 Data collection is fundamental to every branch of science. In order to construct a reliable record of climate spanning a century or more, we need to understand how the observations are influenced by non-climate factors, such as changes in the instruments and their environment. For example, on hot summer days, urban areas are noticeably warmer than rural areas, or even city parks. This is called the urban heat island effect and it's a real effect. Urbanization creates darker surfaces which absorb sunlight rather than reflecting it back to space. Urban areas also have less moisture to cool the air. A number of other factors including waste heat from human activity also play a role. We've built a lot of cities over the last century. When rural weather stations are surrounded by urban sprawl, that can lead to an increase in the observed temperatures at that station. But does this significantly affect our measure of climate change? As an initial check, let's look at a map of global temperature trend of the 20th century from the Berkeley Earth Project. Red areas are warming and blue areas cooling. We can see that some areas have warmed more than others. We can also get a good estimate of human development from satellite images of the Earth at night, showing light pollution from our cities. Let's compare that to the warming map. The areas showing the fastest warming are often areas with very little development, for example the Arctic, the Amazon Basin and Mongolia. Some areas with a lot of development show little warming, like China and the southeastern United States. There is also warming in the oceans. The oceans dominate the surface of the planet, but we don't build cities in the oceans. NASA has done a detailed analysis in which they adjust urban stations to match the nearest rural stations. Here is what the global temperature record looks like using different methods to identify urban stations. The results are almost identical. Another study by the UK Met Office compared temperature trends for still and windy days. The urban heat island effect has most impact on still nights. They found no significant effect on global mean temperatures. These studies and other evidence all tell us that urban heat islands have a minimal effect on the global warming trend. Even so, there's a myth that urbanization and not the greenhouse effect is responsible for a significant part of the 20th century warming. So what is the fallacy in this case? It's the fallacy of jumping to conclusions. Just because localized urban warming could affect the temperature record does not mean that it does. To determine whether 20th century warming is due to urban heat islands, we actually have to look at the data. And the data tell us that the effect is small compared to the total warming over the last century. Another point of confusion arises from the fact that scientists make corrections to the weather station data. Why do we need to correct weather station records? The result of any observation is determined by a combination of what is being observed and the instrument that is being used. If there is a change in the instrument, this will affect subsequent observations. Scientists apply adjustments to observations to address changes in station location, instrumentation and operating practices. They do this by reconciling temperatures before and after the change with observations from other nearby stations. For example, if a weather station is moved up a hill, it may record cooler temperatures than before the move. We need to adjust for that so that the data before the move is comparable to the data after the move. Otherwise, we create a false cooling effect that does not correspond to reality. Even though the adjustments are justified, we can check whether they are significantly affecting the global temperature record. Here is a comparison of the temperature record with and without adjustments. We can see that the adjustments do make a difference, but the difference is small. So adjustments are necessary to maintain a true record of temperature changes. However, one myth claims that these corrections are responsible for a significant amount of the warming over the 20th century. This is another fallacy of jumping to conclusions. Scientists do adjust the observations, but only when there is good scientific reason to do so. And the adjustments only make a small difference to the global mean temperature. We've looked at a couple of myths involving possible problems in the temperature record. In both cases, the effect is much smaller than the warming signal.