The Lu-Hamilton model [1] is a sandpile model which was proposed to study the dynamics of solar flares. In this work, the centroids of the regions affected by the simulated flares are considered as nodes of a growing complex network, using the Suzuki-Abe method [2], which has been extensively used to study seismic sequences [3,4]. Results show that the Lu-Hamilton model yields Gaussian distributions for the degree distribution of the network, which is consistent with a random process, where the probability of all locations to be centers of energy release is the same. Lu \& Hamilton model managed to reproduce a large part of the main characteristics observed in solar eruptions but the physical interpretation of the elements constitutive of the system presented several difficulties [5]. One way to solve this issues was presented by Morales and Charbonneau [7]. We follow that idea and modify the Lu-Hamilton such that the main element of the cellular automaton is a strand of interconnected nodes  that can reconnect when a certain condition is fulfilled.. Thus, we modify the Lu-Hamilton model, such that, instead of running over a regular grid as in a traditional cellular automaton, sites in the grid are reconnected, so that energy is released not to spatial neighbours, but to topological neighbours, as given by the grid network configuration. Rewiring of the grid network is done is such a way that energy release out of the system is always possible, and the number of network edges is conserved. In this case, the Lu-Hamilton model shows a SOC state different to the regular grid case, presenting structures where the appearance of solar flares has a higher probability. Both cases, the regular grid and the rewired grid, present a power-law distribution of released energy per event, which is consistent with previous findings on solar flare dynamics [4, 5, 6], but for the rewired grid, the total energy in the system is lower, due to the larger number of energy release paths that the reconnection of sites introduces. These results suggest a new approach to the simulational study of solar flares and magnetic reconnection, where the dynamics yielding energy release events occurs not over fixed paths, but along a complex network with reconnected paths, as expected during magnetic field line reconnection events in the solar surface [7].

[1] Lu, E. T., Hamilton, R. J., McTiernan, J. M., \& Bromund, K. R. (1993). The Astrophysical Journal, 412, 841.

[2] Abe, S., \& Suzuki, N. (2004). Physica A: Statistical Mechanics and its Applications, 337(1-2), 357-362.

[3] Pastén, D., Torres, F., Toledo, B. A., Muñoz, V., Rogan, J., \& Valdivia, J. A. (2018). Physica A: Statistical Mechanics and its Applications, 491, 445-452.

[4] Abe, S., Pastén, D., Muñoz, V., \& Suzuki, N. (2011). Chinese Science Bulletin, 56, 3697-3701.

[5] Charbonneau, P., McIntosh, S. W., Liu, H. L., \& Bogdan, T. J. (2001). Solar Physics, 203, 321-353.

[6] Aschwanden, M. (2011). Self-organized criticality in astrophysics: The statistics of nonlinear processes in the universe. Springer Science \& Business Media.

[7]Morales, L., \& Charbonneau, P. (2008). The Astrophysical Journal, 682(1), 654.