Re-evaluation of topographic attributes with human population in deforestation framework with spatial dependency

Location: Room “Garda” - Polo Fabio Ferrari - Via Sommarive, 5 - Povo (Tn)

Speaker:

• Ryuei Nishii (Kyushu University, Japan)

Abstract:
This talk explores statistical modeling of forest area rates with two covariates. The forest coverage ratio of grid-cell data is modeled by taking human population density and relief energy into account. The likelihood of the forest ratios is decomposed into the product of two likelihoods. The first likelihood is due to trinomial logistic distributions on three categories: the forest ratios take zero, or one, or values between zero and one. The second one is due to a logistic-normal regression model for the ratios between zero and one. This model was applied to real grid-cell data and fits better than zero-inflated beta regression models. Furthermore, gridded forest area rates are regressed by spatio-temporal models based on natural cubic splines. Altitude and roughness of the grid will be also examined as additional explanatory variables of topographic features.

About the Speaker:
Ryuei Nishii received his PhD in mathematics from Hiroshima University, Japan, in 1981. He is currently a professor of the Institute of Mathematics for Industry, Kyushu University, Japan, where he teaches statistics, mathematics, and machine learning. His research interest includes spatial statistics, particularly the classification of geostatistical data based on statistical and machine learning approaches. Also, he is interested in statistical modeling and analysis of actual data related to industry, environment and space weather. He served as an editor of Japanese Journal of Applied Statistics and associate editors of academic journals including Transaction of Geoscience and Remote Sensing.

Contact: melgani [at] disi.unitn.it (Farid Melgani)