Problems with Analyzing Data with Cyclical Error

There may be some doubt concerning the rate of global warming which is due to the nature of the temperature measurements. For statistical data one usually encounters a normal distribution but with temperatures there are both daily and annual cycles in the data. This can bias the estimate of global warming as seen by what happens to the data in the following example. The solid line is the true curve, the dots are the data and the dashed line is the estimated trend based on the data. Notice that the slope is greater for the estimated trend. This increases the time necessary to make accurate estimates of the rate and rate of change of global warming. When analyzing the data one has to be on guard against subtle errors.