218
H. von Storch, J. S ¨
undermann, and L. Magaard
Was he approximately your age?
Namias was 14 years older, he died in 1997, and Bjerknes was much older—he could
have been my father. This was a time of enormous cross fertilization.
What about the nineties?
The nineties are clearly climate, the chemical and biological cycles in the climate
system. These are the next big topics, not the physical cycles of climate.
Could you say something on the role of experiments? Like in GARP when people came
together to make a big effort to observe the atmosphere or the ocean or the boundary
layers, intensively for a limited time, and then go back into the laboratories?
Experiments are absolutely necessary. Experiments are the basis of physics. We do
process experiments, which are real physics in the ocean, where you try to learn . . .
Could you give an example?
Such as, how does the Ekman layer work? These are physics experiments. But then
you have to make other experiments and these are very often not recognized as
experiments: is Global Change an experiment? Now, you see, one important thing
about geophysics is, and I tried to explain that to my students, physics is based
on experiments where you can control one factor at a time. But in geophysics all
factors are changing simultaneously. Nature is making experiments for us and as
geophysicists we are very often simply put in the role of the observers. We can’t
control the experiment. If you make an experiment on a hurricane, you don’t control
the experiment.
How did you feel the assistance of numerical models, which was increasing with time?
Numerical experiments . . .
Numerical experiments, you mean models. Models are an essential part of physics
and sciences today. There is no question about that. You need models for everything.
You only have to use them in the right way. There are many different kinds of models.
Models are to simulate certain physical processes. But a model is an approximation
that can be used to study physical processes. Then there are models that predict
weather. They have limitations. What are the limitations? In prediction it is chaos
and turbulence. Then you make models of the tides. They are probably very good,
because they have solid physics behind it and the process is truly repeatable, because
it is forced. Then you can make models that are plainly speculative, that means, where
we are trying ideas. The question is what you make with models. There is nothing
wrong with models, but how you interpret the model, that’s the important item.
Could you just give an example of a speculative model?
I would say, modeling 100 years of climate change is speculative.
You are not talking about models like Stommel’s?
No, Stommel’s model is a conceptional model of a process in which he explores the
effect of β. He explains.
H. von Storch, J. S ¨
undermann, and L. Magaard
Was he approximately your age?
Namias was 14 years older, he died in 1997, and Bjerknes was much older—he could
have been my father. This was a time of enormous cross fertilization.
What about the nineties?
The nineties are clearly climate, the chemical and biological cycles in the climate
system. These are the next big topics, not the physical cycles of climate.
Could you say something on the role of experiments? Like in GARP when people came
together to make a big effort to observe the atmosphere or the ocean or the boundary
layers, intensively for a limited time, and then go back into the laboratories?
Experiments are absolutely necessary. Experiments are the basis of physics. We do
process experiments, which are real physics in the ocean, where you try to learn . . .
Could you give an example?
Such as, how does the Ekman layer work? These are physics experiments. But then
you have to make other experiments and these are very often not recognized as
experiments: is Global Change an experiment? Now, you see, one important thing
about geophysics is, and I tried to explain that to my students, physics is based
on experiments where you can control one factor at a time. But in geophysics all
factors are changing simultaneously. Nature is making experiments for us and as
geophysicists we are very often simply put in the role of the observers. We can’t
control the experiment. If you make an experiment on a hurricane, you don’t control
the experiment.
How did you feel the assistance of numerical models, which was increasing with time?
Numerical experiments . . .
Numerical experiments, you mean models. Models are an essential part of physics
and sciences today. There is no question about that. You need models for everything.
You only have to use them in the right way. There are many different kinds of models.
Models are to simulate certain physical processes. But a model is an approximation
that can be used to study physical processes. Then there are models that predict
weather. They have limitations. What are the limitations? In prediction it is chaos
and turbulence. Then you make models of the tides. They are probably very good,
because they have solid physics behind it and the process is truly repeatable, because
it is forced. Then you can make models that are plainly speculative, that means, where
we are trying ideas. The question is what you make with models. There is nothing
wrong with models, but how you interpret the model, that’s the important item.
Could you just give an example of a speculative model?
I would say, modeling 100 years of climate change is speculative.
You are not talking about models like Stommel’s?
No, Stommel’s model is a conceptional model of a process in which he explores the
effect of β. He explains.