Curiosity Spots a Field of Martian Polygons
Rover Curiosity beamed back a 360 degree panorama from Valle Grande, a valley on Mount Sharp. A polygon field this large has never been seen on Mars, and JPL is weighing three explanations.

Rover Curiosity has sent back images of Martian ground that, across a wide stretch, looks like honeycomb. NASA’s Jet Propulsion Laboratory (JPL) released a 360 degree panorama the rover captured on June 19 and 20, 2026, during sols 4,930 and 4,931 of the mission, in a valley the team has nicknamed Valle Grande, on the slopes of Mount Sharp. Polygons have turned up on Mars before, but never in a field this size, and the mission still does not know what made them.
These are polygonal cracks, a network of intersecting fractures that divide the ground into many sided shapes, so from above the surface looks like a honeycomb. Each shape measures between 4 and 8 centimeters across. In the panorama they stretch in every direction, as far as the cameras can see, climbing the flanks of a roughly 6 meter rise the team nicknamed Miraflores, capped with a thick layer of sand.
Earlier in the mission, the rover had come across shapes like these in small clusters at a few points during the climb. This is the first time they have appeared as terrain in their own right, rather than as a local detail.
“We’ve seen a lot of fascinating terrain through Curiosity’s eyes, but this sea of polygons took our breath away. We carefully measured their shapes and chemical makeup, and we’re hoping there are clues in the data about how these structures formed.”
That is what Ashwin Vasavada, the project scientist for the Curiosity mission, said in a JPL statement. The shapes and their chemical composition have been measured, but the data are still being analyzed, so the explanations the team lists for now remain candidates.
What could even crack ground into shapes like these
The team is weighing three processes, and none has been singled out yet as the answer:
- desiccation cracks, mud that dried and split, a mechanism that was clear for some previously photographed polygons
- repeated cycles of heating and cooling, which fracture ground into geometric shapes
- compression, which squeezes water out of sediment once the material is buried beneath the surface
It is also possible that two of these processes were at work here, or all three. The difference is not academic: drying mud calls for water at the surface, while compression calls for water already trapped in a buried layer, two different stories about what that valley once was.
Fourteen years on Mars
Curiosity landed on Mars on August 5, 2012, nearly 14 years ago now, and has been climbing Mount Sharp, a peak roughly 5 kilometers high, since 2014. Along the way it has photographed sulfur crystals and meteorites, but the mission’s most important find was not geometric but chemical: data showed that ancient Mars had water, chemistry and nutrients that could have supported microbial life. The rover has also found carbon based molecules believed to be precursors to RNA and DNA.
Valle Grande is a continuation of that same climb. Billions of years ago the lower slopes of Mount Sharp were dotted with lakes and streams, so any shape that carries a trace of water feeds into that same reconstruction.
The mission is run by JPL, which Caltech in Pasadena manages on behalf of NASA.
What this image does not prove
The panorama shows shapes, not their origin. A field of polygons is not confirmation that liquid water was recently present at this spot, since every mechanism under consideration points back to the valley’s distant past, and one of them does not even require water at the surface.
The find is also not a biosignature and says nothing about life. The carbon based molecules the rover found earlier could be either biological or geological in origin, and this image has not moved that distinction one step.
Finally, this is a mission release, not a peer reviewed paper. The measured shapes and chemical composition are still being analyzed, so the answer to how this field formed will come from that data, not from the picture.



