Como construir e testar modelos para a Economia Ambiental?

As with any impact modelling, the purpose is about taking decisions with some considerations on probable outcomes. You build an economic model not to be accurate but to understand the dynamics of different parameters and promulgate the least worse version of an action.

In designing a model, it is important to understand the goals and what can be changed without increasing complexity mathematically too much. Since we can make computer simulations of the parameters by setting some constants and changing only one to see its effect, it is important to draw these parameters mathematically and scientifically(by domain knowledge) without increasing the computational and space complexity too much as said priorly. Depending on the goals certain variance level can be selected by the expert and experimentations can start, EDA and other statistical measures also will be followed during this procedure

Here's where it gets interesting – you have different 'tools' to choose from, like math, statistics, simulations, or agent-based models. Each tool has its pros and cons, kinda like picking the right tool for a job. And don't forget, you need data to fuel your model, like how a car needs fuel. So, you'll need to see what data's available.

There are some important checks: calibration (getting the right taste), verification (making sure all data are in), validation (seeing if it's okay in practice), sensitivity analysis (figuring out if small changes matter), and uncertainty analysis. These steps help you fine-tune your model, sort out its strengths and weaknesses, and understand how much you can trust it.

Using a model is like using a tool to answer questions and make smart decisions about the environment. It helps us explore different scenarios, evaluate policies, and find the best solutions. Sharing our findings is like collaborating at a session where we get valuable insights and make informed choices together.

Every model should take into cognizance the fact that complex systems such as the environment have multiple causative factors that cannot simply be captured in a model no matter how encompassing that model is? For example pollution, deforestation and other environmental hazards are some of the causes of climate change however not many models take into focus the fact that earth is suffering from 'old age'. Take mammals for example, the human body degenerates with age irrespective of type of food, water, and environment where that body reside. Aging is also a factor that captures the essence of all these other factors that advantage or disadvantage the human body. So models of the environment should also make room for aging.

Remember any applied mathematician will tell you that every model is a faulty model and should not be taken 100% correct. Also as George Box said: “ All models are wrong, but some are useful”