While many science related government agencies are shut down (NASA GISS is deemed ‘non-essential’ for example) some remain open due to statements like this:
Due to the Federal Government shutdown, NOAA.gov and most associated web sites are unavailable. However, because the information this site provides is necessary to protect life and property, it will be updated and maintained during the Federal Government shutdown.
The NOAA Space Weather Prediction Center remains open, and they’ve updated their solar cycle progression graph set. Today, as we watch the sun we find only two small sunspot groups, both rather anemic.
![latest_512_4500[1]](http://wattsupwiththat.files.wordpress.com/2013/10/latest_512_45001.jpg?quality=83)
The latest data is not encouraging for Solar Cycle 24 as the SSN numbers have taken a pretty big hit. In fact, all the solar metrics have taken a hit at a time near the peak when their should be many more sunspots and indications of an active solar dynamo.
The SSN numbers for September dropped to about 37:
Radio flux is also down:
And the Ap Index, an indicator of solar magnetic activity is still bumping along the bottom. Compare it to the peaks seen in Solar Cycle 23 in 2004:
Clearly, we’ve passed solar max, as this magnetic field chart showing the magnetic filed has reversed (a signature of solar max) shows:
Solar Polar Fields – Mt. Wilson and Wilcox Combined -1966 to Present
From Dr. Leif Svalgaard – Click the pic to view at source
It seems that even though the solar magnetic field has flipped, predictions of associated climate doom have not come to pass.
Date: 07/10/13
Sun’s magnetic field about to flip, could affect Earth’s climate
The Sun’s magnetic field is soon going to flip by 180-degrees which could lead to changes in climate, storms and even disrupt satellites, scientists have warned. The Sun’s magnetic field changes polarity approximately every 11 years. It happens at the peak of each solar cycle as the Sun’s inner magnetic dynamo re-organises itself.
http://www.thegwpf.org/suns-magnetic-field-flip-affect-earths-climate/
Rather than an active flip, it’s more like the sun is rolling over and playing dead.
More at the WUWT Solar reference page
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
richardscourtney says:
October 10, 2013 at 7:27 am
It is true that “for the SUN the Earth-distance TSI is meaningless”, but it does not follow that it is meaningless for the Earth’s climate. I argue that the “Earth-distance TSI” probably has significant effect on the Earth’s climate
This thread is about the sun, not about the Earth. The climate models use the Earth-distance version as they should, On time scales of thousands of years the distribution of TSI through the year is very important [the Milankovitch cycles rely on this], but I still have to see any valid correlative claim of Sun-Weather-Climate effects depending on the Earth-distance TSI being different from the 1AU TSI. Perhaps you could show me some.
lsvalgaard says:
October 10, 2013 at 7:37 am
“I still have to see any valid correlative claim of Sun-Weather-Climate effects depending on the Earth-distance TSI being different from the 1AU TSI. Perhaps you could show me some.
Weather is clear – since the solstices are almost in phase with perihelion/aphelion and the Earth distance TSI difference for solstice-perihelion/aphelion is mere ~1 W/m^2, the Earth distance will not much change anything – for now. (We live in age of orbital forcing playing small role, so other factors of comparably smaller magnitude of effect can intermittently prevail).
But when they go out of phase (which is already happening – but we’re still more or less at the top of the precession cycle) it will not just change the seasons, but likely change the climate profoundly (alhough possibly not leading to global iceage in this cycle), because the total effective insolation of Earth will descend quite steeply (yes, the sinus is cruel at points around PI, especially when coupled with square distance) and exactly because of the variation of Earth distance TSI.
The effect in my opinion is more or less exactly opposite to what imagined Milankovitch – the cooling potentially leading to iceage is in my opinion not triggered by less insolation of northern hemisphere but by less insolation of southern hemisphere where is more ocean – surface water, which has very different optical and thermal properties than landmass materials, is much less reflective, which is aided by waviness, especially near terminator at high incidence angles, able to absorb considerably more solar radiation, to higher depths, with lower change of the temperature, given by higher heat capacity, and therefore considearbly lower radiative loss rate, which is even further underlined by the optics of the water interface with air, which much more than solar spectrum from outside, reflects back the IR from inside, moreover in very thin layer due to very low transmissivity for mid-IR, it aides ocean in spending more than third of the received energy on surface evaporation – transporting heat to atmosphere directly, lowering so considerably the lapse rate which further lessens the radiative loss by higher effective atmospheric temperature, slowing the heat dissipation.
The ocean having all this properties will get clearly less insolated during the year due to precession cycle – when summer solstice will get out of synchro with perihelion (every year adds ~16 minutes). The ocean which is not only more present at the southern hemisphere, but it is present there preferably at well insolated latitudes:
equator-tropics: the ocean is ~74.8% (~73% at NH, ~76.5% at SH) of the area,
tropics-polar circle: ocean is still ~70.5% (but ~48.3% at NH, ~92.7% at SH)
polar circles-poles ocean is ~48,5% (~73% at NH, ~24% at SH) , but mostly covered by sea-ice at least for some time in the year. (you can see for yourself in detail if you look into the data: tumetuestumefaisdubien1.sweb.cz/GLOBAL-OCEAN-LAND-STRATIFICATION-1DEGREEresolution.xls – graph: http://tumetuestumefaisdubien1.sweb.cz/SOUTHERN-NORTHERN-OCEAN-STRATIFICATION.png).
To imagine – now the Sun gets in zenit in time of perihelion at the tropic of Capricorn (imagine ~center of Australia), and Earth is closer to Sun so it gets there ~65 W/m2 more on top at solstice, then after the ~10000 years it will get this 65W/m^2 less there. Ocean reflects ~2% when Sun in zenit, while it reflects ~3% at the ~47° angle. Landmass reflects 7-40%, usually 10-20% – and especially much in the deserts around tropic – so when the Sun will be in perihelion around northern solstice African and Middle Eastern (twice the Australia area+) deserts will be even more hot, nevertheless directly reflecting 25-30% of the sunlight back to space, the far eastern deserts more north as well as North American West dtto. -And Kepler’s 2nd law will in my opinion only help the effect, not act against it as some argue, because it will make the perihelion summer over the northern deserts shorter. While south freezed, circumpolar conveyor slowed to arrested, seaice covering more and more ocean (which therefore stops absorbing the heat there) expediting the process by accelerating ice albedo positive feedback.
I very roughly estimate the apsidal/axial precession effect will alone lower the global effective surface insolation (I mean insolation which gets absorbed, converted to heat) 5+ W/m2 when globally averaged. I’m at the time unable to give more exact number, because I don’t have a reliable albedo stratification. But I think this suffices for the idea. The main “culprit” here is clearly the Earth distance TSI (with 1AU TSI clearly playing no role) in complicity with axial precession – it is not exactly a correlation study, but anyway I’m afraid that the Earth distance TSI was, is and likely will be the chief co-factor of most profound climatic changes this planet could experience, although the currently low eccentricity will reduce the effect.
lsvalgaard:
Thankyou for your response to my post at October 10, 2013 at 7:27 am.
http://wattsupwiththat.com/2013/10/07/september-solar-slump-continues/#comment-1443092
in your reply to me at October 10, 2013 at 7:37 am
http://wattsupwiththat.com/2013/10/07/september-solar-slump-continues/#comment-1443099
However, I admit to being somewhat puzzled by your reply.
Perhaps my post was not sufficiently on topic because as you say
However, your post I was commenting and your reply both refer to “climate models” which are of the Earth’s climate and not the Sun’s.
Also, your reply says
I see no reason for me to do that. I was responding to your points about TSI and “correlative studies for CLIMATE [30-yr average weather]”. My concluding two paragraphs said
So, I would be grateful if you were to discuss what I wrote and not what you would have preferred me to write.
Richard
richardscourtney says:
October 10, 2013 at 8:23 am
So, I would be grateful if you were to discuss what I wrote and not what you would have preferred me to write.
It is hard sometimes to distinguish because the topic in question is leading one’s thoughts in a certain direction which is perhaps not what was meant. My point was that there is a distinction between the climate effects of the orientation and orbit of the Earth and that of solar activity. The former effects are large [can lead to glaciations] and are undisputed. The latter effects are small and not even convincingly demonstrated as they tend to drown in the noise. Your ‘harmonic’ effects have not been seen as far as I know and are IMHO irrelevant for the topic of this post. Perhaps you could steer me to observations showing said effects.
Richard, the journals are fairly complete in their coverage of Earth’s distance from the Sun affecting its weather pattern variations. This is an intrinsic factor as in the Earth moves closer to the Sun, not the other way around. This distance factor (which is NOT a solar factor) is already a considered factor of GCM forecasts and projections, as is the cyclic change in TSI (which IS a solar factor).
What is your preferred Solar mechanism and how does it work to force temperature trends?
Dr. Svalgaard Since a magnetosphere is such a small object compared to the magnetic cloud any changes would be negligible. There are other much more important processes that influence the counterstreaming, e.g. reconnection within the cloud. The balance is really not an issue as the suprathermal electrons bounce back and forth between the two ends of the field lines many times [the bounce time at 1 AU of the order of a few hours].
That is really interesting. Ok, let’s for time being leave feedback to the sun aside.
Jupiter magnetosphere isn’t exactly negligible, solar wind at 5AU is at least 25 x (if not more) weaker than at 1 AU, and the J’s magnetic field strength is about 20 times stronger than the Earth’s (the J’s magnetosphere spatial angle could be up to 500 x the Earth’s).
If the first encounter’s reconnection alters the balance slightly, the subsequent bounces would increase the effect. Just speculating, but it is possible.
In reply to:
Leif Svalgaard says:
October 10, 2013 at 6:41 am
William:
General comment on how the solar cycle 24 magnetic cycle change will affect cloud cover on the earth: It is important to note regardless of the physical explanation as to what is currently happening to the sun it is a fact that as a result of the slowdown in the solar magnetic cycle and linear decay of the newly formed sunspots the solar heliosphere has weakened. As result of the weakening of the solar heliosphere there are now more galactic cosmic rays (GCR, mostly high speed protons) striking the earth’s atmosphere.
GCR strike the earth’s atmosphere and create ions that affect cloud formation and cloud optical properties. For some unexplained reason the cloud modulation effects of GCR was inhibited so that even though GCR levels where recently the highest every measured in the space age, the planet did not cool. Whatever has inhibiting the GCR cloud modulation mechanism appears to have stopped, there was record summer cloud cover which caused record low temperatures in the Arctic this summer and there was record sea ice formed in the Antarctic.
Based on what has happened before when there was a Maunder like minimum the same regions that warmed when the solar magnetic activity has high due to the solar magnetic cycle modulation of cloud cover will now cool as there will be an increase in low level clouds. Interesting there will also be a reduction in high level cirrus clouds.
The net effect of low level clouds is to cool the planet by reflecting sunlight off into space. The net effect of high level wispy cirrus clouds is to warm the planet by the greenhouse effect. So a reduction in cirrus clouds will cool the planet. Curiously it has been found that there was a reduction in low level clouds in the Arctic and a increase in cirrus clouds which coincides with the observed past warming in the Arctic. There is hence observational evidence that a significant portion of the Arctic warming in the last 70 years was caused by the reduction in low level clouds and increase in high level wispy cirrus clouds.
The cirrus cloud effect is not observed over the Antarctic ice sheet as it is so cold there is insufficient water vapour to form the cirrus clouds in the winter.
http://cosmicrays.oulu.fi/webform/query.cgi?startday=02&startmonth=09&startyear=1966&starttime=00%3A00&endday=02&endmonth=10&endyear=2013&endtime=23%3A30&resolution=1440&picture=on
William:
In reply to your comment: The observational anomaly is not that the solar northern hemisphere leads the solar northern hemisphere (thanks for the link I was aware that there was been and that there is currently a north/south asymmetry in the solar magnetic cycle that is not however the anomaly). The anomaly is that there are now no sunspots/sunspot groups in the northern hemisphere and there is evidence that the magnetic flux tubes are being shredded/torn apart in the solar northern hemisphere (See comment). i.e. The sunspot group formation mechanism is no longer producing sunspots in the northern hemisphere and it appears the solar southern hemisphere will experience the same fate.
http://sdo.gsfc.nasa.gov/assets/img/latest/latest_4096_HMIB.jpg
http://www.solen.info/solar/
What was happened in the solar northern hemisphere is what to expect in the solar southern hemisphere. If one looks at the solien’s overlay of the magnetogram view of the sun (The magnetogram is due to a magnetic fields effect on elemental emission lines in the solar photosphere. A strong magnetic field causes individual emissions lines of an element to split in two lines which is called the Zeeman effect. The emitted photons from each Zeeman split emission line have opposite polarity. A polarization filter can therefore be used to determine large regions of the sun what have magnetic flux loops passing through them).
Looking at the solen site’s solar magnetogram/optical overlay of the sun the following is apparent. The solar southern hemisphere large sunspots are being replaced by tiny sunspots and in some cases there is no visible sunspots at all only large magnetically active regions which is the residue that would be created if the magnetic flux tubes are shredded as they rise through the convection zone.
Comments:
There are two hypothesis proposed in this forum to explain why the magnetic field strength of newly formed sunspots is decaying linearly:
1) Tachocline Zone Solar Magnetic Dynamo Hypothesis: The current standard model (this is the standard model for the dynamo not William’s hypothesis) for the solar magnetic dynamo has the magnetic flux tubes which when released rise up to the surface of the sun to form produced in the narrow region (deep within the sun) that separates the solar convection zone and the solar radiative zone. It is hypothesized that the shearing motion between the convection zone and the radiative zone creates the magnetic flux tubes in the tachocline.
Theoretical calculations indicate the magnetic flux tubes require a magnetic field strength of around 10,000 to 30,000 gauss when released to survive the turbulent forces in the convection zone. There are multiple different possible explanation as to what could happen to the tachocline to explain the linear reduction in the field strength of the magnetic flux tubes (the tachocline is disturbed or narrower for example which reduces available time for amplification to create the flux tubes for example). Regardless of what creates the magnetic flux tube at the tachocline, that mechanism requires a magnetic field strength of 10,000 to 30,000 gauss to avoid being torn apart by turbulent forces in the convection.
One variation of the tachocline hypothesis has the remnants of the past solar magnetic cycle moving from the poles down into the tachocline. The remnants are then the seeds for the next solar magnetic cycle. This variation of the tachocline solar dynamo mechanism provides an explanation for the butterfly pattern as the solar cycle progresses.
2. Convection Zone Solar Magnetic Dynamo Hypothesis: This is not the standard model for the creation of magnetic flux tubes. It is believed this model hypothesizes that rising and falling regions in convection zone and other motions in the convection zone concentrates the existing magnetic field lines amplifying them.
The convection zone solar magnetic dynamo mechanism was abandoned by most solar theorists as the magnetic flux tubes are buoyant. Calculations indicate the magnetic flux tubes will hence rise to the surface of the sun and therefore cannot create the observed great concentration of magnetic flux which is called a sunspot/sunspot group which can have a magnetic field strength of up 3000 gauss. It is not clear how a convection zone solar magnetic dynamo mechanism could explain the butterfly pattern.
It is important to note that the observed magnetic flux loops on the sun are relatively shallow unlike the magnetic flux lines on the earth which pierce the core of the earth. This explains why sun can have for a short period of time two magnetic north poles.
Also it is important to note majority of the magnetic flux created each solar cycle is destroyed, cancelled or sent off into space.
lsvalgaard:
In your post at October 10, 2013 at 8:34 am you respond to my having said
By writing
Say what!?
I write “unless and until they are identified” and you ask me to “steer {you} to observations showing said effects”.
The important points are
1.
Oscillating systems usually exhibit harmonic effects so it can be expected that the climate system does for the reasons I explained.
And
2.
Any harmonics could be interpreted as indications of ‘something’ when conducting what you called “correlative studies for CLIMATE [30-yr average weather]”.
It is ridiculous to claim that the strong probability of harmonics is “irrelevant for the topic of this post” which is about identifying ‘drivers’ of climate. Observed correlations may merely be observation of harmonic effects. And, conversely, climate fluctuation can be ascribed as being a result of such probable harmonics in the absence of evidence to the contrary. But, as I said, such harmonic effects cannot be quantified unless and until they are identified.
Richard
Pamela Gray:
At October 10, 2013 at 8:42 am you say and ask me
The TSI variation as a function of solar distance IS a climate effect as I explained.
The incorporation of this in climate models does not interest me because those models are flawed in principle so should be scarapped and ‘done over’.
I do not have – and I do not know enough to have – a “preferred Solar mechanism” which works “to force temperature trends”.
Richard
vukcevic says:
October 10, 2013 at 10:06 am
Jupiter magnetosphere isn’t exactly negligible, solar wind at 5AU is at least 25 x (if not more) weaker than at 1 AU, and the J’s magnetic field strength is about 20 times stronger than the Earth’s (the J’s magnetosphere spatial angle could be up to 500 x the Earth’s).
We have been down this road a zillion times. What is important is the cross-section of the Jovian magnetosphere compared to that of a CME or Magnetic cloud. The magmetosphere is a thousand times smaller than the CME. To see it try to calculate the two cross-sections yourself; all the data is readily available on the internet [e.g. wikipedia].
If the first encounter’s reconnection alters the balance slightly, the subsequent bounces would increase the effect. Just speculating, but it is possible.
No it is not possible as the first encounter has a negligible effect to begin with. Even if the balance is altered that has no effect on the sun as the electrons mirror high above the sun’s surface.
William Astley says:
October 10, 2013 at 10:32 am
General comment on how the solar cycle 24 magnetic cycle change will affect cloud cover on the earth
Not relevant for discussion of how the Sun behaves.
The anomaly is that there are now no sunspots/sunspot groups in the northern hemisphere and there is evidence that the magnetic flux tubes are being shredded/torn apart in the solar northern hemisphere
Not anomalous as that happens regularly [and there are again sunspots in the north]. And ALL flux tubes are ALWAYS shredded/torn apart ALL the time.
It is hypothesized that the shearing motion between the convection zone and the radiative zone creates the magnetic flux tubes in the tachocline.
There are other shear layers: one just below the photosphere and one halfway down to the tachocline where the meridional circulation changes direction.
Regardless of what creates the magnetic flux tube at the tachocline, that mechanism requires a magnetic field strength of 10,000 to 30,000 gauss to avoid being torn apart by turbulent forces in the convection.
It is most likely that the sunspots are not created in the tachocline and ALL flux tubes are ALWAYS torn apart ALL the time.
One variation of the tachocline hypothesis has the remnants of the past solar magnetic cycle moving from the poles down into the tachocline. The remnants are then the seeds for the next solar magnetic cycle. This variation of the tachocline solar dynamo mechanism provides an explanation for the butterfly pattern as the solar cycle progresses.
Creation of the butterfly diagram does not require sunspots to be formed in the tachocline.
motions in the convection zone concentrates the existing magnetic field lines amplifying them.
The flux tubes are ALWAYS TORN apart and are not concentrated in the convection zone, but in the photosphere as is directly observed.
It is not clear how a convection zone solar magnetic dynamo mechanism could explain the butterfly pattern.
That it is not clear to you does not mean that it doesn’t happen.
richardscourtney says:
October 10, 2013 at 11:29 am
Oscillating systems usually exhibit harmonic effects
The climate is not an oscillator.
William Astley says:
October 10, 2013 at 10:32 am
The anomaly is that there are now no sunspots/sunspot groups in the northern hemisphere
Not anomalous as that happens regularly
Check out ftp://howard.astro.ucla.edu/pub/obs/drawings
I picked about 20 days at random. Of those, the following showed no groups in the northern hemisphere::
2005/9/14
2004/3/5
2004/6/8
2004/9/1
1983/4/1
1983/5/10
lsvalgaard:
At October 10, 2013 at 2:07 pm you quote my having said at October 10, 2013 at 11:29 am
Then you say
Rubbish!
I explained how and why the climate SYSTEM oscillates within each year such that global temperature varies by 3.8°C during each year. The claimed rise in global temperature anomaly over the past century is a quarter of the amplitude of that oscillation.
As I said, “oscillating systems usually exhibit harmonic effects”, and harmonics occur at multiples of the basic frequency. As you say, “The issue is about excluding possibilities” and, therefore, the possibility (i.e. strong probability) of harmonic effects needs to be excluded and not ignored.
And if you want to argue that the “climate is not an oscillator” then I suggest that you argue with Marcia Wyatt and Judith Curry because their recent paper in Climate Dynamics says otherwise
http://link.springer.com/article/10.1007/s00382-013-1950-2#page-1
This is the Abstract of their paper with emphasis added by me
Richard
richardscourtney says:
October 10, 2013 at 3:04 pm
And if you want to argue that the “climate is not an oscillator” then I suggest that you argue with Marcia Wyatt and Judith Curry because their recent paper in Climate Dynamics says otherwise
First of all they say “quasi-oscillatory”, second, ‘oscillatory’ means ‘wavy’, not implying a physical oscillation where there is a restoring force. What is the restoring force in the climate?
vukcevic says:
October 10, 2013 at 10:06 am
Dr. Svalgaard Since a magnetosphere is such a small object compared to the magnetic cloud any changes would be negligible. There are other much more important processes that influence the counterstreaming, e.g. reconnection within the cloud. The balance is really not an issue as the suprathermal electrons bounce back and forth between the two ends of the field lines many times [the bounce time at 1 AU of the order of a few hours].
That is really interesting. Ok, let’s for time being leave feedback to the sun aside.
—–
Ok let’s say that the magnetic cloud impacts Earth’s magnetosphere, (field lines) reconnection occurs, some particles stream inwards towards Earth on Earth’s field lines at the X lines. At the same time, from the same reconnection event other particles stream back towards the Sun from the same reconnection X line and make it to the corona or super halo region.
Vuks, kinda like a field line assisted acceleration, from Jupiter, but I don’t see it making it back up past 1 AU less maybe on the tailside of the bubble.. The magnetic cloud would have punched up a hole in the current sheath, so …why couldn’t a reconnection event just backhaul?
tumetuestumefaisdubien1 says:
October 11, 2013 at 6:14 am
So I checked the new data and the result was no differences
You also have to pay attention to the Version number. The current version is 14 and the data is different because the whole series was re-calibrated. That is why they publish the entire series every day instead of just the latest value for that day. The re-calibration makes a difference of about 0.3 W/m2.
I know I’m carper, why is it?
at that time the distance was changing during the day and TSI(Earth) was decreasing by -0.177 W/m2 per 6 hours. If there are missing values within the 6 hours then those will have to be dealt with as well. So there can be a different number of high-resolution values in each 6-hr interval. I think they calculate the daily values as the mean of all single high-resolution values and not of the four 6-hr values. It can be argued what is the better way of doing it, but we have to go by what they actually did.
lsvalgaard says:
October 11, 2013 at 7:01 am
“The re-calibration makes a difference of about 0.3 W/m2.
So does it mean the whole TSI level is even 0.3W/m2 lower?
“at that time the distance was changing during the day and TSI(Earth) was decreasing by -0.177 W/m2 per 6 hours.”
I thought it could be something like that, sine can be cruel at some points especially when coupled with square distance. 🙂 But how this can have influence on the Earth-distance TSI? The satelite is still in Earth orbit, which doesn’t make much difference in distance even from one side to another.
I think they calculate the daily values as the mean of all single high-resolution values and not of the four 6-hr values.
But shouldn’t be the daily sum of high-resolution values same for one day as the sum of the high-resolution values for the 4x6hours?
tumetuestumefaisdubien1 says:
October 11, 2013 at 10:19 am
So does it mean the whole TSI level is even 0.3W/m2 lower?
Yes the whole level is lower by a constant amount, approximately 0.3. I forget what the precise number is but it doesn’t matter as one simply discards the old values.
But how this can have influence on the Earth-distance TSI?
Because the distance changes during the day.
But shouldn’t be the daily sum of high-resolution values same for one day as the sum of the high-resolution values for the 4x6hours?
Well it is not the sums, but the averages that are important; here is an artificial data set made up with the following values (1,m,m,m,m),(2,2,2,2,2),(3,3,3,3,3),(4,4,4,4,4) i.e. with 4 intervals (the 6-h data); the ‘m’ means missing data. In each interval there are 5 high-resolution values [one 1 in the 1st interval, and five 2s in the 2nd interval, and five 3s in the third…]. The average values for the four intervals are clearly 1, 2, 3, and 4. The average of those 4 averages is 2.5. The daily average of 1,2,2,2,2,2,3,3,3,3,3,4,4,4,4,4 is 2.875
lsvalgaard says:
October 11, 2013 at 11:02 am
…But how this can have influence on the Earth-distance TSI?
Because the distance changes during the day.
Yes but thats what we actually measure with Earth distance TSI – I would think one should not correct the Earth-distance TSI for the distance, the satelite should measure what comes and that’s all – the satelite is at the Earth distance, the distance difference during the orbit is less than 10^-5 and the rest – that the TSI varies with the distance from Sun is actually desirable to be measured.
“Well it is not the sums, but the averages that are important; here is an artificial data set made up with the following values (1,m,m,m,m),(2,2,2,2,2),(3,3,3,3,3),(4,4,4,4,4) i.e. with 4 intervals (the 6-h data); the ‘m’ means missing data. In each interval there are 5 high-resolution values [one 1 in the 1st interval, and five 2s in the 2nd interval, and five 3s in the third…]. The average values for the four intervals are clearly 1, 2, 3, and 4. The average of those 4 averages is 2.5. The daily average of 1,2,2,2,2,2,3,3,3,3,3,4,4,4,4,4 is 2.875”
Is this really the method how the values are derived from the high-cadence record? This primitive method is quite exactly why there’s the significant bias of about 0.2W/m^2 when one does the average of the incomplete TSI record. (The 1360.8 v. 1361.0) It would be a bit a pity with such a good detector. I would be thrown from statistics exam straight away if I do something like this.
tumetuestumefaisdubien1 says:
October 11, 2013 at 2:57 pm
Yes but that’s what we actually measure with Earth distance TSI – I would think one should not correct the Earth-distance TSI for the distance, the satellite should measure what comes and that’s all – the satellite is at the Earth distance, the distance difference during the orbit is less than 10^-5 and the rest – that the TSI varies with the distance from Sun is actually desirable to be measured.
The measurements are reduced to the center of the Earth [as they should be]. The variation over the orbit is 0.127 W/m2 compared to the instrument precision of 0.007 W/m2. By reducing all measurements to the center of the Earth, anybody can easily calculate what TSI should be at any desired point without having to worry about what the orbit looks like [which may even change with time]. So SORCE is doing it correctly. They are doing everything correctly now.
Is this really the method how the values are derived from the high-cadence record? This primitive method is quite exactly why there’s the significant bias of about 0.2W/m^2 when one does the average of the incomplete TSI record.
There is no systematic bias. And the problem with missing data is very minor as missing data is rare. Column three gives the average time for the measurements within each interval. If that time is significantly different from the ‘nominal time’ in column two you know there were missing data.
I would be thrown from statistics exam straight away if I do something like this.
I don’t see why. This is physics not stats. Some people would fill-in the missing data with an estimated average value, but that would be making up data and not be good.
tumetuestumefaisdubien1 says:
October 11, 2013 at 2:25 pm
But when they go out of phase (which is already happening – but we’re still more or less at the top of the precession cycle) it will not just change the seasons, but likely change the climate profoundly (although possibly not leading to global ice age)
Of course, none of that is disputed [and thus not of any interest]. The important parameter is the 1AU value because that is the one that varies without our foreknowledge. From the 1AU value we can calculate the Earth-distance value at any time. The SORCE people are doing everything right with one possible exception [and I’m still mulling that one over]. They calculate the Earth-distance at the time when the photon left the Sun 500 seconds ago. I think it should be half that, 250 seconds. You may enjoy this little investigation http://www.leif.org/research/TSI-SORCE%20Friday%20Effect.pdf At an early point it seemed that TSI was different on Fridays. It turned out to be caused by calibration activity on Thursday evenings [Boulder time] that interrupted the regular cadence, meaning that simple linear interpolation to recover the few missing values wouldn’t work.
lsvalgaard says:
October 11, 2013 at 3:55 pm
“You may enjoy this little investigation”
Thanks Leif
lsvalgaard says:
October 11, 2013 at 3:28 pm
There is no systematic bias. And the problem with missing data is very minor as missing data is rare. Column three gives the average time for the measurements within each interval. If that time is significantly different from the ‘nominal time’ in column two you know there were missing data.
I just wrote a reply, but it somehow, disapperaed on posting and I’m already quite titred to rewrite it.
So just briefly – there are 2.5% missing values in the 6h record. They are often in block of multiple missing values and the holes have quite uneven distribution, so the dominating carrier signal is lost with the solar activity modulation – as I’ve found the missing values are significantly more in the “perihelion half” so if one does average, even with proper period, the resulting averages are significantly biased.
Some people would fill-in the missing data with an estimated average value, but that would be making up data and not be good.
Is there really so much difference between filtering distance signal out – in case of the 1AU and puting the same signal in -in the case of filling the gaps in Earth distance record – to at least obtain better absolute level average? I of course don’t defent those who make up the data out of thin air or bias them, but I’m not completely sure if the neutral signal filling is worse than do unweighted averages of averages of uneven weight populations and believe it is physics. I always thought the discipline which teaches how to handle data properly is called statistics.
tumetuestumefaisdubien1 says:
October 11, 2013 at 8:08 pm
but I’m not completely sure if the neutral signal filling is worse than do unweighted averages of averages of uneven weight populations and believe it is physics. I always thought the discipline which teaches how to handle data properly is called statistics.
As long as the unevenness is small enough the proper thing is to calculate averages of what was actually measured [reduced to 1AU and to Earth-Center. Statistics cannot create data where there is none. Anyway our little exercise here has shown that it helps to know a bit about the data and how they were aquired rather than just try to use statistics to calculate the averages incorrectly. The SORCE people do the best that can be done and their data set is superb.
In reply to:
lsvalgaard says:
October 10, 2013 at 2:37 pm
William Astley says:
October 10, 2013 at 10:32 am
The anomaly is that there are now no sunspots/sunspot groups in the northern hemisphere
Not anomalous as that happens regularly
William: The maximum for solar cycle 24 was in February, 2012 it is anomalous that their are no sunspot groups that are observable with a small telescope in the solar northern hemisphere now. Extrapolate what is happening to the sun. Professionally I am a senior specialist. I assist specialists in solving problems. I am given problems that others have failed to solve or when there is a crisis. One of the methods use to address complex problems particularly in a crisis situation is to keep all reasonably viable hypotheses on the table as evaluating the problem with different hypotheses may change the consequences and the appropriate action to address and/or manage the consequences.
We are the first humans to observe this particular solar change. Knowledge of how the sun has changed in the last 250 years cannot be used to support the assertion we know how the sun has changed in the last 12,000 years.
As I stated before, there is an enter different group of specialists that have discovered a set of quasar observations that indicate very large objects resist their collapse by a process that creates massive magnetic fields and charge unbalance. There are no black holes. The object that forms changes over time. There is evidence that charge is not conserved, that matter is created in the extreme conditions in these objects. There are massive jets and pieces of these objects that are ejected (knots in the jets). There is observational evidence that the quasars and the baby ejected quasars emit hydrogen and other elements creating astronomically large gas clouds and dust clouds. There is observational evidence of spiral galaxies evolving driven by this mechanism.
We cannot travel into the sun, cannot travel to directly observe other stars, as we do the sun. There are many assumptions concerning solar and stellar physics. The current solar and stellar model is based on limited knowledge, on assumptions. New observations such as an unexpected change to the sun now can change the solar model and the model for other stars.
There are sets of unexplained anomalies (in the paleo record of the earth, in our solar system, and astronomical observational anomalies) that all appear to be connected to the explanation as to what is happening to the sun now. Pretend or imagine that what is observed next is observational evidence that the solar magnetic cycle has been interrupted, followed by massive unusual solar activity.
It is a big deal for science if there is a change in the stellar models and cosmological model, a change in fundamental physics, and for humanity if there is an almost unimaginable solar change, a very, very, powerful solar change that will affects climate on earth.
William Astley says:
October 12, 2013 at 2:07 am
it is anomalous that their are no sunspot groups that are observable with a small telescope in the solar northern hemisphere now.
There is a nice sunspot group in the Northern Hemisphere right now.
We cannot travel into the sun, cannot travel to directly observe other stars, as we do the sun.
We can see into the sun, measure conditions in the interior, even rotation and plasma flows using the same technique as we use on Earth to detect oil and gas inside the Earth.
There are many assumptions concerning solar and stellar physics. The current solar and stellar model is based on limited knowledge, on assumptions.
The current models are based on solid physics. You say ‘assumptions’. Name one.
lsvalgaard says:
October 11, 2013 at 10:14 pm
“As long as the unevenness is small enough the proper thing is to calculate averages of what was actually measured [reduced to 1AU and to Earth-Center.”
It is improper in principle to calculate unweighted averages from incomplete populations with uneven weights.
Statistics cannot create data where there is none.
There are also other methods than fill in carrier signal (- which anyway must be exhaustively described for calculation of the 1AU TSI) to obtain valid averages. (But for obvious reasons the fill-in is the method of first choice -at least for preliminary purposes – for my purposes it suffices for now.) To calculate the averages without any correction for missing data making multiple percents of highly variable population, or even further leave out the another valid values is in my opinion definitely not good way to obtain a valid result in the degree of exactitude which even the incomplete data clearly still allow thanks to the exceptionally high instrument accuracy.
Anyway our little exercise here has shown that it helps to know a bit about the data and how they were aquired.
That’s what I’m trying to for some time – to find how what is available, what quality I can expect, etc. before I do some conclusions.
The SORCE people do the best that can be done and their data set is superb.
Here I fully agree, from what I’ve read about the instrument, data processing etc. it very much looks to me that the data is the by far best choice from what is available for the recent period. The potential further use of statistics to get from the data as much as possible is anyway only an option, which can be used or not. What is important are the primary data, acquired for as much of the covered period as possible. I find very unfortunate there are the problems now and I very much hope the SORCE team will succeed in solving them – do you have some interesting insider?
Btw. It looks like I solved the problem with the TSI-SST dependence calculation and as it looks now to me there’s another significant factor other than TSI for the 2nd half of 20th century SST rise and it even looks it a bit exceeds the TSI forcing when I use the HadSST3 data instead of the HadSST2. But I still must do the confidence analysis, so it is just what came out on the first sight.
tumetuestumefaisdubien1 says:
October 12, 2013 at 10:27 am
To calculate the averages without any correction for missing data making multiple percents of highly variable population, or even further leave out the another valid values is in my opinion definitely not good way to obtain a valid result in the degree of exactitude which even the incomplete data clearly still allow thanks to the exceptionally high instrument accuracy.
Without making up data or making assumptions about the distribution this is the best we can do. There are about 800 individual measurements per day. Now TSI is an average over the solar disk and that also changes with time [even on a minute basis if there is a very big flare], so we are better of not monkeying with the data. The completely RAW data is still available for special studies.
I find very unfortunate there are the problems now and I very much hope the SORCE team will succeed in solving them – do you have some interesting insider?
They are saving the battery power until they can get another spacecraft up so they can make in-flight comparison.