Nowadays everybody is interested in doing software simulations. It is very good for mechanical or civil engineering in terms of designing & simulating various dynamic or static components. But what about when models are used to represent plant/crop growth?
Some people even do not bother to go into the field & even cannot think of staying in the sun or rain. But they are very comfortable in doing simulation in lab. Is it good? Does not it hamper the self-knowledge? Without even knowing the characteristics of plant growth, soil conditions, assumption of a lot of parameters in the model may lead to a very satisfactory or I might say desired results. But actual field conditions may be quite different. Similarly not only for agriculture, even civil people or other engineers may face severe problems while doing field projects rather than doing best-fitting model simulations. Does it make any sense really?
Technology arrives to change ways things are done. Often it erodes human employment. For instance the use of automatic teller machines in the banking sector is a threat to the employment of human tellers. The same thing applies to the use of software. Softwares are often used without understanding the theories underlying the processes and procedures being involved. The machines are more efficient at operating than human beings. Yet they are human creations. Too much of human patronage of these machines may be counterproductive.
Hi Adrij,
Developing mathematical models to represent the physical world is not a new activity, think of the work of Newton, for instance. It has however gone a tremendous growth with the availability of computers.
Computer simulation of some physical system allows analysts to better understand the characteristics of the system they are simulating; and even better, to study possible responses of the system to some external pressure. Think of a building subjected to an earthquake shake, it is impossible to study this event if the building has not been constructed yet! Even if the building is in operation, earthquakes are not frequent phenomena, sometimes you would need to wait for many years to study the building response to a particular shake. That is why it is important to develop computer simulations: you can simulate thousands of earthquakes and study the building response to them.
The art of computer simulation is, however, the validation of the model. This is done by comparing with the real world. If you want to study the behaviour of a crop to drier climate conditions you need to do computer simulations. In this case you capture in your model all you know about the behaviour of the system to the climate conditions, that is where field work is important! Then you compare your model results against your observations of the physical system under climatic stress. Are the results comparable? does the model show the same pattern of behaviour? if the answer is no, then you need to improve your model by better observing the behaviour of the physical system (more field work). Sometimes a few iterations are necessary until you are satisfied with the model results.
As you can see, computer simulations complement and extend the field work.
At the end of the day, you should remember that the simulations provide you with some ideas about the problem, they do not solve the problem completely for you, but when you have to make a decision it is better to do it based on some scientific evidence rather than just trying to guess!
Hope it helps,
Augusto
Yes Augusto, I fully agree with you. As mentioned by you, that field work results should be incorporated into the model parameters, is the good idea behind simulation. For a general idea behind simulation is that it should give a wide range of possibilities that may affect our physical system. But I think the precursor for such simulations should be a good amount of field work. Agreed?
Hi Adrij,
Yes, I totally agree. A good mathematical model can only come from a detailed observation of the real world, what you call 'field work'. Once the main characteristics of the physical system have been captured by the model, it is possible to study the response of the system to a wide range of inputs, something which is impossible to do just by observing the physical system.
I, however, insist that models give you only a guide for action, you must combine other sources of information (your and other people's experience, indigenous people's ancestral knowledge, etc) about the system before taking a decision.
Hope it helps,
Augusto
P.S. Have a look at the article below which deals with the problem of model validation:
https://journals.ametsoc.org/doi/pdf/10.1175/1520-0477%281982%29063%3C1309%3ASCOTEO%3E2.0.CO%3B2
- Badges
- Science topic
- Similar topics
- Geoscience
- Atmospheric Sciences
- Atmosphere
- Air