Falls versus finds: what the difference means
A fall is a meteorite that someone saw or heard arrive and then collected. A find was picked up without a witnessed arrival, sometimes thousands of years after it landed. NASA’s Meteorite Landings dataset (records to 2013) holds 1,107 valid falls and 44,534 valid finds, so falls make up 2.4 % of the meteorites in the table.
Why falls matter
A fall is collected soon after it lands, before rain, frost and soil have altered it. Its date and place are known. A fall sample therefore gives a fair picture of what arrives from space, at least for rocks big enough to survive and be noticed. Finds tell a different story, shaped by where people search and by which rocks survive long on the ground.
Finds come in waves
Most finds in the record are dated 1970 or later: 42,812 of the 44,534. The finds with a complete record, 30,841 of them, are led by Antarctica with 22,094, then Oman (2,993), United States (1,503), Libya (1,465), Australia (618), Algeria (614). Apart from the United States, none of these places is among the ten countries with the most observed falls. Falls are recorded where people saw them; finds are recorded where searchers looked.
| Where | Complete find records |
|---|---|
| Antarctica | 22,094 |
| Oman | 2,993 |
| United States | 1,503 |
| Libya | 1,465 |
| Australia | 618 |
| Algeria | 614 |
| Chile | 402 |
| Kenya | 223 |
Different mixes of rock
The two samples contain different proportions of each family. The table compares falls with complete find records. A family that is much more common among finds than among falls is one that searchers spot easily or that survives well on the ground.
| Family | Falls | Share of falls | Finds | Share of finds |
|---|---|---|---|---|
| L chondrites | 373 | 35.1 % | 10,504 | 34.1 % |
| H chondrites | 345 | 32.5 % | 13,237 | 42.9 % |
| LL chondrites | 84 | 7.9 % | 3,437 | 11.1 % |
| Intermediate and ungrouped ordinary chondrites | 32 | 3.0 % | 126 | 0.4 % |
| Carbonaceous chondrites | 44 | 4.1 % | 737 | 2.4 % |
| Enstatite chondrites | 17 | 1.6 % | 142 | 0.5 % |
| Rumuruti and Kakangari chondrites | 2 | 0.2 % | 20 | 0.1 % |
| Eucrites, howardites and diogenites | 60 | 5.6 % | 526 | 1.7 % |
| Other achondrites | 19 | 1.8 % | 285 | 0.9 % |
| Martian meteorites | 5 | 0.5 % | 39 | 0.1 % |
| Stony-iron meteorites | 11 | 1.0 % | 173 | 0.6 % |
| Iron meteorites | 48 | 4.5 % | 880 | 2.9 % |
| Unclassified stones | 23 | 2.2 % | 6 | 0.0 % |
| Class outside the site’s decoder | 0 | 0.0 % | 729 | 2.4 % |
Irons show the contrast best in absolute numbers: 48 observed falls against 880 finds. Iron masses are dense, look unlike ordinary stones and survive long on the ground, which helps searchers find them. The HED achondrites make up 5.6 % of falls but a smaller share of finds, and ordinary chondrites dominate both samples.
Size works the other way
The median complete find record weighs 27 g, while the median observed fall recovered 2.9 kg. Searches on ice and desert surfaces record very small stones. At the top of the scale, finds win: the 60 heaviest finds on the giant finds page include 8 that outweigh Sikhote-Alin, the heaviest fall.
Falls and finds, decade by decade
The two records grow on different clocks. Falls arrive at a steady pace of a few dozen per decade. Finds stay in the low hundreds until the 1960s and then jump: the 1970s add 4,908 find records, about 15 times as many as the 1960s, and the 2000s add 17,664.
| Decade | Observed falls | Finds |
|---|---|---|
| 1900s | 55 | 83 |
| 1910s | 65 | 88 |
| 1920s | 68 | 89 |
| 1930s | 88 | 218 |
| 1940s | 55 | 138 |
| 1950s | 58 | 152 |
| 1960s | 57 | 335 |
| 1970s | 57 | 4,908 |
| 1980s | 54 | 6,763 |
| 1990s | 54 | 11,531 |
| 2000s | 58 | 17,664 |
| 2010s | 12 | 1,946 |
The table explains why statements about “all meteorites” depend on which sample you use. In every decade since 1970, finds outnumber falls by more than 80 to one. A share calculated across the whole dataset is therefore mostly a statement about finds from recent decades, while the falls keep the older, witnessed part of the record.
How this site uses the two
Every entity page on this site is built around falls, because a fall has a date and a witnessed place. Finds appear only as the sixty heaviest, each set against the observed falls around it. The What fell near tool counts falls only. If you want to see the falls of one country or decade, start from the country list or the decade list; the method is on Reading the record.
The heaviest of each
Side by side, the eight heaviest finds and the eight heaviest falls look like two different worlds. Eight of the eight finds are irons, against two of the eight falls. Only the very top of the fall list, Sikhote-Alin, reaches the range of the large finds.
| Find | Recovered | Fall | Recovered |
|---|---|---|---|
| Hoba | 60 tonnes | Sikhote-Alin | 23 tonnes |
| Cape York | 58 tonnes | Jilin | 4 tonnes |
| Campo del Cielo | 50 tonnes | Allende | 2 tonnes |
| Canyon Diablo | 30 tonnes | Kunya-Urgench | 1.1 tonnes |
| Armanty | 28 tonnes | Norton County | 1.1 tonnes |
| Gibeon | 26 tonnes | Jianshi | 600 kg |
| Chupaderos | 24 tonnes | Knyahinya | 500 kg |
| Mundrabilla | 24 tonnes | Ochansk | 500 kg |
Large iron finds can lie on the surface for a very long time before anyone records them, and the record has no date for their arrival, only the year they were found or reported. For a fall, the year is the year it came down.
Countries with many falls
In the countries with the most falls, the balance between falls and finds varies widely. The table compares the two for the eight countries with the most falls; in six of them (India, France, China, Japan, Germany, Ukraine) the falls outnumber the complete find records.
| Country | Observed falls | Complete find records |
|---|---|---|
| United States | 146 | 1,503 |
| India | 124 | 8 |
| France | 58 | 11 |
| China | 54 | 33 |
| Russia | 48 | 82 |
| Japan | 42 | 12 |
| Germany | 31 | 16 |
| Ukraine | 29 | 11 |
Where falls outnumber finds, the country’s meteorite record rests mainly on witnessed events. Where finds dominate, as in the United States, searches of dry ground have added far more than eyewitnesses.
Counting caveats
The counts here use the class field as written. About 729 complete find records carry rarer labels that the site does not group into a family; they stay in the table as their own row. Relict meteorites, 75 records of heavily weathered material, are left out of every count.