In climate modelling, invariably, scientists make best use of computers (painstakingly integrated for such specific purpose) in a bid to fathom climate change as well as its multiple aspects in every little detail possible. It is not immature to state that computer models are put into use, while climate science is put under discussion.
As a matter of fact, climate models turn out of immense use, in a variety of ways, like, scientists are empowered to uncover the cycles of ice ages, existed hundreds of thousands of years ago, then, they can easily forecast the turn of climatic events likely to turn around in this century or which are due in next century.
But when we rush our thoughts about climate model, people are found to be in dark, at length. Neither do we know how a climate model appear like and are also unaware about its core functioning too.
At Tech And Environment, after having researched broadly and verified the narratives from dozens of climate scientists, we put forth the following account about climate models and their numerous aspects.
Ques. What on earth, is a Climate Model?
Ans. In a layman’s language, a climate model refers to a computer system with hundreds of lines of codes embedded thereby, for fetching accurate forecasting targeting a specific part of the globe or any such portion in a given climate system. Not surprisingly, oceans, ice, vegetation, forests, mountains, land and atmosphere are precisely simulated herein.
Now, if we consider a mega climate model that is designed to cover global climate, it is likely to contain tremendous amount of lines of codes, say something equalling 18000 pages while there will be involved a few hundred scientists in its core development and finally a supercomputer will also be in the need for effective functioning which will be so large and widely stretched, just like a tennis court to run and function seamlessly.
Clearly, such models give an output that is critical for enriching climate science understanding and also to highlight the fault-lines on part of humans, as to what actions of ours damage the climate of earth. Also, in the past 50 years, the agreement reached on climate policy has had such output as is obtained from climate models, on their robust foundations.
Further, it won’t be wrong to maintain that climate models are just the superior machines developed for weather forecasting, in fact, a step ahead of weather forecasting by which climate changes taken place for decades and the dreaded impact of such changes can easily be summarized.
As a matter of factly, Hadley Centre, which is UK’s Met office, plays on such Unified Model to accomplish both the tasks.
Now, for the purpose of simulation of weather and climate, a huge quantity of computing power is needed, which implies that models of today are operationalized on the robust foundation of giant super-computers.
The apt example to cite is Cray XC40 Supercomputers installed at Met office, Hadley centre that is potential enough to conduct 14000 trillion calculations against a second just passed by.
Underlying Physical Aspects:
Now, commoners like us, are by and large, puzzled about the guiding principles that are at play inside the climate model and for such unique purpose, Tech And Environment has rolled into existence. The climate model tends to capture earth’s climate which is guided by a handful of processes and elements with a core inter-relation among them, and such are represented by equations. For instance, oceans, land, atmosphere and the icy regions present on our planet. Further, nothing goes beyond the physical, chemical and biological frameworks which our scientists cherish and the models are designed on such laws and equations which are dominant in earth’s system.
Clearly, scientists have always stressed that the basic physical principles should constitute the climate models, like the First Law of Thermodynamics, (also regarded as the law of conservation of energy reflecting that in any given framework, energy does not extinct rather it is transformed into another state).
Then comes the Stefan-Boltzmann Law, whereby scientists have proved that rising temperature (say 33oC hotter climate) is solely due to the natural greenhouse effect.
Following this and to establish a link between air’s temperature and its corresponding pressure on water vapour, Clausius-Clapeyron equation is taken into use.
And so forth, the list is a bit lengthy.
Nonetheless, there exists a sheer complexity in this set of partial differential equations, because of which, there isn’t found any exact solution to be understood, while there do exist some exceptions in certain cases. In fact, such is the mega challenge involving lengthy mathematics which has invited a cash prize of $1 million for any outlandish genius who would prove and place a viable solution there.
But, there is evidence that the model contains numerical solution which is simply in approximation.
As such, line by line computer codes take the form, when scientists try to decode such physical principles and in case of a climate model meant for the whole planet, such code-lines go so much low that a million lines emerge there.
Programming Languages Utilized:
As for global climate model, Fortran is used to write code in, which IBM developed in 1950. To illustrate this, we arranged a snapshot of a few code from one of the models of our Met Office Hadley Centre.
Following this, climate scientists have nurtured ways to put other programming languages as well, such as C, Python, R, Matlab and IDL. On a supercomputer, C and Fortran are widely put in use, to push the global climate model process info further.
Apparently by now, climate model has code wherein there are equations that represent the physics of our climate system, such as formation of ice in sea and how it melts in Arctic waters and then how does it change into gases and the moist pressure between the land and air.
Following images show how different climate processes have been included into our global models over the earlier decades.
Now, a pertinent question arises, as to how such a large number of calculations happen in a model?
The answer lies in the fact that calculations for every cubic meter of our climate system isn’t possible, since the computing power has some limitations too while the climate system is pretty big and complicated too. For ease of handling, the earth is intently divided into small boxes, regarded as “grid cells” while in any model meant for global climate, there can be scores of layers across the length and breadth of the atmosphere.
Below, we present a 3D image of what a typical climate model looks like and calculations is done in each cell wherein major aspects like temperature, air, pressure, humidity and wind speed are perfectly modelled too.
In a model, there appear grid cells whose size is referred to as spatial resolution and the grid cells may be located over a 100 km on longitude and latitude, in a global climate model that would be rough and tough in design. Now, since our earth is a sphere, the grid cells are designed to be bigger where equator appears along the latitude and longitude, while at the poles, such would be smaller in size.
To overcome such an issue, climate scientists don’t hesitate to cross over to other techniques of gridding, such as cubed-sphere and icosahedral. Not surprisingly, any model that will be high-resolution will contain small boxes in a big number with the advantage that climate insight will be gathered in great detail for a specific world region, when the resolution will be higher, but such systems are time consuming since many calculations have to be done, therein.
In short, if we, in a model, enhance the spatial resolution by a factor of two, there will be needed huge amount of computing potential, say 10 times more, to execute processes in similar time cycle.