Meteorites are an extreme threat to the world that are not given the attention they deserve. While movies like to fantasize about the world ending from a massive meteor strike, the reality is this is not likely. However, what is more common and happens every so often is an object about 15-40 meters strikes the earth. On February 15, 2013, a small asteroid, measuring about 20 meters across, came down over the southern Urals of Russia, barreling in at about 19 km/s, and exploded over Chelyabinsk Oblast, near the town of Chelyabinsk. With a mass greater than that of the Eiffel Tower, the asteroid exploded in an airburst, unleashing energy equal to some 20 or 30 times the energy released in the Hiroshima atomic explosion. The enormous resulting shock wave shattered glass in the town’s buildings, injuring nearly 1,500 people.
Currently, NASA only seeks to find 90 percent of near-Earth objects that are more than 450 feet (140 m) in diameter, 7x the size of the Chelyabinsk Incident. Luckily, there is a plethora of historical meteorite data which varying sizes, locations, times, and classes. The goal of this project is to identify and use a relationship between the historical data points to help predict future dangerous meteorite events in the future.
If successful, this project could help save the lives of thousands by assisting in the prediction of meteorites that could threaten significant population centers. The project would be able to predict the coordinates, size, and timing of a meteorite event.
The negative outcome would be the occurrence of false alarms and also the missing of actual threats, which could scare the population unnecessarily and not correctly inform them of an event.
import pandas as pd
df_meteorites = pd.read_csv('Meteorite_Landings.csv')
df_meteorites.head(15)
| name | id | nametype | recclass | mass (g) | fall | year | reclat | reclong | GeoLocation | |
|---|---|---|---|---|---|---|---|---|---|---|
| 0 | Aachen | 1 | Valid | L5 | 21.0 | Fell | 1880.0 | 50.77500 | 6.08333 | (50.775, 6.08333) |
| 1 | Aarhus | 2 | Valid | H6 | 720.0 | Fell | 1951.0 | 56.18333 | 10.23333 | (56.18333, 10.23333) |
| 2 | Abee | 6 | Valid | EH4 | 107000.0 | Fell | 1952.0 | 54.21667 | -113.00000 | (54.21667, -113.0) |
| 3 | Acapulco | 10 | Valid | Acapulcoite | 1914.0 | Fell | 1976.0 | 16.88333 | -99.90000 | (16.88333, -99.9) |
| 4 | Achiras | 370 | Valid | L6 | 780.0 | Fell | 1902.0 | -33.16667 | -64.95000 | (-33.16667, -64.95) |
| 5 | Adhi Kot | 379 | Valid | EH4 | 4239.0 | Fell | 1919.0 | 32.10000 | 71.80000 | (32.1, 71.8) |
| 6 | Adzhi-Bogdo (stone) | 390 | Valid | LL3-6 | 910.0 | Fell | 1949.0 | 44.83333 | 95.16667 | (44.83333, 95.16667) |
| 7 | Agen | 392 | Valid | H5 | 30000.0 | Fell | 1814.0 | 44.21667 | 0.61667 | (44.21667, 0.61667) |
| 8 | Aguada | 398 | Valid | L6 | 1620.0 | Fell | 1930.0 | -31.60000 | -65.23333 | (-31.6, -65.23333) |
| 9 | Aguila Blanca | 417 | Valid | L | 1440.0 | Fell | 1920.0 | -30.86667 | -64.55000 | (-30.86667, -64.55) |
| 10 | Aioun el Atrouss | 423 | Valid | Diogenite-pm | 1000.0 | Fell | 1974.0 | 16.39806 | -9.57028 | (16.39806, -9.57028) |
| 11 | Aïr | 424 | Valid | L6 | 24000.0 | Fell | 1925.0 | 19.08333 | 8.38333 | (19.08333, 8.38333) |
| 12 | Aire-sur-la-Lys | 425 | Valid | Unknown | NaN | Fell | 1769.0 | 50.66667 | 2.33333 | (50.66667, 2.33333) |
| 13 | Akaba | 426 | Valid | L6 | 779.0 | Fell | 1949.0 | 29.51667 | 35.05000 | (29.51667, 35.05) |
| 14 | Akbarpur | 427 | Valid | H4 | 1800.0 | Fell | 1838.0 | 29.71667 | 77.95000 | (29.71667, 77.95) |
With over 45,000 meteorite incidents and a thousand years of data, there should be more than enough points to claim the results are significant. Additionally, there are multiply variables to analyze to gain further insight of meteorite patterns, so at the very least it should be possible to draw some significant conclusions.
Although there is a lot of data, the data is not detailed enough to provide detailed predictions of the exact day a meteorite would strike. However, it should be able to identify a time specific enough for space organizations to catch the meteorite and monitor it months in advance of a strike.
We will cluster the various sizes of the meteorites and find the variance in strike distances to predict the next strikes and level of danger for each group. Also, we can tie in the location data and find the geographical clusters that are most at risk to strikes.