LIMITACIONES ESPACIALES DE LOS MODELOS DE COMBUSTIBLES FORESTALES AL COMPARAR DOS DISEÑOS DE MUESTREO

Abstract

The fuel model concept (FM) is a practical tool to generate forest fuel maps, which are used to support fire behavior simulations. However, FM mapping has been one of the most difficult challenges that forest scientists have faced. Moreover, this concept has certain spatial limitations because it considers that fuel loading is homogeneous within an area qualified as a given FM. This condition was tested through two sampling methods in "El Largo y Anexos", ejido at Chihuahua state, where supposedly fuels have a stratified distribution, which is
opposite to MC. Data collection was divided in two phases. In the first one, datawere collected through a random sampling design, where 554 sample plots (1000 ²) were evaluated. The following forest fuels were inventoried: 1 hour lag time fuels (1HR), litter (H), and organic matter (OM). In the second phase data collection was made from the 554 sample plots of the first phase, based on a stratified sampling designo Thus, the two sampling designs were assessed and compared: Random Sampling Design (RSD) and Stratified Random Design (SRD). Results show that, in the case of 1HR, SRD was 674% more precise than RSD. In the case of H and MO, SRD was 276 and 360% respectively more precise compared with RSD. The fact that fuels loadings were better evaluated using a SRD suggest that the hypothesis of fuel homogeneity within a FM must be rejected.