Patterned ground, especially polygonal surface structures, are of particular importance for planetary sciences, as they are known from the Earth as well as from other celestial bodies such as Mars, Mercury, Venus and Pluto. They are therefore ideally suited as a basis for analogue studies. However, the classification of these structures is often based on individually and intuitively perceived parameters, which are difficult to determine, especially with remote sensing data. In our work, we therefore propose a new classification of polygonal surface structures. Based on a variety of conventional and established geometric parameters, in combination with the innovative approach of fractal geometry, we suggest a classification based on objective mathematical parameters. Based on remote sensing data from more than 100 sites, we show that polygons of different depositional environments can be distinguished and assigned to specific environmental conditions based on purely geometric data. Polygons formed in periglacial depositional environments can be clearly distinguished from structures formed in arid to hyper-arid environments. Furthermore, the structures can be correlated with the known subsurface conditions. The polygon classes resulting from the geometric investigations show a strong correlation with the ground ice content of the depositional areas. Polygons can thus serve as proxies, for example, to identify suitable landing sites for future Mars missions.
