Late bets are closed!
Everyone please check to verify your bet appears. I accidentally set the “DateMetric” variable to ‘Bet Placed in June, 2012’ in the first form. I manually changed all bets actually placed in August so they should appear here. If your bet does not appear, let me know. Your bet will still be in the database, but a cutting and pasting error may prevent it from showing.
Preliminary observations: Some people who bet early appear to have “already lost”. That is: they have bet for an September average ice extent that is higher than currently indicated by JAXA. Mind you: The September average ice extent could hypothetically exceed the current ice extent, but that is highly unlikely. I’ve skimmed the database to verify none of the “June” labels were accidentally switched to “August”– so I think some people are a bit overly optimistic (or as I noted, bet early and did not revise their bet later on.)
The vast majority of late bettors bet for records low ice extents. Skimming by eye, it appears the lowest bet is that of “ggg” who placed a rather pessimistic the ice extent would be -1.5 million square kilometers; ‘porcozio’ placed the 2nd lowest bet guessing porcozio 0.1 million square kilometers. The 3rd lowest bet is for 2.55 million square kilometers placed by Patrick (Ice buster?).
All are well outside the ±2sigma intervals based on current Jaxa extent, but then so are some of the high extent bets. And anyway, who knows?
The full range of bets is below. If anyone wants a tally of bet/name/quatloo values to do statistics, I can create the .txt file. (Took me a while to explore the phpAdmin features, but it turns out to be easy to create without also exporting things like IP’s emails etc. I can do that after people have a chance to tell me if anything is screwed up below.)
[sockulator(../musings/wp-content/plugins/BettingScripts/UAHBets5.php?Metric=NSIDC September- average NH Ice Extent?Units=millions km^2?cutOffMonth=8?cutOffDay=23?cutOffYear=2012?DateMetric=Bet Placed in August, 2012?)sockulator]
Personally, I can’t see it going much below 0.00
Somehow the estimate I submitted of 3.74 (I think that is what I submitted) got missed
Rob– Click contact lucia. I don’t see anything with your email or under “Starkey” or under “Rob” that matches. I can add your bet– it will just be “late”.
Given that someone else bet 3.75, my probability of winning was quite low. I just wonder what I may have done wrong in submitting. Same name and e-mail address btw
Well… you could tie. But you’d be late you’d come in second. Also, I could make you 3.755 or 3.745 if DRUK doesn’t mind. We don’t want to pen him in too much– but if your bet really did get lost, I’ll add it late. That’s why I was showing these. (I can also look again. But really… I didn’t find it!)
Looks like my bet of 4.31 wins, or comes close! … As long as “September average” means “value as of Aug. 22”.
On the other hand, the rate of ice loss looks rather non-zeroish at the moment. Goodbye, quatloos.
Here is my attempt at the stats based on the larger list.
I have excluded ggg’s bet of -1.5, on the grounds that it is not possible (unless someone tells me it is), but not porcozio’s, although it too is probably not possible.
I haven’t included Rob Starkey’s late bet yet as I am not sure if it is going to be included, or what the value is.
The above/below 2007 count is based on my own calculation of the 2007 Sept. mean, of 4.381, although I am not sure if that is the official figure.
NO. OF BETS 51
MAX 4.820
MIN 0.100
MEAN 3.959
MEDIAN 4.090
STD DEV 0.712
MEAN PLUS 1 SD 4.672
MEAN MINUS 1 SD 3.247
WITHIN +/- 1 SD 44
WITHIN +/- 1 SD (%) 86.275
ABOVE MEAN 34
BELOW MEAN 17
ABOVE 2007 10
BELOW 2007 41
Lucia,
One thing that I have come to be interested in while looking at the ice is the difference between what sceptics, lukewarmers, consensusites and alarmists would predict as the year in which we get an ice-free day.
As an alarmist, I would predict it by the end of the decade, so somewhere between 2012 ad 2019.
The consensusites at the IPCC were predicting it somewhere between 2040 and 2070, although people from NSIDC are saying that some time around 2030 is a more likely date.
All else being equal, I would expect sceptics to predict that the ice will not vanish for a very long time, if at all – after all, those negative feedbacks will keep climate sensitivity to pretty close to zero, so the rapid warming that the Arctic has experienced due to natural variability will soon halt or reverse.
And all else being equal, I would expect lukewarmers to posit an ice-free September day to first occur somewhere in the latter half of the consensus prediction, so somewhere between 2055 and 2070.
Now, I am assuming that you, since you look at the data reasonably closely each year, would agree that the trends do not suggest 2055 and that somewhere around 2030 would be a much more reasonable pick. (Correct me if I am wrong.)
If I am right, it would then appear to me that the real argument between lukewarmers, the consensusites and the alarmists is about climate impacts, not sensitivity. If we come to a relatively close agreement on climate impacts – such as the Arctic will melt out completely one day in summer some year in the next 25 years – then our disagreements on climate sensitivity become pretty pointless.
Obviously, if the impact on the Arctic is the only area on which we agree (relatively speaking) then that is not much of an agreement. But perhaps we can extend that agreement. What is your take on the rapid changes in PDSI, for example (the increase in the average dryness of global soils)?
http://climexp.knmi.nl/get_index.cgi
(If the link does not work, pdsi and climate explorer get you to the page, I think)
And the link does not work. Try here:
http://climexp.knmi.nl/select.cgi?id=someone@somewhere&field=pdsi
No. I think it’s both. Among other things the Arctic is not the whole world.
David Gould (Comment #102101)-“it would then appear to me that the real argument between lukewarmers, the consensusites and the alarmists is about climate impacts, not sensitivity.”
These are not independent matters. The magnitude of impacts is greater with a higher sensitivity simply because the changes are bigger (actually, many analyses suggest net impacts change signs past a certain threshold of temperature change). That being said, that describes the real relationship that would exist. Many people may suppose impacts combined with amounts of temperature change well off the curve of the real relationship between the two. I would tend to argue that, even supposing climate is as sensitive as the “consensus” or “alarmists” believe, the “alarmists” (for the record, I prefer “the alarmed”) the impacts would not be as severe from such as many have suggested.
As far as the fact that there has so far been more ice loss in the Arctic than the “average” model, a loss of sea ice is not really an “impact”, but it seems unlikely that the trend is representative of what one can really expect for the future since it appears to be associated with enhanced North Atlantic warming over the last thirty years that probably can’t be sustained-relative to the rest of the world, anyway.
As for various things that do constitute impacts, there have been no cases I’ve seen where positive evidences of negative impacts is present.
Lucia,
I understand that the arctic is not the whole world, which is why I said:
‘Obviously, if the impact on the Arctic is the only area on which we agree (relatively speaking) then that is not much of an agreement. But perhaps we can extend that agreement. What is your take on the rapid changes in PDSI, for example (the increase in the average dryness of global soils)?’
As to it being both, yes, but if you believed that the world was going to warm by one to two degrees and I believed that it was going to warm by three to four degrees and yet we both agreed that the world was going to go through X changes, our disagreement over sensitivity would be irrelevant.
As such, I am looking at areas in which you, a lukewarmer, and I, an alarmist, might agree. If those areas of agreement start to pile up – arctic, drought, sea level rise, and so forth – then our disagreement over climate sensitivity starts to have less and less signficance.
So: when do you think that the Arctic will experience an ice free day in summer?
Andrew_Fl,
From that, I gather that you are not in agreement that the arctic will experience an ice-free day in summer at some point in the next 25 years, given that the natural variability is something that you think is going to slow, stop or reverse in the relatively near future. That is fine; I am looking for things on which some lukewarmers can agree with some alarmists (or alarmed – I do not care about which label is used provided people understand what is meant by it). There may be other things on which you and I can agree on. What do you think about the changes in PDSI, for example?
I note that the extent figure of 4.209 x10^6 km2, for August 24th was lower than the 2007 daily low figure of 4.255 x10^6 km2, recorded on September 24th.
Since the annual low is normally between September 9th and Sept. 24th, and the 80’s and 90’s average lows were on Sept. 11th, it would be exceptional if this were not the case this year, and we can therefore expect the daily record low to be beaten by a considerable margin this year.
My own crude calculations, on which my Sept. mean figure was based, suggested a low of 3.626 x10^6 km2, on Sept. 17th., but since the daily figure is already lower than my projection, I wouldn’t be surprised if it went as low as 3.234 x10^6 km2, i.e. about 1 million km2 lower than the previous record.
This would surely be a major “victory” for “alarmists”, and I wonder what others thought.
How does our disagreement over sensitivity become irrelevant? I think there will be less warming that you do. I’m not sure what X is above, but if you think the change will be X and I think it will be X/2, I think the change is less than you think it is. I think that’s “relevant” to our outlooks, which differ. Maybe you think it’s irrelevant, but all that means is then we also disagree about what is or is not relevant to diagnosing what we disagree about.
I think we may need to irrigate more places, we may need to shift where crops are grown and we may need to grow different crops.
Oh ice free summer….
http://rankexploits.com/musings/2011/connelly-dekker-bet-actually-robs-got-a-very-good-chance-of-not-losing/
This is the closest I can recall coming to a long range forecast:
Lucia,
I specifically said that we both agreed that the world was going to go through X changes even though we disagreed on the temperature change. I agree that if you think it is going to be X/2 and I think it is going to be X that we disagree, but in my proposal we do not disagree – we agree. And because of that agreement, our disagreement over temperature sensitivity becomes irrelevant.
Can you quantify that a little? If you take a look at the graph closely and do some analysis, since the major downturn started, the trend has been linear. Indeed, if you plot 10-year averages, the world slides into being on average in severe drought (PDSI -3 or worse) by the middle of the century. First of all, do you think that that kind of extrapolation is reasonable?
From just a graph, it is difficult to tell what you think. Do you think, for example, that the linear trend line is a more reasonable than the exponential?
David–
It’s hard to tell when I predict “ice free summer” because I haven’t made a specific prediction. I think there is way too much noise to project based on trend fitting and I think owing to the whole freeze/ thaw thing the physics are difficult too.
For your specific question: There is no exponential fit on the graph. The probabilities are in the legend. Of the tested models, the Gompretz fit best at the end of last summer; that’s in dark blue. Gompretz technically never hits zero. Under that model, we’d get zero ice sometimes owing to “noise”, but the mean trend is for a tiny bit of ice to always linger.
The other fits potentially predict negative amounts of ice eventually. My “best” is just a probability weighted fit– so eventually it predicts negative ice. Clearly, that can’t happen, and clearly trend fitting is going to be dicey for predicting the first ice free September.
I don’t know when we’ll have the first ice free day— but clearly before we have the first ice free September. 🙂
David Gould (Comment #102113)
Could you clarify what you are asking me? The first paragraph… I don’t know what you are trying to say. You want to think we agree about X even though I think the magnitude differs from what you think it will be?
Quantify what? Which graph?
David Gould:
Perhaps we’re back to the a disagreement over the relevance of an ice free period in the Arctic. Not only as an “indicator of climate change”, but in and of itself, what impacts it has:
How does a period of summertime ice free conditions impact us, and are all of those impacts bad (or even, do the negative impacts outweigh the positive ones)?
Serious question.
Seem now would be a good time to begin betting on an ice-free summer Arctic by 2020. Probability seems well over 50%.
That’s too far off for a bet here. But anyway, define “ice free summer”. Do you mean at least 1 day of ice free? Or do you mean no ice is seen during any summer month or maybe July, Aug, Sept? Or shift that?
It’s kind of interesting to watch the progression of ice loss.
Movie.
I ended up with 2-frames per second starting from July 1.
If we don’t get a shift in weather patterns, it looks to me like we could lose about 50% more ice before this is over.
We also started this season with good extent but very little ice thickness.
Here’s a movie, 7-frames a second (one week per second), going back as far as there is “data”. It’s sobering (in a way) to see how little ice there was going into this season that was more than 2-m thick.
The data were take from here.
I used the “nowcast” dates (this is partially model based so for each day, there is “backcast” and “forecast” dates).
And Patrick of porcozio will win the quatloos!
And my initial bet ends up between a factor of 2.4 too high… or so.
If so, you’ll be in good company!
Weather does tend to shift though. That said, it could take a turn for greater ice loss. We’ll see.
I make no predictions save that this year will be the grand minimum and that ice will increase from here out, due to the quiet Sun. Mind you, ice is being underreported this year because the widespread Alaska-Siberia ice is mostly slush and so being missed by the satellite microwaves — as Walt Meier (NSIDC) states at https://stevengoddard.wordpress.com/2012/08/24/ice-free-arctic-stupidity-continues/#comment-121498 . Dr Meier states he expects the slush to melt out in early September, whereas I predict it will not, so that the re-freeze in October will be spectacularly fast.
From the NPEO website the current air temperature is around 0°C at 82.8°N and it looks like there’s a high pressure dome camped over the pole at the moment (this is not that atypical).
The webcams are showing overcast skies, which of course reduces heat loss.
Looking the last major atmospheric heating event (which coinciding with the big Arctic Storm) most likely we’re looking at two more weeks of this, before something big comes along and kicks the high pressure dome off.
I say all of this with a certainty factor of ±2.4.
😕
Carrick,
There are positives and negatives from an ice-free arctic.
As an example of a positive, assuming that an ice-free September day indicates very low ice concentration through the NWP and the Northern sea route for some weeks prior, there will be significant savings in transportation for some the imports and exports of some nations.
A negative would be the effects that the ice-free arctic has on atmospheric conditions in the northern hemisphere. A massive amount of heat and moisture will be cycled through the Arctic ocean into the atmosphere as there is more and more open water not only at the end of the summer but during a significant part of the last month or so.
There is already evidence that this is having an impact in slowing circulation of weather systems, extending periods of drought and rainfall.
And I am looking for the link to that now – it was on Neven’s blog but there have been so many threads posted there in the last few days that I cannot find it quickly.
Lucia,
Regarding my question about quantifying, I would like you to quantify (very roughly) how bad an impact this drying is going to have. You talked about changing which crops we grow in some areas and moving some agriculture to other others. How difficult do you think those things are going to be for us over the next 40 years?
As an example, in Australia we do not really have too many options in terms of moving our agriculture. For many decades, governments and scientists have looked (not because of climate change; just because we wanted to be able to sell more food to the rest of the world) at using the north of Australia for agriculture. But the soils are just not suitable and any effort to create an agriculture base there would require billions of dollars of government infrastructure investment beforehand, particularly on dams. Does America, for instance, have areas of suitable agricultural land that it could move some of its agriculture effort to? Would it require billions of dollars of government investment?
The graph is the one that you can generate through climate explorer via the second link; this one: http://climexp.knmi.nl/select……field=pdsi
David, thanks for the comment. I ‘ll see if I can find that.
Regarding the US and sustainable agriculture, the short answer is we’ve spent many billions already on it.
What the US has as an advantage (west anyway, where it matters the most), is there are tall enough mountains to get a heavy snow pack in winter, even in otherwise arid regions, and this is used to supplement summertime rain via a series of aqueducts and reservoirs. There are areas that are semi-arid already in the US that rely primarily on irrigation, for example the Snake River Valley in southern Idaho. In a way, it’s easier to run a business in these areas, because you know ahead of time whether there’s going to be enough water available for your crops, than in areas that rely on the fickleness of nature to get enough precipitation that year.
As I understand it, the Natural Resources Conservation Service is where it centers (note they are run under the USDA, which also coordinates with local farmers in crop planning), but there are other organizations that are involved too, including the National Forest Service, the Bureau of Land Management and the Army Corp of Engineers, as well as state level projects.
David
Sorry, but I haven’t spent any time trying to quantify that and I don’t plan to. For some reason you’ve brought this up on a sea ice thread. It’s a topic I’ve never blogged on.
That depends on the precise way things change, but I imagine hte answer is it probably wouldn’t require billions in intervention to shift nebraska farms to South Dakota and so on. Of course, it’s a loss to people in Texas to lose farming, but it might be a gain to North Dakota.
Tomorrow I’ll take photos of our “drought stricken” back yard. It’s been raining the past few weeks. I wish I’d taken photos at the peak drought. But I think we are still officially in a drought. You may be surprised by how things look.
We don’t currently irrigate here. But you can imagine that a state whose states song begins “by they rivers gently flowing” can probably drum up water if we really have to do so.
Minnesota is “land of lakes”.
David–
Here’s annual rainfall for chicago:
http://www.climatestations.com/images/stories/chicago/chiprcp.jpg
Lucia,
I brought it up on the sea ice thread because the decline in sea ice and some comments about that decline by people who have previously self-identified as lukewarmers has made me suspect that many people are not as far apart as the lukewarmer/alarmist divide might seem to suggest, given that that divide is largely based on sensitivity difference. I was wondering about your position specifically and the position of some of the other lukewarmers who read and comment on this blog. The arctic and soil moisture/drought are two things that come readily to my mind, as those are things that I have spent time looking at.
Lucia, regarding drought, there’s speculation that ice free periods in the Arctic could influence rainfall patterns (David mentioned that above).
I’m not really sure how that’s supposed to work, given the amount of isolation from the rest of the system there is in the high Arctic. Probably we’ll find out first-hand soon enough if the last six or so years are any indication.
As David said, there is a paper out about the influence of the Arctic on the current extremes which are caused by stationary weather systems. Can’t find it now.
Here is the talk on the paper:
http://www.youtube.com/watch?v=4spEuh8vswE
Here is the abstract of the paper:
http://www.agu.org/pubs/crossref/2012/2012GL051000.shtml
I cannot remember where I originally found it now. I was sure it was at Neven’s, but I could not find it here. Google searching eventually got to it …
Oh, and the full paper is linked with the talk:
http://marine.rutgers.edu/~francis/pres/Francis_Vavrus_2012GL051000_pub.pdf
David
It’s not my intention to chastise you for bringing it up. That’s fine. I’m just explaining why I’m not going to attempt to quantify it. If you want to explain your ideas that’s fine. Maybe we can all learn something.
Sure. But here, David seems to be asking me to quantify my speculation. I’m not going to do that– but if he’s been looking at it, I’d be glad to hear his speculations. (And that doesn’t mean by link that says “read this”. If he has an idea, I’d like him to surround the links with words with some sort of summary of the bit in the link I’m supposed to get something out of. I clicked the link to KNMI and all I get is the KNMI screen I could gill in, click and get a graph. But I don’t even know which choices David wants me to make to create a graph that, if I created, might tell me what David wants to communicate. So I’m obviously not going to tell David what thoughts sprung to my mind when I was the graph. )
On this:
If this occurred, in Illinois, we would likely have to water bank. That is build “capacitance” into the system. The bank would catch water during floods and we could draw water at other times.
We already do this to some extent for flood control. (Rarely to draw water out in later times.)
Some long time farmers have always done it to some extent. That is: create or deepen a pond at a low spot and divert water there during floods to hold up water. Or sometimes it’s a matter of not leveling soil to get more growing area) But mostly it hasn’t been that worthwhile because it’s rarely required and not really worth the capital investments.
The village has been trying to encourage people to put in water catching systems to slow the rush of water to low spots during flood. They’ve explained we could use the water for our gardens. currently, almost no one does it. But if weather patterns did get more persistent, people would do it.
(My mother-in-laws mother had a cistern for her laundry business. That was back in the days when otherwise people used well water which was hard. Now we get lake water, so no one does it. But it’s been done before and if the weather changed, people would likely do it again.)
I’d go out and that the photos I promised David but…. it’s raining!
David Gould-As far as drought goes, this is an extensive study of drought globally over 1950-2000, during which there was warming:
Sheffield, J., K.M. Andreadis, E.F. Wood, and D.P. Lettenmaier. 2009. Global and Continental Drought in the Second Half of the Twentieth Century: Severity–Area–Duration Analysis and Temporal Variability of Large-Scale Events. Journal of Climate, 22, 1962-1981.
“Globally, the mid-1950s showed the highest drought activity and the mid-1970s to mid-1980s the lowest activity.”
In other words, the worst period for drought, during a period of warming, was before most (all?) of the warming over the period began.
Another thing that is important to keep in mind is that drought patterns are highly effected by sea surface temperature pattern variability (ie natural variability like ENSO). The Sahel, for instance, was relatively wet in the fifties and is one of the few large areas with an indication of significant drying-but this is not due to a continuous trend, rather it’s because the Sahel region was very dry in the eighties. That subsequently reversed towards more average conditions. The US has recently seen drought conditions reminiscent of those in the 50’s, but precipitation in the US has been increasing so the droughts can’t compare to the Dust Bowl even though the temperatures are similar or greater.
Lucia,
From wikipedia, North Dakota has about 90% of its land area in farms at the moment. As such, I am unsure of its spare capacity.
Now perhaps Illinois has a reasonable area of spare land available that has good soils that, for whatever reason, is not currently being farmed. In Australia, we farm every bit of available land, even pushing out into areas that are not really suitable for farming. But in the US things might be different.
Regarding the graph, I did already state what it showed: a drying trend that if it continues will see the globe as a whole be, on average, in severe drought. (PDSI of -3 or worse). My initial question was whether you think extending such a linear trend is a reasonable thing to do.
On the link, to see that graph (I cannot link to it directly for some reason) you just need to click ‘make time series’.
Thanks David.
I think I’d had to discuss this with somebody in the field to get their assessment, but I have some concerns.
1) As I’ve mentioned the air mass associated with the Arctic Ocean is largely decoupled from that over Siberia or Canada. It seems like a stretch to argue that summer-time ice loss could be related to polar amplification at much more southern latitudes in the fall and winter, and they mention this without attribution to any modeling study, so it’s hard to see this as anything more than pure speculation. (OTH, the later freezing of the Hudson Bay probably is implicated in a change in fall weather patterns in the US and Canada. That one is almost a no-brainer.)
2) They mention the decrease in pressure gradient between the pole and tropics resulting in a reduction of zonal winds. That would work except for the fact that the evidence is that zonal winds are limited in the current climate by eddy vortex shedding, rather than the magnitude of the pressure gradient from pole to equator (see e.g. Tapio Schneider’s work).
[A similar effect happens in your water pipe to your house, there is a maximum sustainable rate of water flow for a given diameter of pipe before the water flow becomes turbulent and impedes further motion.]
3. They fall into what I think is a trap in arguing about extremes. Extremes by their definition are uncommon weather patterns, let’s define them as conditions that happen 5% or less of the time. Let’s exclude temperature maxima, because those are simply a statement that the center of the distribution has shifted, and those of us possessing brains knows it’s warmed since the 19th century.
Further as has been beat to death (e.g. Taminos blog), looking at extremes of these sorts aren’t just not interesting because they are just a restatement that there is a trend, they are also sensitivity to the location of your baseline and sensitive to the interval being studied (these are both very bad properties to have for statistical tests).. Instead, we’re talking instead about extremes of the sort that increase variability such as flooding or drought.
In order to statistically establish the existence of a shift in the number of extremes associated with increased variability, if you are only looking at 5% of the events, you have only 0.05 the power of looking at shifts in the centroid of the distribution. Estimates have been made (Pielke Jr’s paper) that based upon the estimates of the increase in frequency of drought and heavy rainfall, that it might take as long as 200 years of measurement to quantitatively observe a shift in the number of extremes.
[High energy physicists know all about this, that’s why they are constantly trying shift to higher energy, higher flux beams in their accelerators, in order to get enough events to observe effects in the tail of distributions.]
Finally when somebody in science says “Particular causes are difficult to implicate” that means in plain English “it’s not an observable or measurable effect.” In a way that’s just conceding the point that it isn’t currently a measurable effect.
David Gould, look up the US Conservation Reserve Program. There’s a reserve that is approximately the size of the state of New York.
We’ve got it covered.
Also 90% of Nodak land being farm land?
Yes, probably true if you include the BLM and other ranching land, which is much of it, which has a density of about 1 cow per acre. Much of it is very arid in the current climate and not arable.
RE: David Gould (Comment #102160)
Historically that is not what happened so I wonder why some are suggesting it will happen now. During the last major Holocene warm periods where sea levels rose substantially, global climate was characterized by an increased water cycle, increased precipitation and abundance, not severe drought.
http://en.wikipedia.org/wiki/Older_Peron
http://en.wikipedia.org/wiki/Neolithic_Subpluvial
This is consistent with the physics of climate as we understand it.
David– I am also unsure of the spare capacity in ND.
As a general rule, in the US, owning to set asides, not all farm area is actually farmed. Some is fallow. (My sister– a physician– owns land she restored to prairie to get a tax break. The adjacent plots are soy or corn– depending on the year.)
Also, in some cases, a farm is utilized to ‘graze’ rather than raise grain/ veg etc. So, calories/acre can be modified dramatically if weather and profitability of various crops changes.
FWIW I’ve driven through SD, but not ND. (I think one of those states was left out of a big US map book and the governor joked that was ok by him!) It is vast emptiness….
There is some. But my point is more that what is grown can shift. Currently, Illinois is “corn, corn, corn, soy, soy, soy, corn, corn, corn”. But depending on amount of rain fall, etc. it can shift to wheat, sorghum, rye, potatoes etc. During the 80s, we did have a dry spell and one summer after 2 years of drought in a row, I saw sorghum. My reaction is “what the heck is that!?!?!”
All sorts of things might be grown, but the farmers pick what makes sense based on price and anticipated weather/climate. In Illinois at least, they could change crops or possibly switch to irrigation of some sort. Or a combination. Which they might do would depend on precisely where they are, amounts of capital required to irrigate etc.
I haven’t seen “the graph”. You left a link to something. I clicked. I saw no graph.
My general reaction to PDSI graphs before is to note that PDSI is not an absolute scale. If Australian and Illinois got the same amount of rain one particular year you might be listed as “wet” and we would be listed as “dry” on the PDSI scale.
So, for you in most places in Australia getting drier is certainly a bad thing for growing crops. But that’s not necessarily true everywhere.
Have a look at fields where they grow Minnesota Wild Rice (in the “land o’ lakes” http://bioweb.uwlax.edu/bio203/s2012/ziegler_emil/reproduction.htm
Drought — as defined by PDSI– would cut into the “wild rice” crop because the ‘fields’ would no longer be under water.
But it doesn’t mean farmers couldn’t pick a replacement crop if they knew they could count on those fields no longer being consistently underwater. (Minnesota is east of North Dakota. Watch the movie “Fargo”– most of it takes place in Minnesota. Then the main criminal flees to Fargo ND.)
So… I have no opinion on whether it makes sense to extrapolate PDSI. But more generally, I don’t even know that the whole word in drought– as defined by PDSI really means that we have less crop land.
I would need to know a lot more. If someone has figured that out, I’d be interested. But the last time someone showed me a PDSI plot, the conversation ended with the plot. But the impact of that…. crickets.
Anyway…In the US. I think those underwater wild rice fields count as “farmland”. For that matter, my sister’s prairie might count as “farm land” because if she wanted to she could dump her tax credit and start framing. But in the meantime, it may be fallow farmland. (I don’t actually know. I’d have to ask her.)
I went back and looked more and eating crow now. 90% of ND land is farmland according to this assay. (This incidentally covers technology improvement, so it’s worth a glance through just to see how an agriculturally advanced country adapts to challenges from weather/climate.)
It lists:
I’ve been through the western side recently, and what I saw looked pretty desolate (more typical of Wyoming). This is the sort of thing I remember.
Lucia, regarding rice, they are switching in Mississippi to strains that don’t need constant water. They flood the field a few times a year, then let it dry up. There’s a paper here.
ivp0:
I think the expectation is for greater precipitation, it’s just that the dry-belts will move further northward due to an expansion of the Hadley cell.
Whether that affects the US is a harder call, we’ve got this reserve called the Gulf of Mexico that is already responsible for pumping the central part of the US (and the Southwest in the summer) full of rain water.
Trouble is you can get intervals up to about 45-days where the steering currents maintain a stationary pattern, and if that involves a high pressure dome across the center of the country, that blocks normal weather patterns until it finally gets knocked off.
The other argument is as far as I can tell completely speculative, is the idea that the steering currents will weaken, and increase the amount of time these stationary patterns remain fixed over a given region (and hence by definition increase the frequency of periods of extreme precipitation and drought events).
So you could have a) a decrease in average rainfall in some regions and an increase in rainfall in the rest, and b) less predictable weather conditions making it harder for farmers to maintain solvency, e.g. increase in frequency of extreme weather conditions. (Hence my call for the need for more spending on weather prediction and possibly less on climate prediction.)
Carrick–
I was showing Minnesota wild rice which isn’t
related toquite the same as ‘normal’ (aka asian) ricehttp://en.wikipedia.org/wiki/Wild_rice
Lucia,
If you selected, ‘make time series; after clicking the link below you will see the graph. For some reason, I cannot link to the graph directly.
http://climexp.knmi.nl/select.cgi?id=someone@somewhere&field=pdsi
ivp0 and Carrick,
The thing is, over the last 50 years the world has gotten drier. According to NCAR, the region of the globe at -3 PDSI or worse (severe drought) at any one time has risen from around 12 per cent to around 30 per cent. Over the same period, the region of the world at +3 PDSI or worse has remained steady at around 10 per cent.
One of the other things about a warming world is that more water is held in the atmosphere. This means less is held in – for example – the soils. When it rains, it rains more heavily, of course. But while it is not raining, it is drier.
As to the graph, I understand that there would be quite a high level of autocorrelation, with dry years more likely being in clusters and the same with wetter years. That might reduce our ability to extrapolate.
(The percentage figures in the previous post may be slightly off, by the way – I am going from memory, as the papers on this I have at home. But they would not be too far off)
David
Carrick,
The CRP looks to be a fantastic initiative. Through that, the US has significantly more capacity to absorb long-term drought impacts than Australia.
David Gould:
But not outside of the bounds for the last 100 years, which is the more important fact of the two.
What you are saying is true, but we do know the long-term prediction of climate models is for a wetter world, not a drier one. (Though you probably know they are off by a factor of nearly two in underpredicting the amount of global annual rainfall.)
Colder climates tend to be drier and dustier, that’s something that seems pretty solid from the paleo data.
Sorry, but huh?
What graph?
David
Click on the link you left
http://climexp.knmi.nl/select.cgi?id=someone@somewhere&field=pdsi
It’s not a graph.
Lucia:
I agree with Lucia.
It’s a general assumption/result/whachawannacallit of the models that RH remains constant in a warming climate.
With more moisture in the atmosphere, which presumably comes out of a liquid form from the oceans to maintain a constant RH level, why does this imply less moisture content in the soil? Stoichiometrically it looks like the percent content of water in soil should remain roughly the same, as long as the assumption of constant atmospheric RH held valid. (Actually the pores expand in the heat, which seems to suggest you’d have a greater water holding capacity in warm temperatures than low ones, at least for soil that has vegetation with root systems on/in it to keep it aerated.)
I’ve looked at what the IPCC said about it, and the modeling didn’t seem to make the constant RH assumption. It seems obvious if you reduce the RH of the atmosphere, you’ll get greater net evaporation. However, warm environments need not equal dry environments. There are plenty of counter examples to that, and many very cold environments are very arid too, so the opposite may be actually the case here.
Lucia,
If you select ‘make time series’ after clicking the link below you will see the graph. For some reason, I cannot link to the graph directly.
http://climexp.knmi.nl/select.cgi?id=someone@somewhere&field=pdsi
Carrick,
According to the PDSI graph, the dryness is outside of the bounds of the last 100 years. Whether that is statistically significant, I do not know. But it is lower ‘now’ (ie, in 2005, when the graph ends) than at any time since 1880, which is when the graph begins.
David–
I clicked and I don’t know what the graph that is then plotted supposed to be. There aren’t enough details to explain what it is. (No… the mere word PDSI and time is not enough.)
I have, for example, found “Characteristics and trends in various forms of the Palmer Drought Severity Index during 1900–2008 Aiguo Dai1” which shows a bunch of graphs and is accompanied by words. And I’m not seeing the land getting drier and drier and drier. It also discussed computation of PDSI– and there seems to be a range of ways to compute it, each of which gives different answers.
So, why don’t you:
1) Take a screen shot of the graph.
2) Post it so you can link it in a comment.
3) Tell us in words what you think it is supposed to be telling us. (Definition of PDSI, who came up with the numbers in that graph etc.)
All of this would be much more helpful than just asking me what I think that graph means.
Lucia,
I have done as you requested. If you need any more information, let me know.
http://evilreductionist.blogspot.com.au/2012/08/here-is-graph.html
Lucia and Carrick,
Regarding moisture content in the atmosphere and soils, my understanding is that increased evaporation removes water from the soil, contributing to the increased water content in the atmosphere, and that increased evaporation is expected (and has been observed) in a warming world. My understanding could of course be wrong.
David regarding the comment 102186, I think that assumes constant absolute humidity does it not?
Also look at the NCEP global precipitation data (sorry don’t have the link at my fingertips). I think you’ll find it’s lower than the 1950s, but probably not statistically significant, due if for no other reason to the large amount of noise in precipitation data.
The advantage of this dataset over raw data sets is it is a physical-model based reanalysis. (I am sure EWMCP has one too, I would consider them the “gold standard” if you can lay your fingers on it.)
Edit: Whoops, NCEP goes to 1979. Figure. I see a weak trend, but nothing to write home about.
David, the other thing to keep in mind is there is a diurnal cycle. So if you put more water in the atmosphere during the day, that will be “reclaimed” at night as dew. Many plants of course are adapted to collect dew efficiently.
Just returning for the moment from the American SW (which is anything but dry right now, by the way), I was very reminded of that, seeing all of the succulents out there.
Ah here’s the ECMWF (not EWMCP not sure how that came out … alphabet soup) precipitation numbers:
Global precipitation
land only which is likely what “matters” for humans.
David
If the RH is the same over soil, what would drive the higher evaporation? While it’s true that higher temperatures will tend to warm the soil and evaporate water, the temperature oscillates. So, for example, we’ll often have soil dry during the day, but dew forms in the early hours of the morning and so on. And of course if RH over land is high, it can sometimes rain, which results in more water in the soil. Higher absolutely humidity ought to mean more water falling as rain or snow– so generally, one might expect higher temperatures with same RH to result in more precipitation and wetter soils. (Mind you– this might not work out entirely this way– but the argument has to be rather complex. More run off during violent storms maybe? something.)
I thought mostly, the increased water in the atmosphere was driven by evaporation over bodies of water like oceans and to some extent very large lakes. The moisture laden air them moves and often rains out over land where some ends up in the soil. (It may, of course, evaporate on dry days. But if the RH is the same, it’s not clear why it should evaporate so very rapidly.)
Lucia:
Ah, I realized that. Just pointing out the adaptations that are being made to reduce water consumption. Rice is usually thought of as a “high water demand” crop. Here in the South, even with plentiful water supply, they opt for the lower water usage route. I would guess without looking it up its because there is less agricultural run-off that way (less nitrogen influx into the waterways).
Incidentally, the high Arctic surface temperature is now below zero °C. Nothing to celebrate because 1) sea water freezes at -2°C and 2) the winter Arctic inversion’s likely set in, so a storm would mix the upper altitude warmer air in with the surface air, and I expect you’d see a sudden warming event. Some years, the ice just can’t win.
I didn’t realize until I started looking around that they actually thought inversions helped heat loss.
I don’t know anywhere that’s true. (Wrong sign… inversions stifle convective heat loss which is the predominant mechanism for heat transfer.)
Carrick,
I think that constant relative humidity does not mean that evaporation remains constant. The difference between the saturated vapor pressure and the actual partial pressure in the air is what mainly controls the rate of evaporation, all else being equal. So higher temperature and the same RH for sure leads to more rapid evaporation of water surface.
I rather doubt that matters very much in the real world though, since there are lots of factors which would seem to be important that are not considered by the Palmer index. If there were substantial uptrends in the dryness of (non-irrigated) farm land, farmers would complaining. I also suspect that the total need for water is gradually falling with rising CO2 because the plant stomata need to open less.
SteveF:
Yep. Agree with that, but it’s what happens over land that I’ve been grappling with.
Soil as you know is a very complex mixture, especially if there are organic cycles involved.
The point about the stoma is a good one.
David–
I do need more info. I have “Characteristics and trends in various forms of the Palmer Drought
Severity Index during 1900–2008” He’s got 4 fourms of PDSI. I don’t see that graph or discussion. Which PDSI is it? Does s/he show this graph of that data in the paper and discuss that data in the paper?
http://www.agu.org/pubs/crossref/2011/2010JD015541.shtml
Also, if we look at
http://www.cgd.ucar.edu/cas/catalog/climind/pdsi.html
we see
“Temporal (left, black) and spatial (right) patterns of the two leading EOFs of the monthly PDSI (normalized by its standard deviation prior to the EOF analysis). Red (blue) areas are dry (wet) for a positive temporal coefficient. Also shown in the left-lower panel (red) is the Darwin mean sea level pressure shifted to the right by six months to obtain the maximum correlation (r=0.62). Variations on < 2 year time scales were filtered out in the left panels. The percentage variance explained is shown on the top of the left panels. (From Dai et al. 2004) " The upper left seems to be the time variation of PC 1 of PDSI. Why does it go up and the graph you present go down? (For that matter when you say I can download that data at the link you provide... how? Which link do I look at? So I can look at the read me or try to sort it out relative to the paper?)
That’s the bit I’m not getting. I agree will get more total evaporation over water if the air is warmer.. If we didn’t we wouldn’t even expect constant RH. We’d expect a drop in RH.
But constant RH just sort of a leading order assumption.
But from the soil– I’m not sure. What is the partial pressure of water in damp soil? And how does water absorbed into soil move up as the top layers dry out? Lots of things are involved– capillary action. Plants behavior. So on.
Either way, Lucia, the Palmer Index (I had to look up what PDSI even was) is not global precipitation, which has stayed relatively static according” ECMWF.
The argument has moved now from rainfall to relative “dryness” of the soil. Given how little real instrumentation exist, and how idiosyncratic soil is, I really have no idea how you could get historical numbers dating back to 1950 that would accurately measure this.
If you want to restrict yourself to the satellite error, there are products that purport to measure it. See for example this.
Lucia:
You often get a “skin” called a “crypotogamic crust” on the soil that blocks evaporation (depends on the soil, how disturbed it is, yada yada yada). Bacteria and algae and the local flora (e.g. desert junipers) that cultivate them via nutrient addition to the soil are involved in this (so biologically you have an agent that is actively decreasing daytime moisture loss from evaporation, and by the way increases how much water gets absorbed.)
Here’s a teaser.
Q: How do you write down equations for this in a model?
Carrick,
Yes, evaporation from land surfaces is not simple. One factor which would seem to be important is the temporal pattern of rainfall. 3 cm of rain over 12 hours is for sure more absorbed than 3cm over 20 minutes. 1 cm each three days is not the same as 3 cm each 9 days. There should also be important regional differences ( ground cover, slope, etc.). A messy problem.
SteveF all of that and the history of the soil is involved.
Whether there are hoofed animals compacting it, whether the vegetation is otherwise disturbed, etc. Time of day matters. Even your example is complex, because if you have the right flora, it can accommodate 3cm over 20 minutes. Whether the flora is disturbed matters a lot in that case (which speaks to the effects of land usage changes by humans rather than anthropogenic CO2 forcing).
@carrick
Finally when somebody in science says “Particular causes are difficult to implicate†that means in plain English “it’s not an observable or measurable effect.†In a way that’s just conceding the point that it isn’t currently a measurable effect.
That statement was in reference to particular events, but their general point is that the conditions for these type of events to occur are strengthening.
Lucia,
The EOF is not the same as PDSI. They are looking at different things in the same data. Figure 13 in this paper seems to include it: http://www.cgd.ucar.edu/cas/adai/papers/Dai_JGR2011.pdf
It is also consistent with – for example – figure 9 in this paper: http://www.cgd.ucar.edu/cas/adai/papers/Dai-drought_WIRES2010.pdf
As to getting the data, there are a number of links on that page that go to the data in various formats. As I said previously, I do not have the skills and so forth to turn that data into an easily useable form for me, and at work I do not have the software, either. You would likely be able to do better with it, but you might not have the inclination.
SteveF–
Heck, the number of things that affect evaporation over land are vast. Did the rain fall on a flatish surface? Of a steep hill? Was the ground bone dry when it rained? Or a little moist? Is the ground clay? Rock? Sand? humus? Did you have mulch over the dirt? Heck.. a mist falling on too of my hostas might not even hit the ground. But if it rains hard, the soil under the hostas is shaded stays pretty moist for a long time.
Carrick,
The rainfall/soil dryness things are two different discussions.
I brought up rainfall in discussing the effect that changes in the Arctic could have – extending periods of low or heavy rainfall.
The dryness index was the second point I brought up (completely unrelated to the Arctic) in my attempt to see if there are any things which some lukewarmers and some alarmists can agree on.
David
How? Figure 9 shows 5 traces some going up, some going down and some going sideways. Which of the 5 traces matches your figure? And why is whichever one matches your figure the “right” one to think about?
I’ll read the Dai paper more tomorrow– but on a quick glance, I don’t immediately see which of the 5 or so PDSI’s is the PDSI in the graph you find at KNMI corresponds &etc. If that data is from Dai, I’d like to read what the author who created the data has to say about his own data.
bugs:
This is true if their analysis is right. That underscores the need to confirm the predictions of the model (if you want to call handwaving a model that is, model-driven hypothesis might be a better term).
The importance of being able to measure effects to confirm a hypothesis goes back to my comment about “post-modern science”, where people think that models can replace observation, oddly usually in areas where measurements are difficult to collect. I know it’s a comment you didn’t like the last time I brought it up, mostly because I think you lack the cultural framework to recognize how widespread this phenomena is, not being limited to climatey related things.
David:
Yes I remember this, and while I understand the general concept, I’m still having trouble putting together a model in my head that I think would work in practice.
The barotropic concept of cold arctic-warm tropics controlling the rate of large-scale overturning has been shown to be wrong for a couple of decades now, so it’s surprising to see it still present in this paper. (Interestingly the idea traces back to a model by Lindzen, which, while elegant was wrong, as I think Isaac Held put it in one of his papers.)
It turns out that the temperature difference between the surface and the troposphere near the equator is the main factor driving “real world” large-scale circulation (diabatic heating in other words). We had quite a discussion of this on Jeff Id’s blog at one point. Anyway this is important because it has the opposite sign (diabatic heating has increased while the equator-pole temperature has decreased).
Here’s one paper by Walker and Schneider on the topic.
Meh,
I can’t get too excited about the Palmer index because it runs so contrary to good paleo and historical evidence. Over the last 8000 years, when it was significantly warmer and seas were higher than they are today, it was wetter, with more regular rainfall, and plants, animals and humans thrived. We know warming is in the cards over time due to increased manmade CO2, but history tells us it certainly does not look like gloom and doom. In the absence of strong evidence to suggest that somehow this time it will be different, good times are probably ahead.
Lucia,
In figure 13, the solid red line looks pretty much identical to the graph that I am talking about, although it only starts at 1950.
And if you are concerned about the particular graph that I linked to, Dai’s papers discuss the observed drying. The graph is, imo, a summary of what Dai is saying. But it is not vital to the message.
@ivp0
You doubt studies that go back two thousand years, but are very confident of those that go back 10,000 years.
David-
Ok…. I’ll bite. I’ve looked at the paper. I haven’t noticed any message that would be summarized by the graph you posted and Dai elected to not include in the paper.
What do you think that message of the paper is? Where does Dai deliver it? (Or did he?) And why does he show all the other traces? Do they give messages too? Are their messages different?
bugs:
I’m guessing maybe because it’s a lot easier to get good precipitation proxies using trees, and perhaps because most of us consider the ice core dO18 data to be reliable as temperature proxies.
Carrick,
“Interestingly the idea traces back to a model by Lindzen, which, while elegant was wrong, as I think Isaac Held put it in one of his papers.”
And it is nothing unique to climate science. Elegant theories seem to take on a life of their own, being routinely referred to, even if a few simple experiments can proove them wrong. The “elegant but wrong” explanation seems too attractive to abandon for a “messy but correct” explanation. When it finds its way into textbooks, it becomes extremely difficult eliminate unless the field is “important”, since few will ever have the need for the correct explanation.
I have
I’m a bit perplexed by the projection section especally figure 11. I may be mistaken, but figures (a) and (b) would indicate that the PDSI estimates based on the multimoldel mean for the AR4 would get the results for 1950-2000 totally wrong. Then, he shows us projections into the future using the same models. Whatever PDSI really tells us (and I’m not sure what it does tell us at this point) why would we think the models are able to predict it into the future when they are totally wrong over 1950-2000?
Bugs-
It seems to me @ivp0 was talking about his reaction to the graph David is asking us to look at. Evidenetly, that graph is based on data from Dai– much of which was omitted from Dai.
The Dai paper on PDSI reconstructed based on surface temperatures and measured precipitation and cloudiness doesn’t go back 2,000 years. We don’t have measurements for any of those that go back 2000 years.
I don’t see any global PDSI reconstructions going back 2000 years in his paper at all. What I see is a reconstruction based on tree rings for west. n. American. This is (a) not the method used for the graph David is telling us to think about and (b) not global.
Figure 1 in Drought under global warming:
a review Aiguo Dai∗ which is based on tree rings shows nothing remarkable about the current level of PDSI in Western North America.
Being unconvinced that the graph David shows tells us much about whether the earth’s land became dryer on average over 100 years has nothing to do “doubt[ing] studies that go back two thousand years” because that graph is is not based on anything going back two thousand years!
Carrick (Comment #102205)
August 26th, 2012 at 9:34 pm
bugs:
We do have evidence that the stationary weather patterns are happening, the models are giving us the reason.
Interesting. The Dai paper does cite some actual soil moisture data.
I knew we had thermometers stuck in the ground at a various places and if I hunt around, I can find soil temperature someplace relatively near by. Looks like we have soil moisture detectors too. I wonder if other states too?
Ok… this strikes me as a reason to be dubious of any computed PDSI graph that predates 1948.
If I’m not mistaken, Dai forced the mean precipitation data from 1850-1948 to have a mean equal to that in the the 1948-1878 period. So, in the event that precipitation might have been lower (or higher) before 1948, this would not be reflected in his computed PDSI’s. Meanwhile, CRU is continuous.
On the one hand, I don’t know what else one might have done to create a graph going back to 1850. But on the other hand: I note that the PDSI’s computed for years before 1948 are omitted from DAI. He (she?) doesn’t discuss them in the paper at all.
What this leads me to think any answer to any questions where I compare estimated PDSI from 1900 to estimated values now is highly speculative.
(I also note that graph at David’s site is for -90N to 90S while Dai shows 60S to 75N. Dai also discusses computation of PDSI in the arctic regions as highly speculative. And I’m also a bit uncertain what the soil moisture value underneath piles of snow in antarctica would mean for the earth anyway. )
bugs
Obviously, stationary weather patterns exist. They always have.
Please point to the evidence that stationary weather patterns are happening at a different rate than in the past.
bugs:
Actually, we’ve been down this road… there isn’t any evidence for their hypothesis, just phenomena that vaguely could be related but with no statistical power associated with the observations, this is as much as admitted to in the paper, and the models don’t actually give you reason:
I see you missed that the explanation given in the paper is superseded by more recent modeling, such as the paper of Tapio Schneider that I referenced. Their explanation for the proposed mechanism was known to be wrong by 1990.
SteveF:
Yes indeed. Especially if the “messy but correct” version requires something more than hand waving, and doesn’t yield results that conveniently fit your narrative but instead probably contradict it, which is even worse of course.
That climate becomes more variable with warmer temperatures is a concept that is consistent with the strengthening of diabatically driven large scale circulation together with eddy-limited flow.
Suddenly we get “climate becomes more variable until it’s not”, with the “it’s not” magically appearing to explain what are likely unrelated weather events, themselves not even record setting ones, just above average.
(Expect a push back on this from the modelers, then expect to hear “strengthening of ENSO” as an explanation again. 😉 It’s weaker until it’s stronger. One size fits all.)
RE:
bugs (Comment #102210)
August 27th, 2012 at 5:13 am
@ivp0
You doubt studies that go back two thousand years, but are very confident of those that go back 10,000 years.
######################################################
Physics tells us that a warmer global climate will have a stronger water cycle with more rainfall. Paleo evidence tells us that warmer global climates in the past were wetter with more rainfall. Historical evidence tells us that past warmer climates were wetter with more rainfall. Anthropology tells us that warmer climates in the past were wetter with more rainfall.
Palmer is simply confused. Perhaps he was using inverted tree ring studies 😉
Lucia,
I think that the message from Dai’s work is very clear: global soil moisture has been decreasing over the past few decades.
Obviously, you disagree that that is the clear message. Thus, this is not an particular point on which you as one lukewarmer and I as one alarmist agree.
Thanks for the discussion. 🙂
David–
You showed a graph beginning in 1880 and asked an open ended question. I don’t know why you think I could have guessed what message you thought the paper or graph conveyed. That’s why I asked you what you though the message of the paper as conveyed by the graph was. (I have to say I am rather mystified that would chose to show a graph beginning in 1880 that was not included in a paper to supposedly convey that ‘clear message’ of the paper, particularly if the message is limited to the most recent few decades.)
If you want to find out what we agree or disagree on you actually have to say what you think. I can’t begin to guess if we agree or disagree if you don’t say state what your claims on. That woudn’t necessarily bother me– but in this case, you are the one who introduced the topic in a rather confusing way.
As far as Dai’s paper: I do think his paper (which does not show data back to 1880– or the graph you showed) does claim soil moisture over over areas not including the polar regions has been declining in the recent decades. How much? He equivocates by showing PDSI computed various different ways. Is it unprecendented? his paper doesn’t say. Will it continue– the peer reviewed paper also doesn’t say.
But I’d also say that I’d find the claim that the soil has dried more convincing if he showed measurements of soil moisture instead of PDSI computed 5 different ways. He sites soil moisture measurements (some from Illinois) but shows none. This makes it difficult for me to form much of an opinion.
lucia,
I apologise for it being confusing. It is sometimes difficult to work out what others will find confusing after you have been looking at something for a little while – things that seem ‘obvious’ are not so in many cases.
Again, thank you for your time.
David–
I think the combination of blog comments and thinking about something for a long time is to imagine you’ve actually expressed specific idea. Between the lag in time and attention, people don’t always know what you mean to say– the problem can be them, or you.
The paper does suggest there soils may be drying. I just don’t know how much confidence we can have in that method of detecting it. Unfortunately, if we assume the author got as much direct data as possible on soil moisture (e.g. measurements in illinois, china etc. cited above) the direct data don’t exist. So… that makes things difficult.
(I have to admit that I’m intrigued what little data exists seems to exist in my backyard. I knew when I’d read it that I thought “Why no direct soil moisture? I suspect we measure that.” Now… you might wonder why I suspected that. The reason I thought that is I knew I’d looked at Illinois soil monitoring data in the past… to decide how long I could risk leaving Dahlia tubers in the ground. I know that somewhere on line I can look up soil temperatures measured by buried thermometers in various fields around the state. I thought we had soil moisture data too. And it turns out my recollection must be correct because it was Illinois direct measurements of soil moisture that got cited. Seems we might be unique in collecting it because he didn’t have data from other states.
Oh– his data say Illinois is getting moister. I don’t know if that’s true either. (I don’t farm. So it’s not exactly data I looked up!) I’m almost curious enough to see if I can find the direct soil moisture data.
Lucia, if you don’t mind it not being fine-grained, there are satellite measurements of soil moisture. See this.
Carrick–
That’s pretty cool! It’s probably no less fine grained that on-ground measurements. Of course, they don’t go back forever.
@Carrick
The issue is not the heat getting from the mid latitudes to the poles, it’s the creation of the slow moving and stationary weather systems.
bugs:
They’re related concepts.
Heat is transferred from the mid-lattitudes in the atmosphere via large-scale circulation (polar cell). If you assume baroclinic flow, then the rate at which the transfer/circulation occurs is governed by the temperature difference between pole and mid-laitutudes.
If you have turbulence (eddy) limited flow, it’s not. Most older GCMs assumed baroclinic because they were too coarse to deal with eddies. (And it’s even a harder problem in the ocean as Isaac Held discusses here.)
Next time my bet won’t be so conservative. The ice mass must be less than I thought…..