Temperatures dropped to a record low in Prince Edward Island overnight Tuesday, with reports of frost throughout the province.
An official record low of 3.8 C was set early Wednesday morning at Charlottetown airport.
The previous record for that date was 5.1 C, set in 2005.
Bob Robichaud, a meteorologist with Environment Canada, said that to his knowledge, frost has never been reported before in July in P.E.I.
“That 3.8 we got last night kind of sticks out as being lower than some of the other records for anytime in early July,” Robichaud told CBC News on Wednesday.
“So we’re looking at a significant event,” he said.
Environment Canada has issued a frost risk warning in low-lying areas of the province for Wednesday night. The temperature is expected to dip to 4 C.
The forecast for Thursday, however, calls for sunny skies and a temperature of 22 C for the province.
Tallbloke quote, “Anyone else who has clues should join in, this is an open source climatology which will be built through co-operation rather than the claim and counter claim of individual competing ideas.”
Maybe differing strengths of the earth’s magnetic field and it’s interaction with other bodies in the solar system, including the sun, could have an impact on this? Perhaps the answer to the problem will be found amongst the non-linear bits of ‘noise’.
Hi tallbloke,
In elliptic response to your interesting questions:
You may find this site to be a useful gateway:
http://hpiers.obspm.fr/eop-pc/
I will be bookmarking & sharing other links as time passes.
Here’s one now:
Benjamin F. Chao (2004). Earth rotational variations excited by geophysical fluids. International Very Long Base Line Interferometry (VLBI) Service (IVS) 2004 General Meeting Proceedings, 38-46.
ftp://ivscc.gsfc.nasa.gov/pub/general-meeting/2004/pdf/chao.pdf
–
The lazier reasoning around here:
1) “Correlation does not imply causation.”
2) “You have no mechanism.”
3) “It’s mostly noise.”
My response:
http://www.sfu.ca/~plv/ccLR1CRF.PNG
Tenuc (09:06:54) :
Maybe differing strengths of the earth’s magnetic field and it’s interaction with other bodies in the solar system, including the sun, could have an impact on this? Perhaps the answer to the problem will be found amongst the non-linear bits of ‘noise’.
Certainly does. Temperature seems to almost but not quite follow along in the aa wake. It’s the most fascinating puzzle I’ve come across in thirty years of looking at scientific data.
A very bright man called Brian Tinsley who Leif Svalgaard respects as a careful researcher published a paper last year on the Earth’s global electrical circuit. I couldn’t understand it, but I’ve never been great with working out electrickery. :o)
It’s behind a paywall, find a friend with institutional access.
Paul Vaughan, lovely graph, but you forgot the axis labels. Always playing the man of mystery. 🙂
I’m going to continue pondering ocean heat retention for now.
So far as I’ve been able to work out, the ocean’s thermal inertia works on several different timescales, because of stratification at the thermocline and the base of the near surface mixed layer and varying currents, heat input and rates of emission.
There is an 11 year solar signal visible in the SST data, which gets out of phase with changes in SST when the major ocean oscillations are changing phase. This signal represents an average 0.1C cooling and warming and is the transient response of the near surface waters.
Then there is a multi-decadal response to runs of high amplitude – short minimum solar cycles such as the ones we’ve had in the late C20th. This results in the increase in ocean heat content and thermal expansion we’ve seen between the start of XBT ocean heat measurement in the ’60’s and accurate satellite altimetry in the ’90’s and the levelling off of sea level rise and OHC post 2003. I think these runs of cycles are probably connected with the ~60 year cycles observed in the SST data for the Pacific and Atlantic oceans particularly. Going on the thermal expansion and heat content, I’ve calculated that the heat is stored down to an average of 1000m. The thermocline is deeper than this in some areas, shallower in others.
If as seems likely, we are heading into a period of lower amplitude – longer minima solar cycles, I predict this stored heat will be emitted from the oceans and will cushion the climate somewhat as global temperatures decline further. This graph I’ve produced is a preliminary attempt at understanding what is going on.
http://s630.photobucket.com/albums/uu21/stroller-2009/?action=view¤t=sst-nino-ssa.jpg
Since the heat accumulates in the ocean as evidenced by the thermal expansion picked up by the satellite altimetry, I’ve calculated a running cumulative total for sunspots, as a proxy for accumulated Total Solar Insolation (TSI), and compared it with SST from 1870 to date as a proxy for ocean heat content, plus the pacific decadal oscillation to help get a handle on how the ~60 year oceanic oscillation disturbs the correlation. One of the interesting features of the graph is the upturn in SST in 1910 near solar maximum, several years before the end of the negative phase of the PDO.
Below the thermocline, temperature tapers off towards the seabed in a slow fashion. Differences over time are small, but the volumes are vast. This will be where millenial responses are hidden. I haven’t had the chance to investigate those yet.
“The 1930s warming was part of a warming focused mainly in the northern high latitudes, a pattern reminiscent of an increase in poleward ocean heat transport (Rind and Chandler, 1991)”
“Prior studies suggest that the low precipitation during the Dust Bowl was related in part to sea surface temperature conditions over the tropical oceans (Schubert et al., 2004; Seager et al., 2005).”
http://downloads.climatescience.gov/sap/sap1-3/sap1-3-final-ch3.pdf
Siegfried D. Schubert, Max J. Suarez, Philip J. Pegion, Randal D. Koster, & Julio T. Bacmeister (2004). On the Cause of the 1930s Dust Bowl. Science 303(5665), 1855-1859.
http://www.seas.harvard.edu/climate/pdf/schubert_2004.pdf
“During the 1930s, the United States experienced one of the most devastating droughts of the past century. The drought affected almost two-thirds of the country and parts of Mexico and Canada and was infamous for the numerous dust storms that occurred in the southern Great Plains. In this study, we present model results that indicate that the drought was caused by anomalous tropical sea surface temperatures during that decade and that interactions between the atmosphere and the land surface increased its severity. We also contrast the 1930s drought with other North American droughts of the 20th century.”
Citations:
http://adsabs.harvard.edu/cgi-bin/nph-ref_query?bibcode=2004Sci…303.1855S&refs=CITATIONS&db_key=PHY
“[…] the development and maintenance of atmospheric ridges is the prime ingredient for drought conditions […]”
“[…] the longer the anomalous weather conditions persist, the more likely it is to have some stationary forcing acting as a flywheel (i.e., as a source for inertia) to maintain the anomalies […]”
“The droughts discussed above can be distinguished by their duration, with longer lasting events more likely involving forcing of the atmosphere.”
“The atmosphere does not have much heat capacity, and its “memory” of past conditions is relatively short (on the order of a few weeks).”
“Hence, the forcing required to sustain a drought over seasons or years would be expected to lie outside of the atmospheric domain; an obvious possibility with greater heat capacity (and hence a longer “memory”) is the ocean.”
“Therefore, most studies have assessed the ability of particular ocean sea surface temperature patterns to generate the atmospheric wave pattern that would result in tropospheric ridges in the observed locations during drought episodes.”
“For the longer-term events, the effect of steady forcing through sea surface temperature anomalies becomes more important.”
“The North Pacific SST changes appear to be the result of atmospheric forcing, rather than the reverse; therefore, if they are contributing to drought conditions, they may not be the cause of the initial circulation anomalies.”
“Because a large proportion of the variance of drought conditions over North America is unrelated to sea surface temperature perturbations, it is conceivable that when a severe drought occurs it is because numerous mechanisms are acting in tandem.”
http://downloads.climatescience.gov/sap/sap1-3/sap1-3-final-ch3.pdf
–
A lot of circular logic in ^there.
I think they need to look at this:
http://www.sfu.ca/~plv/ChandlerPeriod.PNG
I can show them a unique combination of events that correlate precisely with the ~1931-centred event. It is an event which has happened only once during the entire polar motion record (1846+).
More insight:
http://www.sfu.ca/~plv/ChandlerPeriodAgassizBC,CanadaPrecipitationTimePlot.PNG
tallbloke (07:34:16) :
That’s why the wiggles don’t match on scales of a few decades, but in the long run, the solar effect is clear.
Contrast this to all the claims that the solar effect is without or with only short delays. In the long run the number of degrees of freedom falls so fast that there is no statistical significance left. Even in the very longest run [billions of years] there is no solar signal, as the Sun has brightened 35%, but the temps have stayed much the same.
Paul Vaughan (14:41:13) :
“The 1930s warming was part of a warming focused mainly in the northern high latitudes, a pattern reminiscent of an increase in poleward ocean heat transport (Rind and Chandler, 1991)”
“Prior studies suggest that the low precipitation during the Dust Bowl was related in part to sea surface temperature conditions over the tropical oceans (Schubert et al., 2004; Seager et al., 2005).”
http://downloads.climatescience.gov/sap/sap1-3/sap1-3-final-ch3.pdf
Heat can only escape from the ocean at a certain rate, because the atmosphere acts like a blanket above it. When the solar cycles are running high, the heat which can’t escape upwards has to go sideways and downwards. A lot accumulates in the currently 400m deep Pacific warm pool. Judging by the high anomaly in the north atlantic in the latter C20th, a lot has accumulated there too.
Speculative idea:
The running count of cumulative TSI (sunspot numbers are a good proxy) was at a long time low point in the ’30’s, so if heat was being transported away from the tropics northwards up the Atlantic at that time. It would be due more to surface currents. Solar cycle 17 was the highest amplitude cycle in 60 years, but it didn’t start ramping up until late 1934. Perhaps due to the low state of OHC in the underlying waters there was a particularly low amount of cloud cover allowing the near surface waters to be warmed more in the tropics, with the rapid flush of polewards driven warm waters extending a natural frequency drought in the US.
Paul: How do you see the ~5% increase in the chandler wobble frequency centred at 1930 tying in? If it is mainly due to seabed level oceanic shifts as NASA says, could the longterm decline in my cumulative sunspot count be connected too I wonder. Maybe things start happening in the deep when the battery of heat energy above the thermocline runs low. The adjusted sunspot count was below 10 for 20 months between the end of 1932 and July 1934.
Leif Svalgaard (22:37:05) :
tallbloke (07:34:16) :
That’s why the wiggles don’t match on scales of a few decades, but in the long run, the solar effect is clear.
Contrast this to all the claims that the solar effect is without or with only short delays. In the long run the number of degrees of freedom falls so fast that there is no statistical significance left.
Hi Leif. You’re probably working your way down through the posts but in my post at
12:54:18 I note that the ‘at some times more obvious than others’ 11 year ocean response is the transient ‘short delay’ response of the near surface mixed layer. Under that is the longer term storage layer down to the thermocline. I agree that the signal from that isn’t clear in the surface data. I meant more that it’s clear to me that there is long term storage of solar energy in the oceans, and that this will be affecting our climates (plural).
Paul Vaughan has improved my 11 year signal graph:
http://www.woodfortrees.org/plot/hadsst2gl/from:1850/to:2009/isolate:156/mean:39/plot/sidc-ssn/from:1850/to:2009/scale:0.001/offset:-.05
I would suggest that during times of high solar cycles when the ocean is gaining net energy, the heat being stored below the surface layers doesn’t interfere with the ‘short delay’ surface response as much as when it’s being emitted during times of low solar cycles.
I’ll just add that stong events like the ’98 el nino will upset the phasing during high solar cycle runs too.
My key piece of evidence is that the satellite altimetry showing a rise in sea level and the SST rise of only 0.3C or so from 1993-2003 proves the oceans must be expanding due to longer term heat stored to a much greater depth than the near surface mixed layer.
Please do comment on that.
tallbloke (12:54:18) :
Quote “…A very bright man called Brian Tinsley who Leif Svalgaard respects as a careful researcher published a paper last year on the Earth’s global electrical circuit. I couldn’t understand it, but I’ve never been great with working out electrickery. :o)…”
Thanks for the heads-up on the paper. Not had time to read it yet, but the abstract certainly looks interesting – Quote “As it [downward current density] flows through layer clouds, it generates space charge in conductivity gradients at the upper and lower boundaries, and this electrical charge is capable of affecting the micro-physical interactions between droplets and both ice-forming nuclei and condensation nuclei.”
The link below provides an open source for this paper if anyone else is interested.
‘The role of the global electric circuit in solar and internal forcing of clouds and climate’
http://www.utdallas.edu/physics/faculty/tinsley/Role%20of%20Global%20Circuit.pdf
Also came across this one, again by Brian Tinsley
‘INFLUENCE OF SOLAR WIND ON THE GLOBAL ELECTRIC CIRCUIT, AND INFERRED EFFECTS ON CLOUD MICROPHYSICS, TEMPERATURE, AND DYNAMICS IN THE TROPOSPHERE’
https://www.utd.edu/nsm/physics/pdf/Tin_rev.pdf
Thanks again Tallbloke – looks like my bedtime reading for the next few days is sorted 🙂
tallbloke,
Your reasoned-speculation is interesting.
I will share a few notes:
1) The cause of the ~1931 phase reversal is still considered a mystery. If you dig around, you may discover that a conventional explanation for this is that AAM & OAM records do not extend back that far.
2) You inquire regarding Chandler frequency; rather than comment on that at this time, I will note that in the days ahead I hope to have some phase plots in presentable form.
3) I’m not on the same page with you regarding solar-terrestrial interpretations. This comment (which could be misunderstood) requires a polite elaboration: The woodfortrees graphs are interesting, but I wouldn’t expect a busy scientist to spend (much) time looking at that work in its current state of development. I’m thinking “cart before the horse” – i.e. rate of acceptance of new ideas depends on sequence of introduction.
4) If you are concerned about losing degrees of freedom (due to time-integration) (in the statistical sense – and assuming you are letting highly questionable [but unyieldingly – & broadly – conventional] assumptions slide), I urge you to reconsider what Currie (1996) has to say and to broaden your investigations beyond global averages to include local temperature ranges. You may recall a recent discussion of solar polar field strength where attention was drawn to some of the benefits of working with ranges vs. averages, such as cancellation of shared (& sometimes unwanted) features. [Note: Such technique can sometimes be applied between stations to gain spatial contrast (vs. blend) information.] This is not to say that range is a “superior” variable to average (for all purposes); it is, rather, to draw attention to the fact that range & average are different variables.
5) I encourage you to always be mindful of confounding (including the possibility of confounding of study variables with lurking variables […that you may not have even imagined]).
Re: rbateman (02:49:07)
Thank you for sharing this comment.
Here’s another signal spotted by Adam from Kansas, plus my comment.
http://icecap.us/images/uploads/amoarticlel.pdf
The interesting part is that there was an El-Nino within 12 months of the last four solar minimums and here it is happening again. It’s almost as if there’s an El Nino lying in wait after solar minimum arrives as of recent O.o
Very interesting observation. I think this is some of the excess heat stored below the mixed surface layers during a run of high solar cycles coming back out of the ocean when solar influence is at a low ebb.
When air temperatures drop, the differential between the atmosphere and ocean increases, allowing stored heat a quicker passage to the atmosphere. This will buildup a momentum of upwards moving warm water. There will be overshoot as it warms the atmosphere. Thus the ‘different’ type of el nino commented on in a recent article. Its all starting to fit together for me. Is anyone else getting it, or am I crazy?
Ok, my calculator is hot.
Please could someone help check the maths on these calculations, paying particular attention to orders of magnitude.
According to a world atlas on the web the area of the world’s open oceans is 335,258,000 sq km.
According to the satellite altimetry, the worlds oceans rose 32mm between 1993 and 2003.
According to the IPCC around half this rise was due to melting ice and around half was due to thermal expansion of the oceans.
335,258,000 sq km x 16mm = 5364.128km^3 expansion
According to an approximate average of the studies of ocean heat content done by Levitus et al, Ishii and Kimoto, and Domingues, the ocean heat content rose by around 5.5×10^22J over the same period.
Lets use Levitus et al’s 700 metre depth to obtain the volume of water containing this quantity of heat.
335,258,000 sq km x 0.7km = 234,680,600 km^3
The global ocean has an average temperature of around 17C and an average salinity of 35 psu. It’s average density is therefore around 1.025 kg/l
If we convert the volume to cubic decimetres and multiply by the density we should get the mass in kg.
234,680,600 km^3 = 2.3468×10^20 x 1.025kg/l
So the mass of the upper 700m of the worlds ocean is around
240547615000000000000kg or 2.405×10^20kg
(please check orders of magnitude carefully for me)
To raise water 1C requires 4.1855KJ/Kg so 5.5×10^22J is going to raise 2.405×10^20kg of water by
(5.5×10^19KJ / 2.405×10^20kg) / 4.1855 C
Which comes out at around 0.05C
There is a gradient in temperature from the surface down to the thermocline, which stays at a pretty constant temperature so I think this figure is reasonably consistent with a rise in sea surface temperature measured at 0.3C over the 1993-2003 period.
So the remaining question is how much the top 700m of ocean will expand if we warm it from 17C to 17.05C
According to a table I found at http://www.simetric.co.uk/si_water.htm 17C water will change in density from 0.998774g/cm^3 to 0.998757g/cm^3 for a 0.1C rise in temperature. This is a factor of 1.000017.
For a 0.05C rise we will get around half this factor so 1.0000085 will be our multiplicand.
1.0000085 x 234,680,600 km^3 (the volume of the top 700m of ocean)
=~1994km^3
Spread this over the world’s oceans and we get a sea level rise due to thermal expansion of ~6mm
But the satellite altimetry measure of sea level rise was ~32mm of which half was thermal expansion, so where’s the missing shilling?
If the altimetry is right, there is much more extra heat hiding in the ocean than a survey of the top 700m reveals.
Or if the Ocean heat content is right, the altimetry has overestimated sea level rise due to thermal expansion by a factor of about 2.7.
The other possibility is that I did my sums wrong. :o)
Place bets now!
I’d be grateful if someone else would run through the calcs and let me know.
Thanks.
tallbloke (06:51:48) :
According to an approximate average of the studies of ocean heat content done by Levitus et al, Ishii and Kimoto, and Domingues, the ocean heat content rose by around 5.5×10^22J over the same period.
For each of the three oceans [approx] for a total of 14E22 J…
Leif, with respect, that is the rise for the entire period of record from 1955, not the 1993-2003 period I’m studying. The more accurate satellite altimetry used by the university of Colarado came onstream in 1993, and the Levitus et al 2009 ocean heat content goes a bit haywire after 2003, doubling the rise of the previous 23 years by 2005.
http://sealevel.colorado.edu/current/sl_noib_ns_global.jpg
http://s5.tinypic.com/24v33t4.jpg/bmi_orig_img/24v33t4.jpg
Scaling off these graphs, the sea level rise is 32mm between 1993-2003 and the ocean heat content rise globally is 5.5×10^22J
I hope this clarifies the situation.
tallbloke (12:42:30) :
I hope this clarifies the situation.
Not at all. You show three rather different curves, telling me that the uncertainty is great, and there seems to be some picking here. For the red curve, I eyeball 1993 to be 3.3 and 2003 to be 12, for a change of 8.7E22 J. And what is the reason for Levitus going ‘haywire’. If you can’t trust them after 2003, how can you trust them before? The whole things simply looks too uncertain, including some of your assumptions [e.g. about gradients], to conclude that something is amiss.
Leif, I think you need to look again, you are looking at the red dot for 2004 on the levitus curve, not 2003, which shows around 8.5×10^22 from about 3×10^22 in 1993
Ishii and kimoto runs from around 2.75 to 8×10^22, and Domingues about 4.5 to a spike at 11which goes lower again the following year.
So,
Levitus et al 5.5X10^22
Ishii & Kimoto 5.25×10^22
Domingues 6.5×10^22 but averaging a bit lower.
I’m not trying to ‘pick’ anything, cherries or fights.
When I said Levitus et all graph goes a bit haywire after 2003, I mean in comparison to the other two studies. Do you think it’s credible that the rise in the levitus curve from 2003 to 2005 is equal to the rise from 1980 to 2003?
It maybe a splicing issue between XBT and ARGO data, or maybe Levitus got his sums wrong. It wouldn’t be the first time.
tallbloke (13:12:09) :
you are looking at the red dot for 2004
OK, but why do you trust their values before 2004, but not thereafter? Because they ‘don’t agree’?
And a minor point: Are the values plotted for middle of year or beginning of year? I.e. does 2003 mean 2003.0 or 2003.5? and for both all the series involved?
Leif Svalgaard (13:25:49) :
tallbloke (13:12:09) :
you are looking at the red dot for 2004
I looked again. The last red dot is 2008, then counting back 2007, 2006, 2005, 2004, 2003 would be the 6th red dot from the right. I read that [again] to be 12.
tallbloke (13:20:23) :
When I said Levitus et all graph goes a bit haywire after 2003, I mean in comparison to the other two studies.
They all look a bit shaky and the ‘spread’ is large. That combined with your various assumptions, make it a bit dubious to draw any conclusions. The order of magnitude is about right, that is how far I would go.
Hi Leif,
I worked on the 1993 2003 period because
1) All three studies are in reasonable agreement over this period
2) I can’t go earlier than 1993 because that’s when the Colorado satellite altimetry starts
3)The rises of Sea level, ocean heat content and SST are reasonably linear during the period.
I think you are right that the red dots are mid year, thanks for spotting that. That means the rise for Levitus et al is around 3×10^22 to just over 10×10^22, about 7.3×10^22
Averaged with the other two studies this will give (5.25+6+7.3)3 (allowing for a bit of averaging on the Domingues spike at 2003)
This gives 6.2×10^22 rather than the 5.5×10^22 I originally eyeballed.
This will make the rise of sea level due to expansion 6.75mm rather than 6mm.
So it’s around a factor of 2.37 too small rather than 2.7.
Did you run through the calcs for me to make sure I haven’t stuffed it up anywhere?
I’d be really grateful for confirmation.
Thanks
Leif Svalgaard (15:19:46) :
tallbloke (13:20:23) :
When I said Levitus et all graph goes a bit haywire after 2003, I mean in comparison to the other two studies.
They all look a bit shaky and the ’spread’ is large. That combined with your various assumptions, make it a bit dubious to draw any conclusions. The order of magnitude is about right, that is how far I would go.
I hear you, but there’s actually a bit more of a back story to this than meets the eye. If you confirm my calcs are ok I think I’ll write it up and put it online for others to ponder.
When you say ‘other assumptions’, which do you have in mind?
Size of ocean?
Salinity of seawater?
Some of you will wonder why the graphs stop at ~1984:
http://www.sfu.ca/~plv/ChandlerPeriodAgassizBC,CanadaPrecipitationTimePlot.PNG
http://www.sfu.ca/~plv/ChandlerPeriod.PNG
The Morlet 2pi wavelet is a bandwidth hog. This is the cost of frequency resolution. Anything beyond ~1984 is contaminated with edge-effect (at the timescale of focus) – until the future arrives – so it is cut off.
Cautionary Notes:
Most authors don’t snip off edge-contamination. This can be very misleading. Also, those who include a cone-of-influence don’t necessarily locate it where it should be for the wavelet they have chosen.
If anyone wants to know what the Chandler wobble period has been doing since 1984, they can see the work of (for example) Richard Gross (Nasa JPL).