Thread for all the board liberals

[vy65]

ElNono

[Cry Havoc]

That makes no sense. Climate *is* weather. Global climate is the sum of all those "highly, locally variable" measures. Measures that comes from a lot of different sources: direct temperate samples, sediments, fauna, etc.

They're volatile sources, which is what makes the model extremely complex.

The "100s, 1000s of years of historical data" can only get you a 5 day forecast with any degree of quality, tops, on a state of the art model.

You can't trust Accuweather to tell you what's the weather going to be next week, but we gotta take your 10-20 year projection seriously? Ridiculous.

Now, if your argument is "from a risk assessment perspective, we should probably strive to be good to our planet", then sure, that makes sense.

But there's a big jump from "strive" to "force"... and the error rate on the projection simply does no warrant that jump, IMO.

I actually have a background in Meteorology, so I feel safe to address this.

The reason our short-term forecasts have a variance in precision is very straightforward. It's almost entirely because of cities. Even massive cities are relatively tiny areas of land. Getting a forecast to be exactly right within a tiny area is insanely difficult because it's not just about forecasting that conditions are right for rainfall, it's getting the rain border within a few square miles of land. That requires an immense amount of data. It is very difficult to acquire immense amounts of data for short-term forecasts, as the situation is constantly changing.

To put this into programming terms, you're comparing the following scenarios:

1) Predict how long it will take to debug a program that is acting completely screwy. Predict what bugs will happen and how long they will take to resolve.

That's pretty difficult to give any certainty to. Versus:

2) Give me an estimate for how long it would take to build a specialized system for, say, hospital equipment in SQL.

The second task is DECIDEDLY more complex and involves many more moving parts, but you could probably give a reasonable estimate of conclusion, versus trying to give a micro-managed level of precision to a specific set of variables that are often changing on the fly (99 little bugs in the code, 99 little bugs).

Notice I used the word precision. The problem is that you're looking for two different levels of measurement, thinking they are direct correlations. We can collect massive amounts of long-term weather shifts, far more than we can for, say, what it's going to be like in Austin on Sunday at 8pm. Yes, we have terabytes more data these days than we did 20 years ago, but it still pales in comparison to how much our climate models are absorbing and responsible for.

And with that said, I still wouldn't call our long-term forecasts precise. They are accurate to the extent that we have an idea of how bad things will be, but we can't for sure say that, given current indications, Miami will be flooded in March of the year 2112. We CAN say that it's likely that water levels will rise by "x" amount of feet due to long-term manmade global warming, and you can extrapolate that from what you will. But that's the difference between long-term accuracy vs. immediate precision in Meteorology and Climatology. It also bears mentioning that if it rains in LA on Monday and we didn't expect it, it's not going to kill anyone. If the climate forecast gets blown, the situation is far more dire. So that adds to the desire for a HUGE amount of resources going into this kind of forecast model.

[DarrinS]

Meanwhile, no statistically significant warming in 18 years, 7 months

[ElNono]

I actually have a background in Meteorology, so I feel safe to address this.

The reason our short-term forecasts have a variance in precision is very straightforward. It's almost entirely because of cities. Even massive cities are relatively tiny areas of land. Getting a forecast to be exactly right within a tiny area is insanely difficult because it's not just about forecasting that conditions are right for rainfall, it's getting the rain border within a few square miles of land. That requires an immense amount of data. It is very difficult to acquire immense amounts of data for short-term forecasts, as the situation is constantly changing.

To put this into programming terms, you're comparing the following scenarios:

1) Predict how long it will take to debug a program that is acting completely screwy. Predict what bugs will happen and how long they will take to resolve.

That's pretty difficult to give any certainty to. Versus:

2) Give me an estimate for how long it would take to build a specialized system for, say, hospital equipment in SQL.

The second task is DECIDEDLY more complex and involves many more moving parts, but you could probably give a reasonable estimate of conclusion, versus trying to give a micro-managed level of precision to a specific set of variables that are often changing on the fly (99 little bugs in the code, 99 little bugs).

Notice I used the word precision. The problem is that you're looking for two different levels of measurement, thinking they are direct correlations. We can collect massive amounts of long-term weather shifts, far more than we can for, say, what it's going to be like in Austin on Sunday at 8pm. Yes, we have terabytes more data these days than we did 20 years ago, but it still pales in comparison to how much our climate models are absorbing and responsible for.

The climate model actually takes the same input multiplied by many times. You can have petabytes over petabytes of data, and we both know that it would take a year to run a millisecond of that model in raw form. What really happens with the historical data is you run a Markov model on it of you're looking for a pattern, or any other kind of linear regression or interpolation if you're looking for a trend, then you statistically project it, because, obviously, you don't have data for the future.

The statistical models used nowadays to do such projections are tremendous, and in part because you do have much more and better statistically significant data. That's why forecasts nowadays are major improvements over the last couple of decades.

Now I agree that if you had to split the forecast over the climate model, you would have to split over locality and number of inputs. When doing locality, you probably are weighing much more recent geographically-close data, than the historical or remote data. It's the sensible thing to do. With a climate model, you're instead handling a lot more data, but you're also looking for different things, like general trends, etc. It would be difficult to argue off-hand what's more complex. I think they both present their own challenges.

And with that said, I still wouldn't call our long-term forecasts precise. They are accurate to the extent that we have an idea of how bad things will be, but we can't for sure say that, given current indications, Miami will be flooded in March of the year 2112. We CAN say that it's likely that water levels will rise by "x" amount of feet due to long-term manmade global warming, and you can extrapolate that from what you will. But that's the difference between long-term accuracy vs. immediate precision in Meteorology and Climatology. It also bears mentioning that if it rains in LA on Monday and we didn't expect it, it's not going to kill anyone. If the climate forecast gets blown, the situation is far more dire. So that adds to the desire for a HUGE amount of resources going into this kind of forecast model.

See, I don't think we're anywhere near the precision (considering the sheer amount of data, the processing, the error rate, etc) to make that claim. If you read what the "consensus" above is, is that there's a "trend" towards warming that's "likely" man-made.

I actually buy that since the industrial revolution we've released a lot of different gases, etc that could've affected the planet temperature. Debatable if it's to dangerous levels. Clearly not dangerous enough yet, and it would be very debatable where that line is.

I think we should take that information as awareness to do better. But I'd like more research over time to have a better grasp of both that trend and the likelihood of coming from man before we actually break the glass and sound the alarm.

[TeyshaBlue]

Kudos to Havoc and EN for the lucid and enlightening dialogue. This is pretty much how a good bbs works.
Very enlightening...

[ElNono]

Expressing opinions, tbh... nothing wrong with that... I actually keep an open mind on the topic, and I think improvements both on statistical algorithms (like back-propagation neural networks, which are all the rage right now) and computing power can give us a much more solid footing both on what the problem might be and how to effectively deal with it.

[TeyshaBlue]

GFY

[ElNono]

GFY

YOU LIE,,,, slapped

[DarrinS]

GFY

??

[FuzzyLumpkins]

There's no such thing as 97% consensus amongst "world scientists" on the topic.

That said, I don't claim to know more than anybody. That's simply my opinion based on what I've read about the topic and my own experience. It could be wrong.

Ultimately, I've not encountered enough compelling reasons to change my opinion on the subject (yet)

Every major science foundation including our own.

As for your 'too many data points' argument, they are never going to be 100% accurate but using empirically tested mechanics they can get with a measurable degree of uncertainty. Given your background I figured you would be familiar with sampling and reconstruction. Even given the uncertainty, the blanket dismissal seems ludicrous to me. Queue the risk management crap that RG posts all the time.

[ElNono]

Every major science foundation including our own.

As for your 'too many data points' argument, they are never going to be 100% accurate but using empirically tested mechanics they can get with a measurable degree of uncertainty. Given your background I figured you would be familiar with sampling and reconstruction. Even given the uncertainty, the blanket dismissal seems ludicrous to me. Queue the risk management crap that RG posts all the time.

Somebody asked me about my opinion on the subject, and I thought I was fairly articulate explaining what my point of view is, and what's based on.

If you haven't read my last few posts in the thread, then that's where it is.

Basically, I don't dismiss the research, but I don't think it's conclusive enough to take rather drastic decisions just yet. More research is certainly warranted in that area, and I think we need to get a better grasp of what we're dealing with.

I don't think it's an irrational position. Then again, I have no expectations of anybody agreeing with me.

[FuzzyLumpkins]

Somebody asked me about my opinion on the subject, and I thought I was fairly articulate explaining what my point of view is, and what's based on.

If you haven't read my last few posts in the thread, then that's where it is.

Basically, I don't dismiss the research, but I don't think it's conclusive enough to take rather drastic decisions just yet. More research is certainly warranted in that area, and I think we need to get a better grasp of what we're dealing with.

I don't think it's an irrational position. Then again, I have no expectations of anybody agreeing with me.

Drastic or any decision at all?

[ElNono]

Drastic or any decision at all?

I don't think it warrants the fear, the regulations, etc at this point in time. To me, those are drastic decisions that have a very real impact on businesses and the economy in general.

If people want to be more proactive about it, only buy electric cars, etc. More power to them.

[DarrinS]

Hey Fuzzy, U.S. Elnono a "sophist piece of "? Or, is that reserved for me? I share his views on this issue.

[Th'Pusher]

Oh no. DarrinS doesn't know the definition of sophistry.

[Th'Pusher]

Hint: ElNono is being upfront and not intentionally trying to deceive using specious argument.

[TeyshaBlue]

Oh no. DarrinS doesn't know the definition of sophistry.

I love hot sophistry with cinnamon and honey.....wait a minute.

[Th'Pusher]

I love hot sophistry with cinnamon and honey.....wait a minute.

[DarrinS]

Hint: ElNono is being upfront and not intentionally trying to deceive using specious argument.

Hmm, but I agree with everything he is saying. I can pull up a five year old post that makes the same points.

[Th'Pusher]

Hmm, but I agree with everything he is saying. I can pull up a five year old post that makes the same points.

Sharing an opinion with someone doesn't mean you didn't use sophist aruguments to demonstrate that opinion.

[FuzzyLumpkins]

I don't think it warrants the fear, the regulations, etc at this point in time. To me, those are drastic decisions that have a very real impact on businesses and the economy in general.

If people want to be more proactive about it, only buy electric cars, etc. More power to them.

Fear? You just admitted that it was a rational possible outcome. The thermodynamic properties of ghg are well understood. The hope is that the Earth has ecological feedbacks to mitigate the effect and indeed that is showing to be the case.

Trying to quantify the probability and weigh risk seems very appropriate and the bet can be hedged actuarially. It's actually one cool thing about the various national and state insurance commissions and rate reporting is that they are quantifying the economic impact with precision.

[FuzzyLumpkins]

Hey Fuzzy, U.S. Elnono a "sophist piece of "? Or, is that reserved for me? I share his views on this issue.

He stated his case and came to his conclusion.

I call you a sophist because you yourself out to a conclusion and find any thing you can to feed it. You intentionally mislead and pay no attention to logical consistency. That you cannot tell the difference is hilarious.

[DarrinS]

K

He stated his case and came to his conclusion.

I call you a sophist because you yourself out to a conclusion and find any thing you can to feed it. You intentionally mislead and pay no attention to logical consistency. That you cannot tell the difference is hilarious.

Consistency

[ElNono]

Fear? You just admitted that it was a rational possible outcome. The thermodynamic properties of ghg are well understood. The hope is that the Earth has ecological feedbacks to mitigate the effect and indeed that is showing to be the case.

Trying to quantify the probability and weigh risk seems very appropriate and the bet can be hedged actuarially. It's actually one cool thing about the various national and state insurance commissions and rate reporting is that they are quantifying the economic impact with precision.

I mean fear in the sociopolitical presentation of "climate-change, the hot button issue", which is where this conversation started. The "if we don't do all this stuff right now, it's gonna blow!". I don't think the science findings warrant that.

What the science warrants is more research. We found a trend, an observation over a set of data. That's a great start. We don't outright know what combination of factors caused that trend to develop (otherwise, it wouldn't be a trend, it would be direct correlation). So let's do more research and try to pin point what is happening. It would be much informative both on the nature of the problem, the overall causes, and what the right solution to it is. To me, that's much more useful than going around making unquantifiable predictions (at this time) about impending doom.

Now, I'm not naive, and I know you can't make people care about something if there's no drama. But that doesn't mean I have to buy into it.

[FuzzyLumpkins]

K

Consistency

We've talked about this several times. I hope youre being intentionally obtuse for effect and not really this dimwitted. There are online dictionaries and you can easily look up the meaning and background of words after all.

You very obviously argue for a conclusion as opposed to allowing the evidence dictate the outcome. RG right from the start in that climate thread had you with like two dozen specious arguments in a month and that didn't include you repeating yourself like an idiot. It is what it is.