The SWIPE (Space Wireless Sensor Networks for Planetary Exploration) project uses Wireless Sensor Networks (WSN) to characterise the surface of the Moon. The envisaged scenario is that hundreds of small wireless sensor nodes dropped onto the Moon surface will collect scientific measurements. An ad-hoc WSN connecting these nodes will propagate the measurement data to sink nodes for uploading to a lunar orbiter and a subsequent transmission to Earth. The data gathered from the sensors will be processed using state-of-the-art data fusion techniques to overcome the restricted energy and bandwidth resources. In this paper, we first provide a short survey of classical data fusion techniques for WSNs. We then introduce data fusion architectures for the SWIPE project. Building on this, we propose data processing algorithms that enable energy conservation and processing efficiency in the proposed SWIPE architectures. The proposed algorithms are evaluated via a series of simulation models. The results show that the proposed algorithms can efficiently reduce the amount of the transmitted scientific data providing a good level of accuracy in the data reconstruction. Furthermore, they are able to correctly evaluate the node health status.