Self-organizing protocols and algorithms in wireless sensor networks rely on knowledge of the topology of the network they manage. In most practical cases, establishing the topology prior to actual deployment is not possible, as the exact placement of nodes and the existence of a reliable link between any two individual nodes cannot be guaranteed. Therefore, this task has to be carried out post deployment. If the network is stand-alone and certain aspects are fixed (such as the identity and placement of the base station), then the task is achievable. If, however, the network has to interact with other systems - such as Unmanned Aerial Vehicles (UAVs) or mobile robots - whose operation is affected by environmental factors, self-organization becomes challenging. In this paper, we propose a dynamic topology construction algorithm, assuming that the network is a part of a joint deployment and does not have a fixed base station. A node carried by a UAV initiates the topology construction process and the ground nodes direct the network traffic towards the initiator. Our approach is collision-tolerant and enables in-network processing, thus reducing the computation burden of the initiator. We demonstrate the scope and usefulness of our approach using both lab and field deployments.