Reports from the geoENV Conference, Greece

Last week the 15_^th International Converence on Geostatistics for Environmental Applications took place in Chania, on the Greek island of Crete. Five members of CERENA travelled to Greece to present the results of their academic achievements in the field of applied geostatistics.

Helga Jordão presented a model that uses a Bayesian neural network to predict landslides susceptibility of mountainous regions and calculate the related uncertainty. These models are particularly useful in times of climate crisis to provide decision makers with the information to implement targeted mitigation strategies to minimize impacts on public safety, economy and ecosystems.

Miguel Gomes introduced a geostatistical predictive model of drought severity applied to Southern Portugal. The study classifies drought levels based on the number of consecutive dry days, and uses machine learning to characterize short-term drought severity in the region.

Amílcar Soares illustrated a new multi-solution seismic inversion method based on genetic algorithms, to assess the risk and uncertainty at early stages of exploration of complex geological targets.

Ana Filipa Duarte presented a poster showing how localized Earth observation data can be integrated with seismic oceanography to predict spatial distribution of ocean temperature and salinity.

Bárbara Fonseca talked about the importance of monitoring the degradation associated to abandoned mines areas using multivariate analysis of satellite images. Her work is effective for delimiting mine tailing areas and to identify associated potentially toxic elements. 

Congratulations on your important contribution!