Road accidents are a common phenomenon in which a vehicle collides
with another vehicle, object, or person, causing damage, injury, or death.
These accidents can occur due to various reasons such as driver error, road design, vehicle malfunction,
weather conditions, and alcohol or drug impairment.
We decide to study this phenomenon since, despite governments around the world have implemented various
measures to reduce the number of road accidents;
road accidents remain a significant problem that can
negatively affect our life.
With our analysis we aim to understand if this phenomenon is growing, in which areas is more present and what the causes are.
From the graph above we can see that in the first years the car accidents number in USA is remained the same.
From the 2017 to 2018 there is the largest increase in cases of about 100'000;
Then there is a huge drop in 2020 probably because of the lockdown caused by Covid.
In 2021 the numbers started increasing again and this year we expect to reach and even surpass the peak reached in 2019.
The number of crashes in the plot are only from the data of 64 counties present in our dataset.
The United States have 3143 counties with over 6 milion car accidents every year.
From studying the number of accidents per year we study the behavior of the number of accidents during the different years.
Although we expected some kind of pattern during the busiest months, the graph instead describes a flat situation, constant for most of the year.
The only change recognizable and common to most years is a decrease in the number of accidents during the period between March and May.
This map shows the number of total accidents that happened in each county.
The three counties that have the most accidents are:
- Queens with 2.270.976 population
- Suffolk with 1.481.364 population
- Nassau with 1.355.683 population.
We can notice that the area where the data is mainly collected is in the state of New York and the ones near it.
For that reason our next visualization are going to be focused on those areas.
Going now to study more precisely the counties close to new york we can have a more concrete map.
Through the second graph we can see more clearly which are the areas with the greatest number of accidents.
From the barplot we can see a big difference between the first four counties and the others:
- Nassau with 277'000 cases
- Suffolk with 276'000 cases
- Queens with 231'000 cases
- Kings with about 215'000 cases
To give more accurate analysis of the accidents we decide to check the accidents that happened per county population.
We can notice that the counties with most total accidents per population are:
- Orange with an average of about 2.67 accidents per person
- Sullivan with an average of about 2.60 accidents per person
- Monroe with an average of about 0.77 accidents per person
To make this visualization we used a Sankey diagram to display the flow between the road descriptor, the lighting conditions and the road surface conditions.
As we can notice by the graph, it is possible to find when most accidents happened.
Thanks to that we can give some advice to when to pay more attention to the driving:
- When the road is wet and grade.
- When there is snow or ice on the road.
- When is dark and there are not lights.
- When there is a curve.