Electrical faults due to leaning trees are common in Nordic countries. This fault type has been studied in and it was found that the initial transients in the electrical network due to the associated arc reignitions are behavioral traits. In this paper, these features are extracted using the discrete wavelet transform (DWT) to localize this fault event. Wireless sensors are considered for processing the DWTs on a residual voltage of different measuring nodes that are distributed in the network. Therefore, the fault detection is confirmed by numerous DWT processors over a wide area of the network. The detection security is also enhanced because the DWT responded to a periodicity of the initial transients. The term for locating the faulty section is based on the polarity of a specific frequency bandpower computed by multiplying the DWT detail coefficient of the residual current and voltage at each measuring node. The fault due to a leaning tree occurring at different locations in an unearthed 20-kV network is simulated by the alternate transients program/electromagnetic transients program and the arc model is implemented using the universal arc representation. Test cases provide evidence of the efficacy of the proposed technique.