Climate is fundamentally chaotic with small, hard to detect changes in the atmosphere which can then develop into significant systems. While scientists have a passion for understanding how climate works, they also take pride in acknowledging what is not yet known and determining the uncertainty associated with data. Climate models, for example, simulate the effects of climate changes but can only provide one range of possible outcomes. Future predictions on longer timescales have a larger uncertainty as there is a greater period for numerous influencing factors to provide an effect. The scientific community applies a multi-disciplinary approach in order to understand areas of uncertainty, an example being EQUIP’s project looking at the prediction of climate impacts.


Topics requiring further research

Our current knowledge of past patterns in climate, modern-day observable processes and general predictions of near-future changes rests on large volumes of research and multiple lines of evidence. It is in the interest of the scientific community and society that the areas requiring more study are recognised and that uncertainty surrounding data can then be reduced.

The principal science journal, Nature, identifies 4 key areas of research which require the attention of the climate science community.


Regional climate prediction is important for people to know how their local conditions will change so they can plan for the future. (General Circulation) Models used to represent a range of physical processes on the planet are at too large a scale at present to provide confident predictions for regions. This is because there is greater complexity at smaller scales, for example where mountains strongly influence the local climate. Therefore while these models can be used with an understanding of their limitations, the uncertainties of model variables need to be reduced for better regional forecasts.


Precipitation – there is confidence that an increase in global temperatures will increase evaporation and cause an acceleration of the hydrological cycle. While the models agree this will result in drier subtropical areas and more precipitation at higher latitudes, the amount of change and the effects on winter precipitation are uncertain. What is required is more data about past precipitation, and a better understanding of moisture dynamics in the atmosphere and the formation of clouds in different regions.


Atmospheric aerosols have been a topic of intense research, with the aim to determine how these tiny airborne particles (such as sea salt, dust and sulphates) affect both temperature and rainfall. Sensors on satellites and on the ground attempt to measure how aerosols scatter and absorb solar radiation. More of this data, together with experiments on atmospheric processes, are required to identify how different particles alter the climate. This is important because some aerosols, like black carbon, produce a warming effect, while particles like sulphates reflect sunlight and therefore reduce warming.


Palaeoclimate data from the numerous natural archives of past climate fluctuations have some statistical uncertainty associated with each source and the patterns they show can vary between sites because of local effects. Analysis of tree ring records is of particular interest. It has been found that temperatures derived from tree rings in certain locations deviate from actual temperatures during recent decades. Identifying a divergence in the records has been beneficial to the science as research can now focus on understanding tree growth behaviour and response to climate changes.


Data uncertainty & Trends

Data is evaluated in terms of accuracy and precision, which can be measures of reliability. Accuracy describes the ability of data, measurements or results to match the actual ‘true’ value. Precision is how close these data, measurements or results are to each other, working as a measure of the spread of data from the average.



Climate data is often displayed through time as a graph, like temperature for the last 1000 years. Data rarely forms a simple linear pattern, with values increasing, decreasing or stationary through time. Instead it will likely appear as a zig-zag representing temporary extremes and possibly an increase of data points. Remembering that climate is defined as a long-term average, these zig-zag variations from the average trend serve as an indication of the trend’s precision.


Person walking a dog. The dog's path may vary as it wanders from a straight line, but both the dog and the person are headed the same way. You can predict where they will end up by looking at the trend of the person walking the dog.