Predicting future climates on planet Earth is an extremely hard task due to the myriad of factors involved.

To make the necessary calculations requires computers with capacities far beyond the average home computer.

However, climate models are become ever more reliable thanks not only to greater computing power, but also to more extensive observation efforts of the current climate, and an improved understanding of the climate system. It is generally accepted that Earth's atmosphere will get warmer and sea levels will rise in the future.

According to Michael D. Lemonick, senior writer at Climate Central, the problem is that warming causes a huge variety of changes that are difficult to predict. For example, warming can affect the polar ice caps, ocean currents, cloud cover, permafrost, ecosystems, and more. Furthermore, each type of change can either exacerbate the warming or create a counter (cooling) effect. An accurate model has to account for each and every factor.

Some factors have only been discovered recently, like for example, the global carbon cycle, which is the interchange of carbon between the atmosphere, biosphere, and oceans. This is extremely important because carbon dioxide is what is causing climate change. Our understanding of the global carbon cycle is just coming around, and is now being inserted into climate models.

What climate modelers are doing is nothing short of building the entire world and all its complexities into a computer, a frighteningly daunting task. But to see what the world would look like in 2100, scientists often use what is known as hindcasts. They look at models from the 1800s and run them through to the present day. If they were reasonably accurate, that can be taken as a good sign.

Climate science is still in its infancy, and will become more sophisticated as time goes by. Methods which are experimental now may become standard in the future. Observations from weather stations around the world will continue to produce data, creating an ever-expanding pool of data. Human comprehension of essential global systems will improve. For example, we will better understand the effects of melting ice on temperature, ocean salinity, sea level, etc., La Nina and El Nino effects, and the long term effects of emissions from the burning of fossil fuel.

Having an accurate model will be useful for a number of reasons. Agriculture experts can determine which crops will grow best given a predicted temperature and precipitation forecast. Coastal city planners can better defend against a rising tide. Everybody can be more confident in planning for their own futures.