Paper detail

Mapping cumulus clouds to scale invariant rough surfaces

Motivated by a recent observation on the self-organized criticality of cumulus clouds (Phys. Rev E 103, 052106, 2021) we study their connection to self-similar rough surfaces, in which $f\equiv \log I$ plays the role of the main field, where $I$ is the intensity of the received visible light. By simulating the light scattering based on a coarse-grained phenomenological model in a two-dimensional cloud, we argue the possible connection of $I$ to the actual cloud thickness. Although in the vertical incident light $f$ is proportional to the cloud thickness, in the general case it is complected. We study the statistical properties of observational data for $f$ with a focus on the conventional exponents of this scale-invariant rough surface. By calculating the roughness exponents, and comparing them with other exponents like the fractal dimension of loops, the distribution function of the radius of gyration and loop lengths, and the exponent of the green function, we prove that this surface is unconventional in the sense that it is the non-Gaussian self-affine random surface which violates the Kondev hyper-scaling relations.