Graphical Perceptions

This is just a couple of graphs that illustrate a couple of Very Important Things… that seem to be visible in the graphs but might just as well be nothing more than a matter of perception…

So I’ve pondered these for a bit of a while, and decided to just toss them here for general mulling over.

The first one we’ve seen before, but with time ‘going the other way’ and without a perception altering curve fitted into it. We saw it in in the 8.2 kiloyear event posting:

http://chiefio.wordpress.com/2011/12/13/8-2-kiloyear-event-and-you/

11,000 years of temperatures

11,000 years of temperatures

Original and larger images, or click this one for a modestly larger image

A nice graph, it shows the spike up of warming out of the Ice Age Glacial, then a massive spike down at 8.2 Kilo Years Ago, a return, then some wobbling sideways and a little trailing off. Time runs from right (old) to left (now).

This other version comes from a posting at Jo Nova’s. Notice that time runs the other way and it has an added reference curve fit to it.

Jo Nova time reversed GISP version

Jo Nova time reversed GISP version

Russ Steel has some history on this version here:

http://ncwatch.typepad.com/media/2010/02/global-warming-or-cooling-depends-where-you-start-on-the-graph.html

And, for reasons I don’t understand at the moment, I’m getting a “forbidden” on attempting to access the original Jo Nova link here: http://joannenova.com.au/2010/02/the-big-picture-65-million-years-of-temperature-swings/

One hopes the idiots on the Warmista side are not doing another hacking cracking site assault. (It really is a very stupid thing to do; at best it mildly annoys folks you are trying to convince of your merit, at worst it generates a LOT of press and attention for someone you would rather folks not visit. Just incredibly dumb to shout to the world: “I’m rude, stupid, and have no morals; trust me, not this honest and fair person I’m pissing at.”) Whatever the cause, time will tell.

What this posting is about is just comparing the emotional impact of those two graphs and the perception shift. I’d stared at the 8.2 Kiloyear version quite a bit. They eye tends to grab onto the peaks and bottoms and make two straight line references (that faint grid in the background); so the effect is of a “bounce up” from the bottom, then a ‘wobble in a range’, then a tail off to now and bit of a flip at the end.

The second one clips the start a bit (i.e. you don’t see the rocket up from the glacial). The effect is more one of ‘starting high’ and dribbling down. Now the curve is fit too, so you tend to see the peaks and bottoms as following a rolling off curve. More mental emphasis goes to the minor peaks and dips near the line, less to the Minoan Warm Period being a match to the first spike up after the 8.2 Kiloyear Event and that the cool period just before the Medieval Warm Period is a match to the bottom of the 8.2 KY Event. Basically, I notice the intermediate pattern of rolling over ALL the peaks more, and that the extremes limit at about the same place less. Interesting perceptual effect.

But there is another effect too. After the 8.2 KY Event, the rise is nearly vertical. As you look at other cool dips, the rise out is ever flatter. 4771 to the Minoan WP is about a 60 degree angle while the rise into the Roman WP is near a 45 angle. The MWP has a fast rise, but is becoming more of an isolated spike. Up and then immediately down to a nearly flat line of the last 600 years. Then the present is an even smaller blip, which the visual pattern would imply will be followed by a slightly down drifting segment as the ‘roll over’ continues.

Which interpretation is correct? Well, I’d vote for the second one based on orbital mechanics and interglacial histories. But the interesting thing to me is the difference in the “impression” the two graphs provide.

On To Mississippi Drying

This graph is from a paper that claims to match the stream flow in the Mississippi River with a strongly lagged solar cycle. One has to wonder why the lag, but I can postulate many possibles; from ocean lag times to lunar tidal forcing perhaps being a bit lagged but driven by the same engine of plant positions. What is interesting here is that a large offset IS applied, and folks tend to just accept the “wiggle match”. Is it right? Who knows… but the point here is about that effect of the visual:

Geomagnetic Solar Cycle vs Mississippi Flow

Geomagnetic Solar Cycle vs Mississippi Flow

What this implies about the future Mississippi Flow is that it will be way lower than the already low levels of now. For about the next 20 years. A 1930’s style Dust Bowl anyone? Personally, I think they have the lag wrong and more of a 10 year lag would still work, while having the Mississippi now being more or less ‘on schedule’…

From:

http://ks.water.usgs.gov/waterdata/climate/

Kansas District, U.S. Geological Survey
AUTHOR: Charles A. Perry, Research Hydrologist
Variations of Solar Activity Affect Regional Hydro-climate

This research focuses on the hypothesis that variations in solar activity affect regional streamflow. Variations in solar activity may control the amount of energy that reaches the Earth’s surface. These variations in solar energy may help create ocean temperature anomalies that can persist for years and move with the ocean currents. The ocean temperature anomalies can have an effect on meteorological factors such as atmospheric vorticity and moisture, both important for precipitation formation. Varying amounts of precipitation controls the regional hydrology including streamflow, groundwater, and lake levels.

Variations in solar activity are manifested in the magnetic structure of the Sun, total solar irradiance, and the strength of the solar wind. Solar wind speed and density directly affect the flux of galactic cosmic rays to the Earth and variations in the Earth’s geomagnetic field. Each of these parameters varies on an approximate 11-year cycle than can range from 9 to 14 years in length which over time displays a unique “fingerprint.” The long-term streamflow record of the Mississippi River at St. Louis, Missouri exhibits a similar “fingerprint.” The historic records of the geomagnetic index aa and 36 month moving averages of Mississippi River streamflow are compared in the following graph.

So now I’m trying to remember exactly what year the AA went to the basement and how far into this “30 Something” years we are now and how bad it’s likely to be how long going forward… all while trying to remind myself “it’s just a wiggle match”… but he does speculate some on mechanism:

The mechanism responsible for the linkage is thought to involve five important processes:

1. Variable solar activity causes variation in total solar irradiance (energy) that reaches the top of the Earth’s atmosphere. Variable solar activity controls the flux of galactic cosmic rays which may affect the formation of low-level clouds that control atmospheric albedo (reflectance).Variations in total solar energy reaching the Earth’s surface are amplified by changes in atmospheric albedo;

2. The variable solar energy is absorbed by tropical oceans creating large pools with different amounts of stored energy;

3. Pools of ocean water with varying amounts of stored energy are transported by major ocean currents (Ocean Conveyor Belt) to other global locations where;

4. Differential evaporation rates from oceanic areas alter global atmospheric pressure patterns (i.e., jetstream position and associated atmospheric vorticity);

5. These patterns dictate regional precipitation and temperature distribution and, consequently, the regional hydrology.

All well and good, and he cites a couple of papers to support the thesis. But then I look at the far right AA in the basement and the Mississippi already nearly shut down and I really really hope “something is different” during a grand minimum or the Midwest is just Royally Screwed for the next 20 to 30 years. Repeat after me: “it is only a wiggle match. it is only a wiggle match. I will not leap off the cliff of conclusion. It is only a wiggle match. It is…”

Back At Perception

Which brings me back to the issue of perceptions. A relationship is postulated. It is put in a graph that looks nice. It seems to fit the data in normal times. Then “something different happens” (and a Grand Solar Minimum” is quite different). Does it still hold? We don’t know. The data don’t cover that situation. But the graph leads us to a clear conclusion; despite the lack of suitable data.

So graphs are interesting, and very useful, and can give interesting insights into possible futures. BUT, it is very important to just look at the data and the proposed mechanisms too. Why? Because the mind brings a lot more to the graph than is IN the graph… Sometimes that is a feature; often it is not.

Subscribe to feed

About these ads

About E.M.Smith

A technical managerial sort interested in things from Stonehenge to computer science. My present "hot buttons' are the mythology of Climate Change and ancient metrology; but things change...
This entry was posted in AGW Science and Background, Human Interest and tagged , , , , , , , , . Bookmark the permalink.

11 Responses to Graphical Perceptions

  1. Fitting a curve to a set of points depends largely on what you think the relationship should be. Do you put a simple least-squares line, a parabola, an exponential or just connect the points? You then look at the line and think “yeah, that’s a pretty good fit” and it strengthens the idea that made you go for that fit. It’s easy to get fooled into going down the wrong path. Plotting the points and error-bars (note that these graphs don’t have error-bars) and eyeballing it can give a totally different idea than when it has the computer’s “best fit” applied.

    In the case of the Mississippi flow data, I personally think the fit isn’t that good, so I’d put it in the “unproven” bucket. I suspect there is some statistical maths in the fit here, but maths doesn’t always tell the truth. GIGO again.

    In graph-drawing, as in statistics, it’s easy to make a misleading picture of what’s there.

  2. Ian W says:

    There is a tendency in all these graphical (and other) approaches to assume that _the_ variable that is being plotted against the dependent variable is _the_ driver. The problem is that their may be several drivers all with different lags and effects. The circulation of the atmosphere is affected by more than ‘just’ the ocean heat content and will have feedbacks and chaotic response to the inputs so a strong input at the wrong time will have no effect while a weaker input at the ‘right’ time could have a huge effect. (Think about the timing of an ankle throw). This is lost in linear projections and curve fitting. So looking for simplistic wiggle watching correlations is probably that – simplistic and likely wrong.
    I get the feeling we should be using Poincare Sections rather than simply graphing variables.

  3. E.M.Smith says:

    @Simon:

    I’d not be as impressed with the Mississippi correlation were it not for the present ‘near halted’ flow rate and the AA in the basement. That can act as an ‘alignment marker’ if big enough. Now looking back at their alignment: Slide their cut off point for the Mississippi Flow to on top of the last spike of the AA graph. The rest of the peaks now line up about as well as they did before on some other peaks. But now the drop off of the Mississippi to date has a better match. Looks like about a 1 solar cycle (decade) offset to me.

    Of course, this is all just an exorcise in self confirmation bias based wiggle matching ;-)

    @Ian W:

    “Ankle throw”? Had to look that one up. Jujitsu, eh?

    http://www.ultimatejujitsu.com/learn-jujitsu-techniques/traditional-jujitsu-techniques/orange-belt/drawing-ankle-throw.shtml

    OK, I’ve done intro Aikido and Judo, some basic boxing, and wrestling (was pretty good at it in high school) and several years and belts of Karate, but not Jujitsu… I’d have likely walked right into that one with a lunge punch…

    OTOH, thanks to the Aikido, when thrown I just roll right back up to a standing position. Being thrown is just a bit of fun in the action… I know, not a Karate “slap the mat and absorb” behaviour at all… but I think it works better and it feels a whole lot better ;-) Also surprises the heck out of some folks….

    Then we have: Poincare Sections

    http://www.physics.emory.edu/~weeks/research/tseries7.html

    https://en.wikipedia.org/wiki/Poincar%C3%A9_map

    OK, that chaos theory stuff… I’ve read about it, but not gotten good at it.

    OK, score yourself a “2 fer” on things I didn’t know ;-) It’s been a long time since I had to look up a “2 fer” ;-)

    Both do look like useful techniques, though…

  4. BobN says:

    I am very skeptical of a lot of curve fitting. Over the years I noticed we always seemed to find a trend line when searching for solutions, only to discover the trend we saw was not there. It needs to be real strong before I agree to a trend line. Been Burnt!

    Speaking of Boxing, My buddy and I signed up for the golden Gloves in High school, not the brightest thing I ever done. I did have people marvel at how well I could take a punch without going out. I did get to be the corner man for my buddy and learned how to sew up a cut while doing so. He always griped I left a scar, I thought it looked great and was proof of my workmanship.

  5. Pascvaks says:

    Thoughts –
    The old TV series “Connections” has been mentioned a few times. As the season changes we notice ‘Signs”, or the signs make us notice changes. Those who put graphs together try to show us signs and connections and frequently give us their interpretation of what it all means; these can be outstanding new insights or absurd outright lies. (The Hockey Stick comes to mind for some reason;-)

    These days, people (in “Developed Countries”;-) likely can’t read signs or make connections, and rely on gurus and others of every stripe and trade to keep them pointed in the right direction. There’s a lot of ‘trust’ inherent in any civilization. (So if civilization has a tendency to dumb us down as a trade off to other ‘advancement’ then in 10,000 years of uninterrupted advancement people get kinda’ dumb?)

    Problems develop when people lie, or stupid people are believed rather than smart people. When politicians who are supposed to be leading us lie and we don’t find out until it’s way too late, that hurts much worse. This all started with some graphical perceptions then there were connections and then there were signs and then I got a picture, in my mind, and I remembered, and I thought, and for some strange reason I felt stupid and like a child in a wilderness, lost, and I wondered “Why?”

    We’re still in a wilderness. This one is ‘civilized’.

  6. adolfogiurfa says:

    I will reproduce here a commentary I have just made at Tallbloke´s talkshop, about TSI and “temperatures”:
    @Vukcevic: Is there anywhere data about the GMF of the same period?, just because following your ideas about the correlation between the GMF and temperatures (IR irradiance), it is like everybody is too much thinking about how much irradiates an oven without taking care of the electric resistances or whoever makes it work or the power which runs it. Just talking about temperatures or IRR is as intellectually advanced as a group of primitive men around a camp fire talking about their personal and subjective perception of warmth.
    Let us apply the socratic method of maieutics: What is it there behind such warmth being produced in the atmospheres of planets?

  7. Pascvaks says:

    Another Perception – Kids today don’t get into Science Fiction the way their grandparents and some of their parents did years ago. They much prefer to see the authors stand in front of a classroom and give them several daily doses in a number of subjects in person.

  8. Pascvaks says:

    PS: Or is it that the would-be author’s (who couldn’t write and sell a book if their life depended on it) prefer to talk to a captive audience of supple young minds who are forced to sit there? Hummmm.. gotta’ think about that a little more;-)

  9. davidatqcm says:

    Reblogged this on The Hobart Chinaman and commented:
    Recently, and foolishly I diiped my toe into the cesspit of the global warning [ no typos there] hoax? debate. D’oh! That was what I call an intro course in Silly101. i like data, here’s some, like all data its subject to interpretation, but one thing is incontrovertibly true …. there’s a lot of variation over time and what is happening crystal clear, and that is that the variations we are currently experiencing are within the historical range of climate variability

  10. Pascvaks says:

    @DavidAtqcm
    Beware of “Authorities Without Borders” — especially those who give ‘kindness’ and only expect respect;-)

    I guess I generally feel that we’re in a whirlpool, it all started when –with nothing better to do– a group began to collect and listen to a Hyde Park philosopher one day. Funny thing about people, if they have nothing better to do, they sometimes stay and listen, and listen, and listen. Think, maybe, something else will grab their attention soon? Something more important? I do;-)

Comments are closed.