The basic idea of proposed cloud statistical mechanics approach to study cloud behavior is that clouds are viewed as composed of numerous basic cloud elements and their behavior is manifested as the collective effect of ensemble interactions among these basic elements. This is analogous to the idea of statistical mechanics, thus the name. With this perspective, cloud statistical behavior may be understood once we determine the micro-scale interactions between cloud elements. The understanding can be achieved through mathematical tools such as probability theory or network theory. In this paper I propsed two basic interactions: clumping and merging among basic cloud elements for boundary layer cumulus clouds. Using a complext network model cloud organization is sucessfully simulated. The appeal of this approach is that it offers a framework where studies at the micro-scale and macro-scale can be united. It offers a viable route to efficiently represent clouds in large-scale models as well as a way to understand cloud behaviors in a new way.