Drought alert systems for forest fire prevention often rely on vegetation water content (VWC) monitoring which is a key parameter in forest fire hazard. In southern France, VWC is now monitored... Show More

Drought alert systems for forest fire prevention often rely on vegetation water content (VWC) monitoring which is a key parameter in forest fire hazard. In southern France, VWC is now monitored through regular field surveys. Thanks to the theoretical sensitivity of shortwave infrared reflectance to VWC, MODIS satellite data are potentially able to monitor VWC depending on plant species VWC magnitude. In this paper, a specific statistical approach based on temporal cross-correlations is developed in order to test the correlation between two MODIS water indices and VWC measurements coming from field surveys. This test assesses the ability of daily MODIS data to monitor Mediterranean shrubland canopy water content and detect any delay effect between MODIS and field survey temporal series. Statistical tests are carried out for 29 sites containing 18 dominant shrubland Mediterranean species. 67% and 54% of significant correlation were found using respectively the NDII and NDWI indices from MODIS data. Correlation were found low with a dominant negative delay effect, i.e., with a MODIS signal that reacts a few days after the field VWC. Test results show that, even if deeper pre-processing of MODIS data may be required, site soil, site vegetation cover, and heterogeneity at MODIS pixel scale, as well as species VWC sensitivity make correlation between field VWC and MODIS water indices non univoque and highly variable. Many obstacles are still to overcome, for an accurate monitoring of Mediterranean shrubland canopy water content using MODIS daily data.

This article presents a lowland mapping method for the Forested Guinea (Guinée Forestière) using a 30 m resolution Digital Elevation Model (DEM) that is currently the best option to analyze large... Show More

This article presents a lowland mapping method for the Forested Guinea (Guinée Forestière) using a 30 m resolution Digital Elevation Model (DEM) that is currently the best option to analyze large forested areas. This low cost DEM method applies surface topography analysis processes to better discriminate areas with stagnant and/or accumulated water. The main index selected is the immediate proximity of flat areas to drainage network, the former with slope gradients not exceeding 5% (2.86°). The mapped lowlands potential cover a surface of 4516 km2, i.e., 10% of the total area of the region with hydro-agricultural potential.

With an increasingly technological improvement, sensors infrastructure actually supports many current and promising environmental applications. Environmental Monitoring Systems built on such sensors... Show More

With an increasingly technological improvement, sensors infrastructure actually supports many current and promising environmental applications. Environmental Monitoring Systems built on such sensors removes geographical, temporal and other restraints while increasing both the coverage and the quality of real world understanding. However, a main issue for such applications is the uncertainty of data coming from sensors, which may impact experts’ decisions. In this paper, the authors address this problem with an approach dedicated to provide environmental monitoring applications and users with data quality information.

The landscapes are highly dependent on the dynamics of local land use and land cover, which directly affects landscape structure and determines the spatial patterns of forest patches, as well as to... Show More

The landscapes are highly dependent on the dynamics of local land use and land cover, which directly affects landscape structure and determines the spatial patterns of forest patches, as well as to the major land uses within a specific region. The calculation of landscape metrics can support the understanding of such spatial distribution. In this study, 16 landscape metrics were analyzed in a drainage watershed in a high relief region in the Rio de Janeiro state, Southeastern Brazil, with the aim to evaluate the use of landscape metrics as indicators for agricultural management. Metrics calculation was followed by a Principal Component Analysis, which indicated the metrics that were most effective in evidencing the landscape structure in analysis. The results showed that the late-succession forest is the dominant component in the landscape. This class also presented the highest MPS metric value, related to the mean patch size by class. Some PCA results suggest that the metrics association was less effective in clustering the overgrown pasture, clean pasture, and annual crops classes, but this could result from the intrinsic association among those classes, by crop rotation, meaning the abandon of a site formerly occupied by an annual crop. Some metrics better suggested an interaction among land use classes and have potential to be use in the analyses of agricultural landscapes in high relief sites.

Fire is acknowledged to be a factor for explaining the disturbance of vegetation dynamics interacting with other environmental factors. In this study, the authors want to clarify the importance and... Show More

Fire is acknowledged to be a factor for explaining the disturbance of vegetation dynamics interacting with other environmental factors. In this study, the authors want to clarify the importance and the role of fire on the dynamics of savanna vegetation. The study area is the Marovoay watershed located on the north-west coast of Madagascar. In this site, burning herbaceous cover is the main practice in the extensive grazing system. They analyzed the relationship between two indicators, one related to vegetation activity changes and one about fire regime that results from a combination of fire frequency and seasonality. All indicators were measured between 2000 and 2007 using a time series of MODIS images. In this work, the authors implemented two approaches of spatial analysis. The first one analyzes the spatial structure of the residuals of a per-pixel non-spatial GLM model. In the second approach, a spatial GLM model is directly computed. In both approaches, the authors proposed two levels of stratification for the study area according to the spatial variations of the relationship established between vegetation activity changes and fire regime. The use of spatial statistical tools produces parsimonious models which they found to be consistent with expert knowledge. The authors demonstrated that a statistical analysis based on spatial GLM is able either to stratify an area when non ancillary data on land use exist or to validate an existing stratification.