Maritime pine (Pinus pinaster Ait.) trees in the Landes de Gascogne, south west France, were suffered from catastrophic damage caused by Storms Martin (1999) and Klaus (2009). To find appropriate approach for examining damaged maritime pine trees in the region, first, the GALES model settings (mechanistic approach) and bias-reduced logistic regression models (statistical approach) were prepared with two airflow models, ARPS and WAsP, and a detailed field survey data from experimental plots located in Nezer Forest in the same region. Second, these models were examined using the French national forest inventory (NFI) data by a method of receiver operating characteristic curve. Both approaches significantly classified damaged and undamaged maritime pine trees when the detailed data was used, but using the coarser data sets all models did not prove adequate discrimination.