NUFORC Geographic Data
Geographic data analysis using Pandas and Plotly.
In a previous article, I created a notebook to look at NUFORC data, but I only focused on the features of the dataset. I want to do this to help researchers locate regions of interest and focus on those regions. I relooked at the data again using geographic features such as latitude, longitude, city, etc., using the simple maps dataset. This dataset can be merged with the NUFORC data by the city to determine the relationship between geography and the sightings.
url = "https://nuforc.org/webreports/ndxloc.html"
data = load_data(url)
cities = pd.read_csv("uscities.csv")
cities.head(5)
df = data.merge(cities, on="city")
df.head(5)By plotting the first 200,000 sightings, we see a scatter plot that maps the density of UFO sightings. There is a large number of sightings in the Arizona and Mississippi regions, primarily close to Huntsville and Phoenix. Another cluster is close to New York City and the eastern seaboard. Other sightings are scattered throughout the eastern and western sides of the continent.
fig = px.scatter_mapbox(df[:100_000],
lat="lat",
lon="lng",
color="density",
text="summary",
hover_name="city",
hover_data=["city", "date / time", "population", "shape", "military"],
zoom=3,
height=400
)
fig.update_layout(mapbox_style="open-street-map")
fig.update_layout(margin={"r":0,"t":0,"l":0,"b":0})
fig.show()
By county, the counties with the highest number of sightings can be counted using a Choropleth map. We see southern New Mexico and Clark County Vegas, along with a strip of counties in the Northwest, as having a high number of sightings. On the east coast, areas like Miami, New York, and Maryland have high counts.
One claim often made is that the 37th parallel has a higher concentration of sightings and strange activity. I found that to not be true; actually, the 40th parallel should be the “strange” parallel. I wondered whether this was due to the high population density because of cities like New York and Chicago, which lie close to that latitude. Researchers should be looking at that latitude rather than the 37th.
By looking at sightings around the world we get a different perspective of where sightings are located. Britain as Canada has a large number of sightings, this is likely due to the anglophone tendency of NUFORC reports.
A heatmap of the continental United States provides the intensity of sighting reports or a representation of how frequent sightings are. A heatmap provides data on how frequent sightings are and what the best locations are to set up instrumentation for looking at the data.
df_count = df.groupby(["lat", "lng"]).count()
df_count = df_count.reset_index(level=["lat", "lng"])
df_count[["count"]] = df_count[["date / time"]]
df_count.head(5)
fig = px.density_mapbox(df_count, lat='lat', lon='lng', z='count', radius=12,
center=dict(lat=0, lon=180), zoom=0,
mapbox_style="stamen-terrain")
fig.show()
Notice how most of the sightings concentrations are in the northeast and midwest. This is likely due to the higher population density. This region should be a better choice for a region to focus on in UFO research because of its high number of sightings. If some of these objects are extraterrestrial or interdimensional, then it makes sense they would focus on highly populated areas if they wish to be seen. Regions of interest in the midwest, would include Chicago and St. Louis. On the east coast, it would be New York, Baltimore, and Washington. In the west LA, San Fransisco, Seattle, Phoenix, and Las Vegas would be areas of interest.
What’s most surprising is despite my last analysis and other findings, traditional locations for sightings like Las Vegas, New Mexico, etc. aren’t the hotspots necessarily, and a lot of the reason for their notoriety has more to do with historical associations (Area 51, Roswell). Many large historical sightings happened on the east coast as well, like in New York, Washington, and the Hudson Valley. These areas should be the focus of research and investigation. Both notebooks will be available on Colab and can be viewed there.
