Русский New site

Advanced search

[ New messages · Forum rules · Members ]
Page 32 of 231«123031323334230231»
Forum » SpaceEngine » Off-topic Discussions » Totally off-topic thread (Talk about anything.)
Totally off-topic thread
WatsisnameDate: Tuesday, 29.01.2013, 00:56 | Message # 466
Galaxy Architect
Group: Global Moderators
United States
Messages: 2607
Status: Offline
midtskogen, your discussion of scientific practices is very good and I agree with you in all respects. Your question about scientists being unwilling to share data basically comes down to the period of time where one's research data is proprietary -- this is perfectly normal in science (and not just climate science), because it gives some insurance that the original researchers will be able to publish without someone else beating them to it. Another group can of course publish first, but they would be doing their work independently. Once it is out for publishing all the relevant material is generally made available to everyone, so that it can be checked.

The 'ClimateGate' scandal, although it did not find any objectionable practices by the scientists involved, has lead to a push for climate scientists to be more open with their work, and I don't think this is a necessarily bad idea.

Quote
In this respect I'm a bit uncomfortable about several of the temperature reconstructions, or rather that much depends on them.

Err, which reconstructions are we referring to here? I don't think there's any dispute over the last century of temperature because there are multiple reconstructions from independent researchers. (See Harb's initial post). Yeah, the details for how they are created are incredibly complicated, but the fact that the agreement is so good and that nobody seems to be coming up with temperature records that disagree significantly from them is rather remarkable.

Quote
Humlum can offer valuable additions to the understanding of our climate. For instance, I think he's been involved in the study of 1000 something years old plants found in melting tunnels under Svalbard glaciers which adds to the knowledge of how changing the climate is, but even though the glaciers evidently were much smaller or gone in relatively recent history at that place at that time, that and other pieces of evidence, cannot be used to tell the cold future of our climate which Humlum frequently is involved in.


I don't dispute Humlum's value to scientific progress at all; much of his work that goes into the big journals is very good. smile I just think his ideas on the relationships between CO2 and temperature have some slant.

Quote (HarbingerDawn)
The mass extinction level we've seen so far is probably almost entirely unrelated to climate. The potential connection between extinctions and climate change doesn't start until around now, and especially going into the future. Up to this point the primary causes have likely been other human ones, like deforestation and other habitat destruction, hunting to extinction, etc. I don't think that Watsisname, meant to imply that climate change was a major driving force in the extinction levels we've seen recently. More likely he meant that the effects of climate change - which are only just starting to manifest - could contribute to that.


Yes, I should have been much more clear. wacko My point was that the rate of species loss is comparable to rates during a typical mass extinction episode -- not the quantity of species actually lost so far. (That's generally considered to be when species loss is at 75% which we are not even remotely close to.)

And as for causes, I think most of the recent biodiversity loss can be safely attributed to alteration and destruction of habitat (particularly deforestation as you said). Hunting to extinction is another factor, but by the numbers I don't believe it is quite as significant. For the present, the effects of climate change are adding stresses to various ecosystems and the severity of this will increase in the future if warming continues unabated. And we may consider natural climate variability as well, particularly if we extend our view to the last glacial period.

There is some recent research on this subject, for example by Wake & Vrendenburg (2008) and Baronosky et al. (2011).





 
anonymousgamerDate: Tuesday, 29.01.2013, 01:58 | Message # 467
World Builder
Group: Global Moderators
United States
Messages: 1011
Status: Offline
Quote (werdnaforever)
If VII is full of lens flares and as fast paced as star trek 11


I'm not sure about the action pacing, but I'm sure there won't be as much lens flaring. Just look at the trailers for Into Darkness, it looks a lot toned down.





Desktop: FX-8350 4.0 GHz, 8 GB DDR3 RAM, EVGA GeForce GTX 1080 FTW 8 GB, 2 TB HDD, 24 inch 1920x1080 screen
Laptop: Core i5 480M 2.66 GHz (turbo 2.93), 8 GB DDR3 RAM, AMD Radeon HD 6550m 1 GB, 640 GB HDD, 17.3 inch 1600x900 screen
 
midtskogenDate: Tuesday, 29.01.2013, 10:29 | Message # 468
Star Engineer
Group: Users
Norway
Messages: 1667
Status: Offline
Quote (anonymousgamer)
Err, which reconstructions are we referring to here?

I gave one example, since it had a closer look at that one myself, but it's probably not unique. But even if there are many such records depending much on the intuition or guesswork of individuals, one should expect that these roughly cancel each other out on the global scale, so for the past century the consensus ought to be roughly correct. I'm a bit sceptical regarding the hockey stick, though. I find it quite remarkable that the global temperature supposedly has been pretty much dead stable for centuries, which I think require extraordinary evidence. And perhaps extraordinary evidence for the case that it's at all possible to deduct a millennial record with an accuracy sufficient to compare it to current observations (1 deg C or better). Since proxies involve more noise and the more proxies you add and the more noise you have, the flatter the signal becomes, perhaps? But I'm no statistician.





NIL DIFFICILE VOLENTI
 
HarbingerDawnDate: Tuesday, 29.01.2013, 14:33 | Message # 469
Cosmic Curator
Group: Administrators
United States
Messages: 8711
Status: Offline
Quote (midtskogen)
I'm a bit sceptical regarding the hockey stick, though. I find it quite remarkable that the global temperature supposedly has been pretty much dead stable for centuries, which I think require extraordinary evidence.

Here's a comparison of several reconstructions:



Who knows how accurate any of them are (time to build a time machine, eh?) but the point is that they do show variation over past centuries. It's not like "hey, everything was mysteriously exactly the same, and then industry happened and it got really warm suddenly". There were ups and downs and natural variations including prolonged cool and warm periods. What it does show is that the modern temperature record for the last century shows a rather obnoxious increase in temperature in a manner which is inconsistent with any reconstruction of past climate, and which cannot be explained without anthropogenic carbon dioxide.





All forum users, please read this!
My SE mods and addons
Phenom II X6 1090T 3.2 GHz, 16 GB DDR3 RAM, GTX 970 3584 MB VRAM
 
midtskogenDate: Tuesday, 29.01.2013, 15:46 | Message # 470
Star Engineer
Group: Users
Norway
Messages: 1667
Status: Offline
Harb, the red and cyan graphs show some variation (still less than a degree, and I would call establishing a figure for the global mean temperature for the years, say 770 to 775, with an accuracy of a single degree pretty remarkable). The other graphs show a ±0.1 or ±0.2 variation over a thousand years or more, which is quite incredible for a world that is very much alive. Extraordinary claims require extraordinary proof...




NIL DIFFICILE VOLENTI


Edited by midtskogen - Tuesday, 29.01.2013, 15:53
 
WatsisnameDate: Tuesday, 29.01.2013, 22:52 | Message # 471
Galaxy Architect
Group: Global Moderators
United States
Messages: 2607
Status: Offline
Quote (midtskogen)
Since proxies involve more noise and the more proxies you add and the more noise you have, the flatter the signal becomes, perhaps? But I'm no statistician.


Mmm, no.

A proxy is a measurement of a desired variable (temperature) through the use of some physical quantity that is correlated to that variable in a predictable way. Of course, as an indirect measurement, it can have (sometimes significant) uncertainty and thus error bars. When you reconstruct the temperature record from multiple proxies, you will get multiple curves representing the temperature record, again each with its own error bars. (You do not simply add them, and they do not smooth to a flat signal.) You can then do a statistical analysis of this collection to get the average, standard deviation, etc. By having multiple and independent proxies, you have increased confidence in the validity of your result. The better the agreement between proxies, the better your confidence.

The general concept here isn't too difficult to understand. You don't have direct measurements of the variable you want, but you can make measurements of other things that are related to it. The relation may not be perfect, and you do not necessarily know which of your proxies is closest to the true value. If you only looked at one proxy you might be way off the mark and not even know it. But by looking at several proxies together you can get a much better grasp of what the reality is.

Quote (midtskogen)
The other graphs show a ±0.1 or ±0.2 variation over a thousand years or more, which is quite incredible for a world that is very much alive.


I'm not sure why you find this surprising, unless you've never seen the reconstructed temperature record over geologic (way more than a few thousand years) timescales. The Earth's temperature record is very complex and has experienced incredible changes as well as long periods of relative stasis. smile There are variations on timescales from years, to millennia, to giga-years. There were times when the temperature was many degrees warmer than now and tropical forests extended into the polar regions. There were other times when the planet was completely glaciated. Nobody knowledgeable about Earth's climate claims that this is the only time the Earth's temperature has changed. But with respect to the this interglacial period, the temperature was fairly stable in comparison, as you can see from the data. And yes, we can achieve <1° uncertainties in this record.





 
midtskogenDate: Wednesday, 30.01.2013, 07:13 | Message # 472
Star Engineer
Group: Users
Norway
Messages: 1667
Status: Offline
Quote (Watsisname)
(You do not simply add them, and they do not smooth to a flat signal.) You can then do a statistical analysis of this collection to get the average, standard deviation, etc. By having multiple and independent proxies, you have increased confidence in the validity of your result. The better the agreement between proxies, the better your confidence.

I can see how you can extract a common signal from something noisy in some cases, but what about temporal noise, i.e. something that time shifts your signal in unpredictable ways?





NIL DIFFICILE VOLENTI
 
WatsisnameDate: Wednesday, 30.01.2013, 07:41 | Message # 473
Galaxy Architect
Group: Global Moderators
United States
Messages: 2607
Status: Offline
Sorry mate, I don't know the details for that. I imagine it's the same concept just on the time axis. E.g, a peak would have lateral uncertainty, but with multiple datasets you would hopefully get a Gaussian distribution and would take the median/average if you were interested in the time coordinate of that peak.






Edited by Watsisname - Wednesday, 30.01.2013, 07:44
 
midtskogenDate: Wednesday, 30.01.2013, 09:19 | Message # 474
Star Engineer
Group: Users
Norway
Messages: 1667
Status: Offline
Quote (Watsisname)
but with multiple datasets you would hopefully get a Gaussian distribution and would take the median/average if you were interested in the time coordinate of that peak.

If you do, and if you can. If not, I intuitively would think that you the more, perhaps multidimensional, noise you bring in, the flatter your result will be.

Admittedly, if you only look at the uncertainties/error bars in most of the hockey stick graphs, the hockey stick disappears. It makes me wonder if the hockey stick appears mainly because there may or may not have been 1 °C degree changes in the past, but the one in the past century is the only one that we know pretty much as a fact. If your 95% confidence range for a 5 year mean is a full degree, wouldn't the probability that there is a 1 degree swing increase quite a bit from 5% as you increase the length of your time axis? So in a timespan of 1000 or 2000 years it becomes reasonable to think that there could have been a few, but you can't plot them because you have no idea where they would be, but the last one in the past century you know of from direct sources, so that one you have to show, and hence the hockey stick. This is the big question that leaps into my face when I see these graphs, but I haven't really seen it discussed anywhere, except things like "statistical analysis shows", which is fine but have actual experts in statistics gone through this (which would require ALL data to be released, which we know from climategate that some have been thwarting)? After all, the people who did the research are experts in climate, not in every thinkable field that might be relevant. It's not easy to navigate in the minefield of statistics. "Lies, damn lies and statistics", etc...

Added (30.01.2013, 12:19)
---------------------------------------------
Published on forskning.no today (use Google translate, for some reason I couldn't get a google translate link to work from this forum). Not peer reviewed yet, but I suppose it will attract discussion.





NIL DIFFICILE VOLENTI


Edited by midtskogen - Wednesday, 30.01.2013, 09:22
 
HarbingerDawnDate: Wednesday, 30.01.2013, 18:03 | Message # 475
Cosmic Curator
Group: Administrators
United States
Messages: 8711
Status: Offline
I was having a bit of a discussion with some people over this, so all you math wizards (and anyone over the age of 12) please try to crack this problem, and - though I loathe saying this - please show your work.

6÷2(1+2)





All forum users, please read this!
My SE mods and addons
Phenom II X6 1090T 3.2 GHz, 16 GB DDR3 RAM, GTX 970 3584 MB VRAM


Edited by HarbingerDawn - Wednesday, 30.01.2013, 18:22
 
werdnaforeverDate: Wednesday, 30.01.2013, 18:16 | Message # 476
World Builder
Group: Users
United States
Messages: 897
Status: Offline
Quote (HarbingerDawn)
I was having a bit of a discussion with some people over this, so all you math wizards (and anyone over the age of 12) please try to crack this problem, and - though I loathe saying this - please show your work.

6÷2(1+2)

wizard
The order of operations:
PEMDAS: "Please excuse my dear aunt sally"; parenthesis, exponents, multiplication and division, addition and subtraction

6+2(1+2)
6+2(3)
6+6
12

Thus,
6+2(1+2)=12

I just hope it's not a trick question which I completely missed the point of.
 
Antza2Date: Wednesday, 30.01.2013, 18:22 | Message # 477
World Builder
Group: Global Moderators
Finland
Messages: 1049
Status: Offline
Quote (werdnaforever)
6+2(1+2)

It's 6÷2(1+2)
Not 6+2(1+2)





Go to antza2.deviantart.com for cool photos!
 
HarbingerDawnDate: Wednesday, 30.01.2013, 18:23 | Message # 478
Cosmic Curator
Group: Administrators
United States
Messages: 8711
Status: Offline
Quote (werdnaforever)
I just hope it's not a trick question which I completely missed the point of.

You obviously missed where that was division and not addition /facepalm

I increased the font size so that no one else would do that.





All forum users, please read this!
My SE mods and addons
Phenom II X6 1090T 3.2 GHz, 16 GB DDR3 RAM, GTX 970 3584 MB VRAM
 
midtskogenDate: Wednesday, 30.01.2013, 18:30 | Message # 479
Star Engineer
Group: Users
Norway
Messages: 1667
Status: Offline
Quote (werdnaforever)
12

I think ÷ is the divide operator. The insides of patentheses must be calculated first, so then you get 6/2*3. The question then is (6/2)*3 or 6/(2*3), i.e. whether to evaluate left to right or right to left since multiplication and division have the same precedence, but are not associative (except multiplication alone, of course). When operators have the same rank, the order is left to right. Therefore, the answer is 9 (not 1).





NIL DIFFICILE VOLENTI


Edited by midtskogen - Wednesday, 30.01.2013, 18:30
 
DoctorOfSpaceDate: Wednesday, 30.01.2013, 18:30 | Message # 480
Galaxy Architect
Group: Global Moderators
Pirate
Messages: 3595
Status: Offline





Intel Core i7-5820K 4.2GHz 6-Core Processor
G.Skill Ripjaws V Series 32GB (4 x 8GB) DDR4-2400 Memory
EVGA GTX 980 Ti SC 6GB
 
Forum » SpaceEngine » Off-topic Discussions » Totally off-topic thread (Talk about anything.)
Page 32 of 231«123031323334230231»
Search: