From Otis Redding's Sitting on the Dock of the Bay
Sittin' in the mornin' sun
I'll be sittin' when the evenin' come
Watching the ships roll in
And then I watch 'em roll away again, yeah
I'm sittin' on the dock of the bay
Watching the tide roll away
Ooo, I'm just sittin' on the dock of the bay
Wastin' time
I left my home in Georgia
Headed for the 'Frisco bay
'Cause I had nothing to live for
And looked like nothin's gonna come my way
So I'm just gonna sit on the dock of the bay
Watching the tide roll away
Ooo, I'm sittin' on the dock of the bay
Wastin' time
Sunday, December 26, 2010
Sunday, December 12, 2010
Perhaps the best article about fat I have ever read
Everything you wanted to know about fat
I'm serious, I'm not talking about "fat people" really as much as fat itself in different forms and the risks possible for CHD based on consumption.
Also is a good example of evidence chain statistics.
I'm serious, I'm not talking about "fat people" really as much as fat itself in different forms and the risks possible for CHD based on consumption.
Also is a good example of evidence chain statistics.
Friday, December 10, 2010
Let's talk about climate change models
Seriously. It frusterates me a bit to hear terms like "sequestration" and "global warming" used and abused with no pretense.
So to clear a few things up.
Sequestration is a gerund-- that is, it's a noun that comes from a verb, and that verb is "to sequester"-- that word means "to take up." The arguement is often made that grasslands sequester more carbon than trees, and therefore we should convert forestland to grass land. Yes, grasslands do sequester more carbon than trees-- they are more metabolically active, and they grow and die quicker. Which also means that the carbon they take up is-- you guessed it, being returned right back to the environment when they die. Trees on the other hand live for a very long time, and although they aren't as active, consider the size of a bole as compared to the amount of grass that would cover that same basal area. Now consider the timber industry-- harvest that bole on a regular basis (carbon removal) and plant new trees (more carbon storage) and you've got yourself a good natural atmosphere cleaning factory.
But no matter the opinions on what to do in this situation: remember-- sequestration is only half (or less) of the issue.Storage is also to be considered.
Global warming-- now obviously we don't "know" the future. And that means that we are using "climate models"-- essentially we are taking data from the past and interpolating it to future conditions. So statistically speaking, models are already on the fritz of the whole extent issue, which is that a model can't realistically describe something outside of its spatio-temporal extent. But there are things we know about climate patterns (summers are hotter than winters, ENSOs and SOs and PDOs have a certain pattern of spatiotemporal ossillation) that we can use to guide our extrapolations. Or even more generally, for as long as we have record of it (or understand the processes by geological reasoning),climate has oscillated. Which means that just because we have one cold winter or one warm winter doesn't mean that a global warming model is valid or crap. To be honest, the best way I could think to determine if there was warming would be exactly what Dr. Keeling did (filter the Mauna Kea temperature data set), and that did show that the filtered trend increased. Whether 55 years was long enough to show a trend, I don't know.
In short-- one years specific weather conditions in any one place is not enough to validate or invalidate a model.
In short, remember that words are words (sequestration is not "good!" and harvesting "bad!") and models are models.Neither may be accurate but both are quite precise.
/rant
So to clear a few things up.
Sequestration is a gerund-- that is, it's a noun that comes from a verb, and that verb is "to sequester"-- that word means "to take up." The arguement is often made that grasslands sequester more carbon than trees, and therefore we should convert forestland to grass land. Yes, grasslands do sequester more carbon than trees-- they are more metabolically active, and they grow and die quicker. Which also means that the carbon they take up is-- you guessed it, being returned right back to the environment when they die. Trees on the other hand live for a very long time, and although they aren't as active, consider the size of a bole as compared to the amount of grass that would cover that same basal area. Now consider the timber industry-- harvest that bole on a regular basis (carbon removal) and plant new trees (more carbon storage) and you've got yourself a good natural atmosphere cleaning factory.
But no matter the opinions on what to do in this situation: remember-- sequestration is only half (or less) of the issue.Storage is also to be considered.
Global warming-- now obviously we don't "know" the future. And that means that we are using "climate models"-- essentially we are taking data from the past and interpolating it to future conditions. So statistically speaking, models are already on the fritz of the whole extent issue, which is that a model can't realistically describe something outside of its spatio-temporal extent. But there are things we know about climate patterns (summers are hotter than winters, ENSOs and SOs and PDOs have a certain pattern of spatiotemporal ossillation) that we can use to guide our extrapolations. Or even more generally, for as long as we have record of it (or understand the processes by geological reasoning),climate has oscillated. Which means that just because we have one cold winter or one warm winter doesn't mean that a global warming model is valid or crap. To be honest, the best way I could think to determine if there was warming would be exactly what Dr. Keeling did (filter the Mauna Kea temperature data set), and that did show that the filtered trend increased. Whether 55 years was long enough to show a trend, I don't know.
In short-- one years specific weather conditions in any one place is not enough to validate or invalidate a model.
In short, remember that words are words (sequestration is not "good!" and harvesting "bad!") and models are models.Neither may be accurate but both are quite precise.
/rant
Saturday, December 04, 2010
PCA for optimization?
To do a multiobjective optimization over chi-square statistics of simulated data versus experimental data, I wonder if it would make sense to weight the individual objectives using the eigenvalues (standardized or not) of the covariance matrix of the different stats? That is, use PCA to assign weights in a non-arbitrary way. Hmm...
Subscribe to:
Posts (Atom)