Polski
Many people watch the weather where they live. Some in a more advanced way – sometimes placing weather ministrations with a thermometer, barometer and windmill on the wall of the house. Others, rather in passing, will check the cloud cover over the city or peep at their neighbors to assess how to dress for work, especially when duties have to be performed outdoors. Based on such observations, they acquire certain statistical beliefs over time, which can be reduced to two types of generalizations: “It rains a lot around here, but my locality always misses the rains” or “It rarely rains around here, but, when it does, it always rains on my locality.”
Such observations, however, are essentially incidental, and thus lose their substantive importance in the discussion, gaining an anecdotal role. The adjective is intended to discombobulate the argument. While in colloquial language, anecdote means an amusing story, in meta-scientific language (i.e., concerning the science of science) it means a single observation with no reliable statistical foundation.
Indeed, one resident may argue that, in his mind, the cloud cover over the city is particularly high, while another may argue the opposite. In order to avoid such anecdotal arguments, researchers try to collect a large data set and only draw conclusions based on it. Then it may turn out that what one casual observer thinks is the norm is more common, and what the other thought was rare is indeed the exception.
Cloudiness over the city – examples from the United States
An example of big data analysis is this spring’s publication from the Proceedings of the National Academy of Science (PNAS). Its authors considered data from 447 U.S. cities with populations over 50,000. Cities were defined as urban units, which sometimes meant combining neighboring units into a single larger one, such as the “City of the City”. New York and Newark. Satellite data on cloud cover over cities and their surroundings from the first two decades of the current century were collected for the analysis.
Statistics on the average humidity of the region were also used. The peculiarities of mountain and coastal cities have been taken into account, with both ocean and Great Lakes coasts. After they were singled out, the largest group were the cities laboriously named inland cities. The smallest was Manhattan, Kansas, while the largest was Atlanta. Water areas were excluded when analyzing the environment.
Of course, as with the results of anecdotal observations, so in the case of collecting data from several hundred cities, it was possible to find both examples of dense housing being conducive to cloudiness and the opposite. Several locations noted the weekly cyclicity exhibited by the changing cloud cover over the city. Their residents can confidently argue that differences in economic activity during weekdays and weekends translate into weather. However, this is not the case for the other several hundred sites surveyed.
Research confirms – there are more clouds over cities
Reviewing data from multiple cities allows one to infer what is the norm and what is the exception. In general, it can be considered that large cities are cloudier than their neighborhoods. In winter it is more pronounced during the day, from spring to autumn – at night. Statistically, this effect is strongest during the day in July and at night in June. It turns out that cloud cover over the city is a fairly cyclical phenomenon after all.
Considering data from multiple cities, located in different regions, throughout the year and at different times of the day, is key to drawing this conclusion. For example, if one were to perform the analysis only for July and daily values, a similar conclusion could be drawn for cities in the eastern part of the United States, but the results for the western part would not be so clear. On the other hand, in July, but at night, one would be able to notice the differences between cities located in the north and south. In spring and autumn for that any effects would be weaker.
The observed effect is stronger during the day in inland cities in humid and temperate regions. In locations with colder and colder climates, it is weaker or even the opposite. However, this only applies to daytime cloudiness, not nighttime. In coastal cities, on the other hand, the influence of the breeze reduces the daytime effect but increases the nighttime effect. Phenomena can overlap or cancel each other out. For example, metropolises on the Pacific coast generally lie at the foot of mountains. This leads to the formation of low clouds, similar to fog. Eventually, Los Angeles became an eponym for one type of smog.
Cloudiness over the city – European research
The authors acknowledge that their analysis confirms conclusions published several years earlier by other researchers. They were drawn from a smaller data set, for London and Paris only. The mechanism is similar, related to the urban heat island. Interestingly, the phenomenon was initially predicted to rather reduce cloud cover over the city.
The daytime warming air in densely built-up areas rises more intensely than the cooler air from rural areas, taking water vapor with it. This increases moisture loss and could result in less cloud cover over the city. The resulting thermal chimney, however, draws in moist air from the surrounding area, and urbanization dust becomes a collection of condensation nuclei. This creates clouds over cities that can last quite a long time. In the case of London and Paris, this effect occurs with some delay and clouds over mainly in the afternoons.
In conclusion, statistically, those who claim that it clouds over a city (especially a large one) more often than over its surroundings are more correct.
