Are weather forecasts lying to you?
For years, individuals have relied on numerous different mediums in order to obtain the weather forecast. From newspapers to smartphones, there are a plethora of different ways that individuals can check the weather. Have you ever wondered which medium is the most/least accurate? Or which of these mediums for weather forecasts are most popular? Well, the two data sets that I analyzed may help to answer those questions. The csv file that I will be analyzing came from a survey that was given to individuals in order to learn more about how people check weather. The dataset asks questions about age, gender, annual income, as well as geographical region the participant lives in. The API file that I am analyzing looks at the weather forecast vs the actual weather that occurred in a plethora of different U.S. cities.
While studying the data set, one may ask “why is this data important?”. First, we learn about how many people actually check for the weather before they leave their house in the morning. From this, we can see how gender, income, age, etc. affect this action. Lastly, combining the data from the dark sky API as well as the “weather-check” csv file, I will be able to determine if the daily weather forecasts that individuals are receiving are accurate or not.
Who was surveyed?
In this data set, “weather-check.csv”, I examined the gender, US region, age, and annual income of all of those that participated. I took multiple values from the CSV file and created a plethora of different graphics using python pandas that accurately showed the representation of each corresponding group in the dataset. From this data, I was able to determine that the survey mostly contained women, people 45–59 year olds, people from the Pacific region of the US, and individuals who made between 25,000–50,000 dollars last year. Lets take a look at who was surveyed:
Who actually checks the weather?
The most important piece of data from “weather-check.csv” came from the “Do you typically check a daily weather report?” column. From this column, we learn about all the people in the United States that actually check a weather report of some kind before they leave their house in the morning. It is so important to know this data because once we know WHO actually checks a weather report, then we can examine HOW various types of individuals check said report. After hours of digging and sorting through the data, I found important data regarding groups of Americans that actually keep up to date on weather reports. In the charts below, I determined who checks a daily weather report based on age, sex, US region, and annual income.
How do you check for the weather?
I believed that it is most important for me to analyze HOW people check the weather report. I iterated through all of “weather-check.csv” data and examined all the different ways that different types of people hear about the weather forecast. After sorting through the data from the CSV file, I was able to create the visualizations below. These charts demonstrate how/how likely individuals are to check a weather report. From this data, I learned that people are most likely to check weather on a smart device (i.e. smartphone, tablet, etc.), many people still learn about weather from the local news, and how the US region you live in determines if you check a weather report or not.
Is your forecast accurate? Or is it bullsh*t?
In order to add onto my original analysis, I decided to pull an API file from the web that contained data regarding the overall accuracy of weather forecasts in specific U.S. cities. I decided to look at two cities in particular: Los Angeles and New York and determine if the weather forecasts in these cities are generally accurate or if they’re a load of crap.
The code on the left comes from my python notebook regarding the accuracy of weather reports in LA. As you can tell, weather reports in Los Angeles are generally accurate. This would go well with my data about how many individuals check the weather (bar graphic in section 2). A large amount of the individuals surveyed in “weather-check.csv” that are from the Pacific region of the U.S. constantly check the weather report because the reports that these individuals receive are generally extremely accurate.
On the other had, individuals from New York experience a completely inverse situations. In the API that I analyzed, New Yorkers reported that the weather reports that they receive are extremely inaccurate. As you can tell from the data on the left and below, it is a common reoccurrence for the citizens of New York to receive incorrect weather reports. This goes in line with what we already know from the graphic in section 2. New Yorkers (North East/Mid Atlantic) will check their weather reports less often, considering that the reports are generally less accurate than of those in the Pacific part of the U.S.
Conclusions
First, I learned how gender plays a role in checking a weather report. Females seemed to be more likely to check the weather than males. I also discovered that people that live in the Pacific region of the United States are most likely to check a weather report. Older people (those above 40) are much more likely to check a weather forecast than those younger than them. I also learned that people are more likely to check the weather report based on where they live. For example, it is probably easier to formulate an accurate weather report for a summer day in Los Angeles than it is for an autumn day in New York City. These are certainly factors that influence users to check the daily weather prediction.
In this data set, we also learned a lot about how people check the daily weather report. I was determined that most people check the weather forecast via their smartphone and the Local TV News. But after looking at this study, the numbers are more evenly distributed than I had originally thought. Weather data is an essential part in our everyday lives. So, make sure that the weather forecasts that you are receiving are reliable and up-to-date!
