Jump to content

Scientific Method in climate science


NortonH

Recommended Posts

I am helping my son with a high school project so I want to make sure i get this 10000% right!

He has been asked to apply the formal scientific method analysis to the question of 'climate change'.

I went to wiki and summarised the method to ..

1. Observe world and propose something

2. Show the Null Hypothesis does not stand

3. Propose new hypothesis

4 Produce quantitative model and falsification criteria

5 New theory holds a long as model is never falsified

I must say, although it all looks so simple I am rather stuck. There seems to be a lot of vagueness in what is discussed and the model bit really baffles me.

Any help appreciated.

PS> The more sources I look at the more wide and vague the whole question becomes, HELP!

Link to comment
Share on other sites

Start with the null hypothesis in this case.

hint the clues are in the words 'change' and 'null'

So suggest a sensible null hypothesis about climate change, qualitative at first.

Decide something that would be true if the null hypotheis was true.

Then put numbers to it so you can measure it.

 

Edited by studiot
Link to comment
Share on other sites

20 minutes ago, NortonH said:

Thanks for your rather snarky response. Actually I was wondering whether someone else might be able to help.

!

Moderator Note

Giving a hint is not being "snarky". It's what we do with homework help here, we help rather than give answers.

You're either being far too sensitive, or you aren't being honest about your goal in discussing climate science. Dial back the troll, bro.

 
Link to comment
Share on other sites

hint the clues are in the words 'change' and 'null'

That sounded like a snarky comment. In any case it was not particularly informative was it?.

I was rather hoping that someone could provide some info about what model I am supposed to use. 

String Junky, Yes I want to put numbers to something - but what?

The example given as a guide was Einsteins theory of relativity. It is described by a single equation, provides it falsification criteria, has been tested thousands of ways and has never been falsified. I realise that i cannot even get of the starting blocks with this one. Which model do we consider?

 

Link to comment
Share on other sites

Well I'm sorry you took umbrage at my quickly dashed off scribbling yesterday.

It was designed in a hurry to 'get you started'.

One very important point you need to realise is that many words with wider meanings in the general English dictionary are used scientifically with only with specialist meanings.

Furthermore scientists will assume that you are using these specialist meanings, if you introduce them.

So if you need clarification about a technical term ask, don't waste precious time arguing about a definition.

From your posts I don't know if you know what a null hypothesis is, but my hint contains everything (scientifically) needed for a starting point.

How about

There is evidence for global warming.

or

There is not evidence for global warming.

 

Which of these do you think is null?

Which of these can you best falsfy?

 

A further hint, don't make your scheme too grand. Go for something achievable by a 14/15 or 17/18 year old scholar (you didn't say which and it makes a difference).

Another important question is. What subject is this for? Geography, Geology, Environmental Science, Chemistry...   ?

 

Where will you look for evidence (material to put into numbers)

 

Well, as I said it depends upon the subject.

You might look for  changes to migratory patterns of birds, insects and animals

You might look for changes to how far North od South various vegetation limits eg Tundra vary

You might look for direct measurements of some physical property - Can you suggest one?

 

Incidentally marks will also be awarded for the thinking behind rejecting certain markers, in the writeup of such a project.

So, for instance, should you accept or reject the obvious one Is the concentration of Carbon Dioxide changing?

Will knowing this actually assist?

 

All this is totally consistent with the roadmap I offered initially.

 

So shall we start again?

 

 

 

 

Edited by studiot
Link to comment
Share on other sites

On 2/23/2018 at 10:22 AM, NortonH said:

I am helping my son with a high school project so I want to make sure i get this 10000% right!

He has been asked to apply the formal scientific method analysis to the question of 'climate change'.

I went to wiki and summarised the method to ..

1. Observe world and propose something

2. Show the Null Hypothesis does not stand

3. Propose new hypothesis

4 Produce quantitative model and falsification criteria

5 New theory holds a long as model is never falsified

I must say, although it all looks so simple I am rather stuck. There seems to be a lot of vagueness in what is discussed and the model bit really baffles me.

Any help appreciated.

PS> The more sources I look at the more wide and vague the whole question becomes, HELP!

I would make it compulsory for all schools to view the documentary called "Chasing Ice" Besides the spectacular awe inspiring photography, I believe it answers your question...In fact I would make it compulsory for all interested in the foreseeable fate of this planet to view.

https://chasingice.com/

Let me add that at one time I was also doubtful as to human activity and the effects of climate change, but I also was always of the opinion that considering the stakes involved, if we were/are to err, it should be/must be on the side of caution.

Link to comment
Share on other sites

That is a case of Pascals Wager, I believe. 

I suppose the basic problem I have is that it is impossible to find a model for the climate so there is nothing to falsify. If I suggest this then the teacher (who seems to be a denier) will claim that it is not scientific. No Model means No Science.

 

Link to comment
Share on other sites

21 minutes ago, NortonH said:

That is a case of Pascals Wager, I believe. 

I suppose the basic problem I have is that it is impossible to find a model for the climate so there is nothing to falsify. If I suggest this then the teacher (who seems to be a denier) will claim that it is not scientific. No Model means No Science.

 

 

Now I will be sarky and offer a counter saying.

You can lead a horse to water, but you can't make him drink.

 

I'm sorry you don't want to be helped.

Link to comment
Share on other sites

The point is that if you cannot produce a model you do not have a theory. 

Regarding Pascals Wager, it does not add any information to your system, it just makes you reassess the risk/reward matrix. If you are unable to accurately predict this matrix (because you have no model) then you are still at square one.

My sons teacher made this a term project and has already sent kids feedback along the lines of 'no model, no science'. I am beginning to wonder whether he kind of knew in advance that there was no answer.

 

Link to comment
Share on other sites

38 minutes ago, NortonH said:

The point is that if you cannot produce a model you do not have a theory. 

Regarding Pascals Wager, it does not add any information to your system, it just makes you reassess the risk/reward matrix. If you are unable to accurately predict this matrix (because you have no model) then you are still at square one.

My sons teacher made this a term project and has already sent kids feedback along the lines of 'no model, no science'. I am beginning to wonder whether he kind of knew in advance that there was no answer.

 

The point is that if you had to look up the Scientific method you may well not have properly understood it in the first place.

Your replies lead me to believe that.

Also I warned you about scientifically improper use of scientific terms.

You also need to understand the words model and theory properly as you have used neither correctly.

 

The good news is that finding out these things is part of the learning process and taken care of in the road map I offered.

All that is required is for you to start listening instead of pontificating.

 

How about trying to answer one of the questions I asked?

Link to comment
Share on other sites

10 hours ago, NortonH said:

I suppose the basic problem I have is that it is impossible to find a model for the climate so there is nothing to falsify. 

Well, that is obviously untrue. There are models of the climate.

The problem for a high school project is that the climate is a very complex thing and therefore the models are very complex. If you think the teacher is a "denier" then this may be a trick question: "well if you kids can't produce a model of this incredibly complex thing in term project, then obviously thousands of scientists working for decades with access to masses of data and the world's fastest supercomputers must be lying".

A simpler thing to look at is the basic physics of how CO2 absorbs infra-red radiation. This has been known for about 200 years and is relatively straightforward.

Other things you could look at are the basic correlations between rising CO2 levels and average temperature (and other measures of climate change). Or the similar correlation of economic activity (which correlates with CO2 production) and temperature. 

And then, because correlation doesn't mean causation, you would also want to look at the [lack of] correlation between solar activity and temperature change (and whatever other excuses the deniers use). 

On 24/02/2018 at 6:51 AM, NortonH said:

The example given as a guide was Einsteins theory of relativity. It is described by a single equation

Which is why that is a terrible example. 

Link to comment
Share on other sites

Pontificating?
I find it odd that you are so dismissive of the scientific method as i had assumed it was the basis for everything.
What have I not understood?
Do we need a model or not?
Not sure which question you are referring to.
What needs to be falsified? Whatever theory is being proposed to replace the null hypothesis.
So far I am unable to get anyone to tell me what that theory is in any non-vague way.
Can you?
Can you provide a model?
As I understand the SciMethod that is required. No?

Subject - General Science.


Strange - The teacher expects something to be said about models and some indication that one exists and has not yet been falsified. Few people even got this far on the first feedback session. We know we need a model but cannot find one anywhere.
Einsteins model is easy to find, in contrast. Why is that a terrible example???? I thought it was fantastic. What is wrong with that theory?

A simpler thing to look at is the basic physics of how CO2 absorbs infra-red radiation.

Yes but that is a single component of the model which needs to describe  the climate of the planet. 

When I see the number of mechanisms going on in the Earths climate i doubt a model can exist really. I suspect that is what the teacher is driving at. Like a model for the stock market.

Link to comment
Share on other sites

1 hour ago, NortonH said:

Subject - General Science.

Thank you for this vital piece of information, though how you can honestly say "not sure which question you are referring to", when I said specifically

 

On 24/02/2018 at 10:19 AM, studiot said:

Another important question is. What subject is this for? Geography, Geology, Environmental Science, Chemistry...   ?

 

1 hour ago, NortonH said:

I find it odd that you are so dismissive of the scientific method as i had assumed it was the basis for everything.

Nothing I have said is dismissive of the scientific method. (SM)

However I did say you need to understand it.

The SM is not a one shot activity it is an ongoing process.
Often it is cyclical in that you go back to the beginning and repeat the steps in the light of the end conclusion many times in order to refine the output.

As such you can enter the SM at any stage and proceed forwards - It is even self correcting in the long run since the 'final' stage of verification returns you to the beginning if verification is not forthcoming, or subsequent new observations cast doubt.

There are many examples of this in the history of Science, and we can explore them if you wish, but I will leave it there for the moment.

 

The Model

This is not an appropriate start point for you.

This is because the purpose of the model is to answer some question or other.

But no model can answer every question, that is why you will need to choose an appropriate model at some stage after you know what question you want answered.

And you cannot have an appropriate model until you have decided what question you want to ask.

This is the second important piece of information I asked for.

On 22/02/2018 at 11:22 PM, NortonH said:

He has been asked to apply the formal scientific method analysis to the question of 'climate change'.

You did not tell us whether there was a specific question given or if you can choose your own.

In fact there should have been a project brief provided and it would be wise to post it. It may even contain a description of what is meant by the SM in terms of the task.

Otherwise the course notes should provide such a defintion.

Another answer needed is the answer to the question Is this a practical exercise or theoretical.

Will the students collect data on their own or will they just look stuff up?

 

One thing I can determine is that you have introduced the idea of the 'null hypothesis"

This is a statistical technique which implies some statistical input is required.

You need a particular type of model to use statistics; not all models work that way and in particular your Einstein example doesn't. That is what Strange meant by saying it refers to a single equation.

Strange also picked up my point about keeping it simple and within your means, so provide the information and we can proceed to talk about appropriate models.

Always remembering that even (perhaps particularly) in the planning stage you can go round the cycle again and again if your initial idea shows itself to be too wooly, until you get the final one.

I certainly did not expect my proposed null hypothesis to be your final one, I'm sure it can be greatly improved upon, but it is a start.

Which reminds me, you did not reply to my question what do you think a null hypothesis means?

Why is it null and what does that mean?

 

Link to comment
Share on other sites

3 hours ago, NortonH said:

Strange - The teacher expects something to be said about models and some indication that one exists and has not yet been falsified. Few people even got this far on the first feedback session. We know we need a model but cannot find one anywhere.

I’m not really sure what you expect. There are obviously models because they are used by climate scientists. But these are probably not something you will be able to get hold of. They are complex simulations of the behaviour of the atmosphere, land and oceans consisting of millions of equations to be run on supercomputers. 

What you might be able to find are some of the (relatively) simple equations describing specific bits of behaviour. I already mentioned the basic physics of CO2. You may also find some information on how temperature affects water content of the atmosphere and cloud formation. But then it starts getting congrats located. Clouds reflect sunlight (a cooling effect) but also trap heat (a warming effect). 

But none of these are much use by themselves because they all interact. 

Because of this the simplistic “falsification “ idea doesn’t really apply. The models are tested to see how accurately they recreate past climate and then used to predict what might happen in future. 

So climate modelling is not itself a theory to be tested but is using underlying scientific theories (which are well tested) to try and model the behaviour of the world. 

4 hours ago, NortonH said:

Einsteins model is easy to find, in contrast. Why is that a terrible example???? I thought it was fantastic. 

The problem is that climate models cannot be reduced to a single equation. It is simulation of complex interactions of a very complex system. 

4 hours ago, NortonH said:

When I see the number of mechanisms going on in the Earths climate i doubt a model can exist really.

I guess you don’t think weather forecasts exist either? (I’m beginning to wonder about the existence of this “teacher”)

Link to comment
Share on other sites

Studiot you asked more than one question - that is why i was unsure which one you were referring to. I have tried to answer them all.
In any case the subject is not relevant - my title was the Scientific Method applied to climate science. That is what it is about.

I do understand the SM and I suggest that it be used as the sole guide to resolving matters of science. I do not believe that any parts are optional.

Regarding the model - I see plenty of predictions being made about future climate and so my instinct is to ask "what model was used to make that prediction?". So far I have not seen any models capable of making credible predictions. As I understand the SM this tells me that such predictions are therefore baseless.

Your second question was also answered in my original post:
He has been asked to apply the formal scientific method analysis to the question of 'climate change'.

That is clear and unambiguous. The whole point of the exercise is to get students to start thinking about the terms being used. We hear about climate change, people tell us what we have to do about it but it is only when we started this project that people realised it was all a bit vague. This was the subject of one of the first feedback sessions.

To answer your next questions - students will apply the formal scientific method analysis to the question of 'climate change'.

Why do you say that the concept of the Null Hypothesis is a 'statsitical technique'? It is not just statistical, it is the basic SM. Why are you now focusing specifically on statistics?
??

The Null Hypothesis in the case of climate is basically that everything is behaving normally and there is nothing happening that is out of the ordinary.

I note that this is round three and still you have not produced any model. Are you trying to tell me that a model is not necessary?


 Strange

"There are obviously models because they are used by climate scientists. "

Well that is the question. Should these models not be open to scrutiny?
If they cannot be scrutinized then how do we test them? Do we even need to?

"But none of these are much use by themselves because they all interact. "
Yes. That seems to be where it gets complicated.

I am not clear how the concept of falsification does not apply - I thought it was fundamental to the SM.

I am glad that you brought up the concept of weather forecasts - they exist but are notoriously unreliable.
So let me rephrase my point - I am sure that some models exist (as discussed above) but are they reliable? How can we know if we are not able to test them?

So this seems to come back to what I wrote earlier - No model means no scientific theory. To me the model is crucial and I do not see how it can be sidestepped.


 

Link to comment
Share on other sites

6 minutes ago, NortonH said:

I am not clear how the concept of falsification does not apply - I thought it was fundamental to the SM.

It does apply, just not in the simplistic way you imagine it does.

At one level, models are tested and if they give the wrong results they are changed or discarded.

At another level, individual bits of science that are used to build the models are tested and either included  to build better models or rejected. For example, the basic physics of CO2. Or the effects of dust in the air (acting as nucleation centres for clouds). Or ...

10 minutes ago, NortonH said:

I am glad that you brought up the concept of weather forecasts - they exist but are notoriously unreliable.

They are very reliable. There is a popular (mis)conception that they are often wrong, mainly based on confirmation bias I suspect. Short term forecasts (24 hours) have accuracies of 80 to 90%, I believe. This falls off for longer term forecasts, which is why you don't see forecasts for more than 1 weak or so.

12 minutes ago, NortonH said:

So let me rephrase my point - I am sure that some models exist (as discussed above) but are they reliable? How can we know if we are not able to test them?

They are tested. Against past events and against the current climate.

14 minutes ago, NortonH said:

So this seems to come back to what I wrote earlier - No model means no scientific theory. To me the model is crucial and I do not see how it can be sidestepped.

So that's fine then. There are models and it is all based on sound science. Your attempts to sow doubt about this aren't going to work. 

1 minute ago, studiot said:

I'm sorry you wish to cut off your son's nose to spite your own face but I can't see any way to help him further.

It's fine. My hypothesis is that there is no son and no teacher.

Link to comment
Share on other sites

23 hours ago, NortonH said:

That is a case of Pascals Wager, I believe. 

I suppose the basic problem I have is that it is impossible to find a model for the climate so there is nothing to falsify. If I suggest this then the teacher (who seems to be a denier) will claim that it is not scientific. No Model means No Science.

 

Wasn't familiar with the term "Pascal's Wager" but after checking this is the definition I Got...."Pascal's Wager is an argument in philosophy presented by the seventeenth-century Frenchphilosopher, mathematicianand physicist Blaise Pascal(1623–62).[1] It posits that humans bet with their lives that God either exists or does not."https://en.wikipedia.org/wiki/Pascal's_Wager

Well considering how this is in regards to science and observational climate change, while Pascal's little wager is with regards to whether one needs to accept some unscientific myth on whether there is an after life or not, with either reward or punishment, I don't see it as a case at all. Again, I urge you to watch the doco "Chasing Ice"

Edited by beecee
Link to comment
Share on other sites

Studiot I gave you polite answers to all your questions so I do not undertand your petulence. 
It seems that any question requiring a scientific answer gets your back up.

Strange.
I get that the individual components of a model are established (eg CO2 IR absorption) but the point is that a model which combines all necessary components is a very complex thing that also needs to be confirmed in its own right. I can give you the model for two ping pong balls colliding or hitting a solid surface but I cannot tell you which ball will be chosen from one of those lottery machines.

I have do disagree about weather models. I have compared forecasts with results a week later and the accuracy is not good. 

Regarding the testing of the models - to me this is the critical part. 
Can you tell me of any models I can look at? Do you have any links? As far as I am concerned this would clinch it as far as the assignment is concerned. Demonstration of a credible model makes or breaks the whole theory.

Am I wrong to ask this question? You accuse me of sowing doubt. Well I HAVE DOUBT. I have not seen a model, let alone seen one tested so am I wrong to say this? Am i wrong to express scepticism of something I have not seen? Is this not precisely what the SM says I should do???

The point of this exercise is to apply the SM to climate change. It is not about proving or disproving any theory.


 

Link to comment
Share on other sites

Hi Moderator. What is your point exactly?

I get the impression that I have rubbed some people up the wrong way for asking what I consider to be legitimate questions.

Is there a problem with that? Are my questions not legitimate?

Science is supposed to be scrutinised and tested. 

Link to comment
Share on other sites

1 hour ago, NortonH said:

Hi Moderator. What is your point exactly?

I get the impression that I have rubbed some people up the wrong way for asking what I consider to be legitimate questions.

Is there a problem with that? Are my questions not legitimate?

Science is supposed to be scrutinised and tested. 

The problem is you trying to play an insincere game of cheap tricks on people much smarter and much more knowledgeable than you. Is that clear enough for you?

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.