The problem of 'extrapolating outside your historical environmental conditions' must include the possibility that we can make no reliable projections of climate change impact. That takes us to the difficult area of how we can make decisions about things we can't be sure of. I suspect that for the harder questions (looking a long time into the future, or at effects on very long-lived species, or at substantial departures from historical conditions) It may be necessary to adopt a risk management approach - since we probably can't satisfactorily remove the uncertainty. That generally means hedging - not seeking a single optimum solution, but applying a range of solutions in the hope that some will perform well, or at least adequately.

.

A model to be credible must go through two phases, calibration and verification, ie, taking a historical series should divide it into two sections, a first that will be possible to determine the constants that exist in this model, and the second with different period of time, one should check the fit of the model.

.

Any model that does not pass through these two phases is a mere speculation.

.

Predicting what will happen in the future normally takes into account what has happened in the past and present. This is where the historical data become important! However, since natural phenomena are never linear in response to changing circumstances any prediction may be susceptible to large errors but to do something is better than doing nothing at all! I order to predict the future with minimum errors the two previous contributors have given some good suggestions on modeling and testing of models against controls. This approach means that you should be able to study expected outcomes of climate change under controlled conditions which will give you some data for simulating the possible outcomes of climate change in the future. How many times and locations that such studies can be carried out is a problem to be solved by the objectives of the study and the number of environmental factors (subject to change) taken into account, otherwise errors will continue to occur in such predictions.

.

Maybe it was not made clear my opinion in my previous comment, but motivated by the brilliant responses from others will make clearer my opinion.

.

For me with or without climate change, with heating or cooling, the ability of man to design scenarios for periods over 30 years is null and can be a real joke when they are read again in the future.

.

I will start my explanation of my opinion in an unconventional way by analyzing the literature of science fiction of the past 60 to 50 years.

Whereas famous authors like Isaac Asimov, Arthur Clarke and even film director Stanley Kubrick, we see that the ability of these icons of science fiction is truly zero in terms of the scientific future prediction.

.

Isaac Asimov puts in one of his book an astronaut to calculate the solution of a problem with a slide rule (who is younger suggest you look on the Internet which is a Slide Rule). Likewise Stanley Kubrick (1968) in the movie "2001: A Space Odyssey" and Arthur Clarke (1969) in the book of the same name (just freeze the political profile of the cold war translate him for 2001, posing as a sort of villain film a mainframe HAL 9000 (HAL => IBM) that interferes on a trip to Jupiter that is made by 2001!

.

That is, the authors of science fiction that has the poetic license to propose anything, does not have the ability to a range of 30 years to provide for things like computers (Isaac Asimov) loss of supremacy of mainframe computers in relation to micro (Arthur Clark and Stanley Kubrick) or the existence of interplanetary Spaceflight to Venus in 2001!

.

Now what is the explanation of all this, the man just generally locked into a horizon of knowledge applied, even in basic science is already possible to predict major steps, and try to design the resolution of problems of the future with the existing tools in this and worse with the political and cultural reality of the present or even the past (kings and queens interplanetary). Who in 1980 could predict that 2013 a response to an intriguing question as made by a Chinese were answered by teachers and / or researchers from Europe, South America, Africa and the United States (put in order of appearance!).

.

Extrapolate science is science fiction literature, partially known what happened in the past, we know the situation at present, but it completely ignores the interaction of variables that will in the future.

.

Another problem that no scientist or science fiction writer can predict is how will the political, economic, social and philosophical structuring of humanity 30 years from now, much less what are the capabilities of future cooperation to solve large problems.

.

What will happen to environment or on human health in the future? I do not know and only clairvoyants can know, but the only certainty we have is that we will be better, both on the environment or human health, because humanity like challenges and the challenge is to restrict its action to nature and further improve their quality of life to humanity respond.

.

Many thanks for all valuable answers!

The errors are had until in the weather forecasting in a short time (i.e. one week). The point is the climate changes are projections: different future conditions that will probably happen according to the measures that take to mitigate their impacts according to the scenario. In the forecast has a grade smaller of the error because measured data every 15 seconds, every minute, per hour and per day; that which can reduce the error of the forecasting. But like other people comment to your question, the error will persist because the weather is not lineal and which transforms it into variable. The most useful thing is to generate projections of climate change and with base on the results; generate a specter of alternative or options to mitigate or to adapt to the impacts that it generates in a small, medium and large period of time.

Look on these papers.

• https://www.researchgate.net/publication/266742210_THE_ANALYSIS_OF_THE_RELATIONSHIP_BETWEEN_CLIMATIC_WATER_DEFICIT_AND_CORN_AGRICULTURAL_PRODUCTIVITY_IN_THE_DOBROGEA_PLATEAU
• https://www.researchgate.net/publication/263775361_Detecting_climate_change_effects_on_forest_ecosystems_in_Southwestern_Romania_using_Landsat_TM_NDVI_data
• https://www.researchgate.net/publication/259605805_Changes_in_the_forest_ecosystems_in_areas_impacted_by_aridization_in_south-western_Romania

Article The analysis of the relationship between climatic water defi...

Article Detecting climate change effects on forest ecosystems in Sou...

Article Changes in the forest ecosystems in areas impacted by aridiz...