Not only technological limitations hinder the use of close meshed sensor networks based on low-cost and miniaturized particle sensors. It is the physics and the nature of aerosols themselves that severely limits the development of integrated sensors. In this work, we examine the limitations related to statistics, which play a role for small sample volumes. Mass concentration metrics as PM10 and PM2.5 are especially prone to high uncertainty due to the low number of particles in small sample volumes. Those metrics give weight to large particles that are not frequently observed. The overall uncertainty consist of a Poissonian counting uncertainty, an uncertainty connected to the size distribution of particles as well as a minor contribution form the mass density distribution of airborne particles. We examine this aspect analytically.