"Assessment of the spread of 7 invasive species under three climate change scenarios in Korea using rule learning of elementary cellular automata"Jin, Hong SungThe spread of 7 invasive species (Lithobates catesbeianus, Trachemys spp., Pseudemys concinna, Mauremys sinensis, Macrochelys temminckii, Pseudemys nelson, Chelydra serpentine) is assessed in Republic of Korea. According to three climate change scenarios, SSP1-2.6, SSP2-4.5, and SSP5-8.5 in CMIP6 (Coupled Model Intercomparison Projects 6), the presence probabilities of invasive species are estimated by region using Maxent software. Expected distributed data consisting of cells with a value of either 0 or 1 are generated using the presence probabilities estimated by Maxent software. When the number of observation data is small or absent, we used occurrence records of species from GBIF (Global Biodiversity Information Facility) and climate data from the Worldclim database as inputs for Maxent modeling. The South Korean terrain is divided into 20km x 20km units according to latitude and longitude, and each unit is further divided into 400 cells of 1km x 1km in size to generate a sequence that evolves according to elementary cellular automata (ECA) rules. A CNN (Convolutional Neural Network) method is used to learn ECA rules and generate regional rule distributions. The degree of spread of each rule is assessed by averaging the number of 1s or pseudo-presences over 400 generations. |
« back