. I worry a good hypothesis has now been tossed out basically due to sample contamination.
Because the result is now the mixture of the protoearth and the colliding body, which likely formed at different times and at different places. Perhaps the protoearth first, because it is larger. Another point is that the dating methods assume that the parent and daughter products have been together since the ‘beginning’, but if the material was vaporized or melted and then recrystallized that assumption does not hold.
The Giant Impact Hypothesis still has some unresolved issues, currently it is the best fit to the available data. The alternate theories primarily suffer from not being able to readily explain the high angular momentum of the Earth-Moon system. After reading up on the nebular hypothesis and figuring in the chaotic nature back then, I’m surprised that apparently no one has ever proposed that a planetary-sized object passed nearby a larger proto-Earth, causing a Moon-sized glob to get pulled off. That would account for the angular momentum and a host of other things. The Sun currently has 99.86% of all the mass in the solar system. We can accept the young Sun having a very strong solar wind when it fired up, ignition and its subsequent effects were somewhat violent. So why couldn’t it have “spit out” some large blobs as well? I presume that would have happened during the rapid expansion of gas near the core after ignition, destabilizing the upper layers. Relatively little mass would be involved. This would also take care of some issues regarding the formation of giant planets.
Now with the dating, as Wikipedia words it while citing “Dalrymple, G. Brent (1994-02-01). The Age of the Earth. Stanford University Press. ISBN 0804723311.”:
Important note: “In older literature, the Hadean is included as part of the Archean.” So those ores should currently be properly considered Hadean, in the period from Earth’s formation to about 3.8 Gyr ago, not Archean.
These very old ores are agreeing with the meteoric data, thus it seems they were not affected by the proposed impact and thus are not afflicted with the proposed errors, and are representative of the original material that formed the proto-Earth.
Of course, we are pretty much in a region where the “accuracy” leaves something to be desired, and arguing between time of formation and time from post-impact seems a bit silly until the dating gets a lot more precise.
kadaka (14:44:37) :
Yup, a Population I star, metal-rich. But first generation in the sense that there was no predecessor star at this location (in terms of relative position).
There never was.
which suggested there were nearby supernovae while our system formed.
There very likely was a [relatively] nearby supernova. Such explosions often compress the interstellar medium leading to star formation within several tens of light-years.
The alternate theories primarily suffer from not being able to readily explain the high angular momentum of the Earth-Moon system.
The best evidence for the collision theory is the identical isotopic composition of Earth and Moon.
We can accept the young Sun having a very strong solar wind when it fired up, ignition and its subsequent effects were somewhat violent.
The strong solar wind was a result of strong magnetic fields [resulting from compression of the interstellar field]. No real violence needed. The magnetic field in the solar wind and the Sun were [and are] strongly coupled and acted as a brake on solar rotation [from 1 day to now 25 days period]. If the Sun rotates slower its angular momentum must be transferred to the circum-solar disk causing it to move away from the sun and ending up with 98% of the angular momentum. By moving away, the ‘nebula’ could cool and protoplanets form [hence younger than the Sun].
These have returned age dates of 4.54 billion years with a precision of as little as 1% margin for error.
It is important to remember that the ‘age’ is not since the elements were formed [some atoms in your body are more than 10 Gyr old], but since they crystallized out of liquid or gaseous form. So, there is no reason for the meteorites and the Earth to have the same age. If the protoearth was formed from meteorites 5.57 Gyrs old, but was completely vaporized and melted 0.03 Gyr later, we would measure the Earth to have an age of 5.54 Gyr.
Vaporizing ‘resets’ the age.
Of course, we are pretty much in a region where the “accuracy” leaves something to be desired, and arguing between time of formation and time from post-impact seems a bit silly until the dating gets a lot more precise.
Which is why I initially noted that the age difference between the Sun and the Earth was an ‘estimate’. Real-world planetology is messy 🙂
kadaka (14:44:37) :
fired up, ignition and its subsequent effects were somewhat violent
The solar ‘furnace’ is extremely gentle. It takes a million tons of Sun to produce enough energy to run a hair dryer…
I love watching artistic renditions of that theorized collision. Two protoplanets on an eventual orbital collision course. And than WHAM! The combined behemoth mass and gravitational pull wadding up the exploded but orbitally ensnared hot, sticky debris into our moon. Physical science at its best.
“Steve Goddard (05:57:52) :
Michael,
So what you are saying is that snow extent “may” have declined to a decadal record high.
Make perfect sense in Alice in Wonderland, where up is down and down is up.”
No not at all I think we are on the same side of this issue if not in exactly the same place. That suggestion along with Mr Tamino’s suggestion that “Absence of Significance is not Significant of Absence” can be properly categorised as Boll.. err Bullsh….err Horse Puc…. err Cra… err not in accordance with the science of statistics as practiced outside the twilight zone. A steaming load of dingoes kidneys, there I knew the correct technical definition would occur to me eventually.
Now the data assessed on the decadal scale shows no reason to divert from the null hypotheses that snow extent is randomly hovering about a stable mean, although we have identified some concerns with variability (In winter snow only (periods 12,01,02)) The question still remains as to whether there are shorter period trends or particular years that are unusual.
In the world of washers and widgets we would test that with a mean and variance chart a la the relevant ANSI standard.
So let’s be boring again :-
Looking for statistically aberrant behaviour in the northern hemisphere snow extent data
Introduction
Standard Industrial statistics method particularly a mean and variability (X-Bar and R) chart will be used to examine the “Process” of winter snow extent in the northern hemisphere.
Method
The period 1967 to 1987 will be used as a baseline to generate the capability benchmark for this process, NB this time period is being arbitrarily chosen from its relation to a relatively flat portion of the GISS global temperature anomaly chart rather than by proper shopfloor assessment criteria. The mean for each winter will be calculated as the arithmetic mean of Dec, Jan and Feb in each year. The Grand mean will be calculated from the average of the winter means. The range for each year will be the difference between the maximum and minimum values of the three winter periods. 1981 will be excluded as on inspection it seems to be an aberrant point.
The Mean Range will be calculated and the standard factors for ANSI control charts used to calculate the Range control limits; and in conjunction with the Grand Mean, the lower and upper Mean control limits.
The chart will be plotted and examined using standard industrial criteria.
Results
Capabilty Calculation:
Year Winter Mean Winter Range
1967 46072701 5474596
1968 44442349.67 2534117
1969 46297133.67 4397930
1970 45242413.67 4876778
1971 46315551 1282701
1972 46517476.67 4745268
1973 45632660.67 1790136
1974 44977140.33 1790246
1975 42902865.33 5398008
1976 44193418 3153827
1977 44957337.67 6676508
1978 48401983 6343339
1979 46730553.33 5470955
1980 43845123.67 10535361
1981 40818640.67 6196657
1982 45810383.33 4463569
1983 45198443.67 3781045
1984 45089016.33 2101285
1985 46726720.33 7771010
1986 46578984 1651032
1987 44872522 5823326
Grand Mean 45540238.87 4503051.85
R(u) 11590855.46
R(l) 0
X(u) 50146860.91
X(l) 40933616.82
Data to be graphed:
Year X-Bar Range X-UCL X-LCL R-UCL X-Mean
1967 46072701 5474596 50146860.91 40933616.82 11590855.46 45540238.87
1968 44442349.67 2534117 50146860.91 40933616.82 11590855.46 45540238.87
1969 46297133.67 4397930 50146860.91 40933616.82 11590855.46 45540238.87
1970 45242413.67 4876778 50146860.91 40933616.82 11590855.46 45540238.87
1971 46315551 1282701 50146860.91 40933616.82 11590855.46 45540238.87
1972 46517476.67 4745268 50146860.91 40933616.82 11590855.46 45540238.87
1973 45632660.67 1790136 50146860.91 40933616.82 11590855.46 45540238.87
1974 44977140.33 1790246 50146860.91 40933616.82 11590855.46 45540238.87
1975 42902865.33 5398008 50146860.91 40933616.82 11590855.46 45540238.87
1976 44193418 3153827 50146860.91 40933616.82 11590855.46 45540238.87
1977 44957337.67 6676508 50146860.91 40933616.82 11590855.46 45540238.87
1978 48401983 6343339 50146860.91 40933616.82 11590855.46 45540238.87
1979 46730553.33 5470955 50146860.91 40933616.82 11590855.46 45540238.87
1980 43845123.67 10535361 50146860.91 40933616.82 11590855.46 45540238.87
1981 40818640.67 6196657 50146860.91 40933616.82 11590855.46 45540238.87
1982 45810383.33 4463569 50146860.91 40933616.82 11590855.46 45540238.87
1983 45198443.67 3781045 50146860.91 40933616.82 11590855.46 45540238.87
1984 45089016.33 2101285 50146860.91 40933616.82 11590855.46 45540238.87
1985 46726720.33 7771010 50146860.91 40933616.82 11590855.46 45540238.87
1986 46578984 1651032 50146860.91 40933616.82 11590855.46 45540238.87
1987 44872522 5823326 50146860.91 40933616.82 11590855.46 45540238.87
1988 44862964 4833078 50146860.91 40933616.82 11590855.46 45540238.87
1989 43581924.67 4562448 50146860.91 40933616.82 11590855.46 45540238.87
1990 44471979.67 2141723 50146860.91 40933616.82 11590855.46 45540238.87
1991 45096895.67 2002991 50146860.91 40933616.82 11590855.46 45540238.87
1992 44015979.67 3171496 50146860.91 40933616.82 11590855.46 45540238.87
1993 45701485 1592322 50146860.91 40933616.82 11590855.46 45540238.87
1994 44816110.67 1889408 50146860.91 40933616.82 11590855.46 45540238.87
1995 43877959.67 3406431 50146860.91 40933616.82 11590855.46 45540238.87
1996 45096536.33 2788638 50146860.91 40933616.82 11590855.46 45540238.87
1997 44499508.67 4759492 50146860.91 40933616.82 11590855.46 45540238.87
1998 44813952.33 1762057 50146860.91 40933616.82 11590855.46 45540238.87
1999 43844289.67 963027 50146860.91 40933616.82 11590855.46 45540238.87
2000 44738855 4698555 50146860.91 40933616.82 11590855.46 45540238.87
2001 45034486.67 2726359 50146860.91 40933616.82 11590855.46 45540238.87
2002 44512401.67 3956045 50146860.91 40933616.82 11590855.46 45540238.87
2003 46829439 2628742 50146860.91 40933616.82 11590855.46 45540238.87
2004 45134885.33 4402515 50146860.91 40933616.82 11590855.46 45540238.87
2005 45559808 3378163 50146860.91 40933616.82 11590855.46 45540238.87
2006 45526618.67 3532290 50146860.91 40933616.82 11590855.46 45540238.87
2007 43817340 3037615 50146860.91 40933616.82 11590855.46 45540238.87
2008 46909030.33 5694332 50146860.91 40933616.82 11590855.46 45540238.87
2009 45028342 3694968 50146860.91 40933616.82 11590855.46 45540238.87
2010 47508592.33 2531736 50146860.91 40933616.82 11590855.46 45540238.87
The Graph
Chart Interpretation
1) 1981 seems to be an aberrant point (mean is below the lower control limit) indicating a particular disturbance to the process for that year.
2) The process seems to have run at lower than normal mean between 1988 and 2002 (Two period of more than 8 points below the grand mean)
Conclusion and critique
This year’s snow extent although above the long term mean and part of a decade mean also above the long term mean should not be considered as exceptional or abnormal
The period prior to and including 1981 should be examined for meteorological and geophysical events that might produce dramatically lower snow extent
It is obvious by inspection that 1982 to 2004 would be a better period to establish capability, and indeed in an industrial setting that’s what I’d do. This yields
Year X-Bar Range X-UCL X-LCL R-UCL X-Mean
1967 46072701 5474596 48472759.96 41545129.03 8715406.65 45008944.49
1968 44442349.67 2534117 48472759.96 41545129.03 8715406.65 45008944.49
1969 46297133.67 4397930 48472759.96 41545129.03 8715406.65 45008944.49
1970 45242413.67 4876778 48472759.96 41545129.03 8715406.65 45008944.49
1971 46315551 1282701 48472759.96 41545129.03 8715406.65 45008944.49
1972 46517476.67 4745268 48472759.96 41545129.03 8715406.65 45008944.49
1973 45632660.67 1790136 48472759.96 41545129.03 8715406.65 45008944.49
1974 44977140.33 1790246 48472759.96 41545129.03 8715406.65 45008944.49
1975 42902865.33 5398008 48472759.96 41545129.03 8715406.65 45008944.49
1976 44193418 3153827 48472759.96 41545129.03 8715406.65 45008944.49
1977 44957337.67 6676508 48472759.96 41545129.03 8715406.65 45008944.49
1978 48401983 6343339 48472759.96 41545129.03 8715406.65 45008944.49
1979 46730553.33 5470955 48472759.96 41545129.03 8715406.65 45008944.49
1980 43845123.67 10535361 48472759.96 41545129.03 8715406.65 45008944.49
1981 40818640.67 6196657 48472759.96 41545129.03 8715406.65 45008944.49
1982 45810383.33 4463569 48472759.96 41545129.03 8715406.65 45008944.49
1983 45198443.67 3781045 48472759.96 41545129.03 8715406.65 45008944.49
1984 45089016.33 2101285 48472759.96 41545129.03 8715406.65 45008944.49
1985 46726720.33 7771010 48472759.96 41545129.03 8715406.65 45008944.49
1986 46578984 1651032 48472759.96 41545129.03 8715406.65 45008944.49
1987 44872522 5823326 48472759.96 41545129.03 8715406.65 45008944.49
1988 44862964 4833078 48472759.96 41545129.03 8715406.65 45008944.49
1989 43581924.67 4562448 48472759.96 41545129.03 8715406.65 45008944.49
1990 44471979.67 2141723 48472759.96 41545129.03 8715406.65 45008944.49
1991 45096895.67 2002991 48472759.96 41545129.03 8715406.65 45008944.49
1992 44015979.67 3171496 48472759.96 41545129.03 8715406.65 45008944.49
1993 45701485 1592322 48472759.96 41545129.03 8715406.65 45008944.49
1994 44816110.67 1889408 48472759.96 41545129.03 8715406.65 45008944.49
1995 43877959.67 3406431 48472759.96 41545129.03 8715406.65 45008944.49
1996 45096536.33 2788638 48472759.96 41545129.03 8715406.65 45008944.49
1997 44499508.67 4759492 48472759.96 41545129.03 8715406.65 45008944.49
1998 44813952.33 1762057 48472759.96 41545129.03 8715406.65 45008944.49
1999 43844289.67 963027 48472759.96 41545129.03 8715406.65 45008944.49
2000 44738855 4698555 48472759.96 41545129.03 8715406.65 45008944.49
2001 45034486.67 2726359 48472759.96 41545129.03 8715406.65 45008944.49
2002 44512401.67 3956045 48472759.96 41545129.03 8715406.65 45008944.49
2003 46829439 2628742 48472759.96 41545129.03 8715406.65 45008944.49
2004 45134885.33 4402515 48472759.96 41545129.03 8715406.65 45008944.49
2005 45559808 3378163 48472759.96 41545129.03 8715406.65 45008944.49
2006 45526618.67 3532290 48472759.96 41545129.03 8715406.65 45008944.49
2007 43817340 3037615 48472759.96 41545129.03 8715406.65 45008944.49
2008 46909030.33 5694332 48472759.96 41545129.03 8715406.65 45008944.49
2009 45028342 3694968 48472759.96 41545129.03 8715406.65 45008944.49
2010 47508592.33 2531736 48472759.96 41545129.03 8715406.65 45008944.49
And charts as :
As, however there is an open question concerning the influence of warming; choosing that period might be contentious in some quarters. Under these circumstances the lower mean period of ’88-02 disappears and we pick up extra aberrant points at 1979(Infringes upper limit on mean) and 1980(Infringes upper limit on range).
This method operates at higher confidence limits than the previous work I’ve put on this thread (.03% as opposed to 5%)
The base line is still being selected by eye rather than more rigorously, and it’s still from Hansen which will still be contentious in some quarters. In this case “rigourous” would require selection of a period that is mostly after significant “warming”
I’ve dropped 1981 from my first capability test run, although the standard industrial practice permits this.
End of boring bit……
So again record high winter snow decade that can be ascribed to natural variation. But i’ll 99.7% agree with you that its snowier if:
1) 2010/11 mean winter snow coverage exceeds 48473000 Km2
2)Mean winter snow coverage exceeds 45000000 Km2 for the next 5 years
3)2010/11 mean snow coverage exceeds 45000000 Km2 and the differnce between the lowest and highest winter coverage exceeds 8700000 Km2
Chances are it will continue to bobble on as normal. The usual 5 Qatloos on the outcome…?
One for the warmers is the question of why the output seems more regular and less variable after 20 years of AGW than before. I thought we were driving climate out of control…..?….:-)