Mt. Ainos in Kefalonia Island, Greece, hosts a large variety of plant species, some of them endemic to the region. Because of its rich biodiversity, a large portion of the mountain area is designated as National Park and is protected from human activities such as hunting or logging. Therefore, the area presents a lot of opportunities for ecotourist activities, such as trekking, birdwatching, and mountain climbing. In order to estimate its touristic activities potential, it is essential to assess the mountain's biometeorological conditions. To achieve that, the human thermal index PET (physiologically equivalent temperature) was used, which is based on a human energy balance model. However, it is difficult to get the specific meteorological data over mountainous areas (air temperature, humidity, wind speed, and global solar radiation), appropriate as input variables for PET modeling. In order to overcome this limitation, artificial neural networks (ANNs) were developed for the estimation of PET index in ten sites within the Ainos National Park. In the process, the spatiotemporal distributions of the PET thermal index were illustrated, taking into consideration the ANN modeling. The findings of the performed analysis shed light that Mt. Ainos offers the greatest touristic opportunities from May to September, when thermal comfort conditions appear. The study also proves that the highest frequency of thermal comfort appears within the aforementioned time period over the highest altitudes, while on the contrary, slightly warm class appears as the altitude decreases on both sides of the mountain.
Keywords: Ainos Mt.; Artificial neural networks; Greece; Human thermal comfort; Kefalonia Island; Physiologically equivalent temperature; Tourism.