Inmaculada Aguado

Assessment of different topographic corrections in Landsat-TM data for mapping vegetation types (2003)

Different methods for topographic correction of Landsat Thematic Mapper images have been assessed in the context of mapping vegetation types. The best results were obtained with a variation of the C method, which takes into account the overcorrection of low illuminated slopes by the original C method. The performance of this method was tested using two criteria: the changes in the spectral characteristics of the image and the reduction in standard deviation of each vegetation type after the correction.

Estimation of fuel moisture content from multitemporal analysis of Landsat Thematic Mapper reflectance data: Applications in fire danger assessment

The objective of this paper was to define indices based on reflectance measurements performed by the Landsat Thematic Mapper (TM) sensor for estimating water content of live Mediterranean fuels for fire danger estimation. Seven Landsat TM images were processed and correlated with fuel moisture content (FMC) of several live species of Mediterranean grassland and shrubland. Raw bands were converted to reflectances, and several indices potentially related to water content were calculated from them. Pearson r correlation coefficients and linear regression analysis were computed in order to estimate FMC. Those indices based on the short wave infrared bands (SWIR: 1.4-2.5 w m) and on the contrast between this band and the near-infrared band offered the best estimations. For grassland, the integral of visible and SWIR bands provided the highest correlation, but also raw reflectances and Normalized Difference Vegetation Indices (NDVIs) provide significant r values ( r 2 above 0.8). For shrub species, indices that include SWIR reflectances performed much better than NDVI, because the SWIR band is more sensitive to water absorption, whereas NDVI estimates FMC indirectly, only from the effects of chlorophyll changes due to water variation content and leaf area index. The most significant relations were found with the derivatives of bands 4-5 and 2-3, and again the integral of visible and SWIR bands. Multiple regression analysis provided adjusted r 2 values of 0.84 for grasslands and 0.74 for shrublands. Average errors of 23.45-40% in the estimation of FMC for grasslands were found, depending on which variables were included in the multiple regression. For the FMC estimation of shrub species, errors were lower (from 7.94 to 19.40%), since the range of FMC values was also lower.

Design of an empirical index to estimate fuel moisture content from NOAA-AVHRR images in forest fire danger studies

An empirical estimation of live fuel moisture content (FMC) was generated from National Oceanic and Atmospheric Administration (NOAA) Advanced Very High Resolution Radiometer (AVHRR) derived variables. This estimation was based on the ratio of Normalized Difference Vegetation Index (NDVI) and surface temperature ( T s ), and the relative greenness (RGRE). It was derived for the summer season (from June to September). The estimation was calibrated from field measurements collected in central Spain, and included consideration of both herbaceous and shrub live FMC values. The proposed estimation worked reasonably well for other periods in the same study area and also for other similar study areas in Mediterranean environments. This estimation of FMC from remote sensing could be used as an input to standard fire danger and fire behaviour programs, providing spatially comprehensive data, which is critical for regional planning of fire prevention.

Integrating geospatial information into fire risk assessment

Fire risk assessment should take into account the most relevant components associated to fire occurrence. To estimate when and where the fire will produce undesired effects, we need to model both (a) fire ignition and propagation potential and (b) fire vulnerability. Following these ideas, a comprehensive fire risk assessment system is proposed in this paper,whichmakesextensiveuseofgeographicinformationtechnologiestoofferaspatiallyexplicitevaluationoffirerisk conditions. The paper first describes the conceptual model, then the methods to generate the different input variables, the approachestomergethosevariablesintosyntheticriskindicesandfinallythevalidationoftheoutputs.Themodelhasbeen applied at a national level for the whole Spanish Iberian territory at 1-km 2 spatial resolution. Fire danger included human factors, lightning probability, fuel moisture content of both dead and live fuels and propagation potential. Fire vulnerability was assessed by analysing values-at-risk and landscape resilience. Each input variable included a particular accuracy assessment, whereas the synthetic indices were validated using the most recent fire statistics available. Significant relations (P,0.001) with fire occurrence were found for the main synthetic danger indices, particularly for those associated to fuel moisture content conditions.

Prediction of fire occurrence from live fuel moisture content measurements in a Mediterranean ecosystem

The present paper presents and discusses the relationships between live Fuel Moisture Content (FMC) measurements and fire occurrence (number of fires and burned area) in a Mediterranean area of central Spain. Grasslands and four shrub species (Cistus ladanifer L., Rosmarinus officinalis L., Erica australis L. and Phillyrea angustifolia L.) were sampled in the field from the spring to the summer season over a 9-year period. Higher seasonal FMC variability was found for the herbaceous species than for shrubs, as grasslands have very low values in summertime. Moisture variations of grasslands were found to be good predictors of number of fires and total burned surface, while moisture variation of two shrubs (C. ladanifer L. and R. officinalis L.) was more sensitive to both the total burned area and the occurrence of large fires. All these species showed significant differences between the FMC of high and low occurrence periods. Three different logistic regression models were built for the 202 periods of analysis: one to predict periods with more and less than seven fires, another to predict periods with and without large fires (>500 ha), and the third to predict periods with more and less than 200 ha burned. The results showed accuracy in predicting periods with a high number of fires (94%), and extensive burned area (85%), with less accuracy in estimating periods with large fires (58%). Finally, empirical functions based on logistic regression analysis were successfully related to fire ignition or potential burned area from FMC data. These models should be useful to integrate FMC measurements with other variables of fire danger (ignition causes, for instance), to provide a more comprehensive assessment of fire danger conditions.

Estimation of dead fuel moisture content from meteorological data in Mediterranean areas. Applications in fire danger assessment

The estimation of moisture content of dead fuels is a critical variable in fire danger assessment since it is strongly related to fire ignition and fire spread potential. This study evaluates the accuracy of two well-known meteoro- logical moisture codes, the Canadian Fine Fuels Moisture Content and the US 10-h, to estimate fuel moisture content of dead fuels in Mediterranean areas. Cured grasses and litter have been used for this study. The study was conducted in two phases. The former aimed to select the most efficient code, and the latter to produce a spatial representation of that index for operational assessment of fire danger conditions. The first phase required calibration and validation of an estimation model based on regression analysis. Field samples were collected in the Cabaneros National Park (Central Spain) for a six-year period (1998-2003). The estimations were more accurate for litter (r 2 between 0.52) than for cured grasslands (r 2 0.11). In addition, grasslands showed higher variability in the trends among the study years. The two moisture codes evaluated in this paper offered similar trends, therefore, the 10-h code was selected since it is simpler to compute. The second phase was based on interpolating the required meteorological variables (temperature and relative humidity) to compute the 10-h moisture code. The interpolation was based on European Centre for Medium Range Weather Forecast- ing (ECMWF) predictions. Finally, a simple method to combine the estimations of dead fuel moisture content with other variables associated to fire danger is presented in this paper. This method estimates the probability of ignition based on the moisture of extinction of each fuel type.

Assessment of forest fire danger conditions in southern Spain from NOAA images and meteorological indices

Traditionally, the estimation of fire danger is performed from meteorological danger indices that are computed for single locations, where the weather stations are located. Frequently, these locations are far from forested areas, and there is a need to spatially interpolate danger variables. Methods for spatial interpolation are always prone to error, especially for those variables that show a greater spatial variability (wind, mainly). Satellite images may be considered a good alternative for interpolation of danger values, since they perform a spatially exhaustive observation of the territory. This paper analyses the spatial distribution of the Canadian Drought Code (DC), part of the Canadian Forest Fire Weather Index System (CFFWIS), in the region of Andaluc{@a (south Spain) following two procedures. First, maps of DC values were obtained from spatial interpolation of a network of 30 weather stations using the squared inverse distance algorithm. These results were compared with interpolation based on linear regression analysis, using National Oceanic and Atmospheric Administration (NOAA) Advanced Very High Resolution Radiometer (AVHRR) derived bands as independent variables. The most significant variables found for these empirical fittings were relative greenness, the ratio of Normalized Difference Vegetation Index (NDVI) and surface temperature, and a temporal variable, which accounts for the variations in day length throughout the fire season. After several empirical fittings were obtained, the most precise estimation was found after adjusting the coefficients to the time period considered.

Estimation of dry matter content in leaves using normalized indexes and PROSPECT model inversion

This work applies remote sensing techniques to estimate dry matter (DM) content in tree leaves. Two methods were used to estimate DM content: a normalized index obtained from the radiative transfer model (RTM) leaf optical properties spectra (PROSPECT) in direct mode and the inversion of the PROSPECT model. The data were obtained from the Leaf Optical Properties Experiment 93 (LOPEX93) database, and only 11 species were used in this study. The species selection was based mainly on the availability of data on fresh and dry samples. The estimation of DM content was obtained from an exponential function that correlated the values of the index proposed, (R2305 − R1495)/(R2305 + R1495), against the DM content of fresh and dry leaf samples. The determination coefficient obtained (r 2 = 0.672) was higher than the coefficient obtained from the inversion of the PROSPECT model (r 2 = 0.507). The data set used to validate the normalized index was provided by the Accelerated Canopy Chemistry Program (ACCP). The determination coefficient between the values obtained from ACCP data and the values estimated for the normalized index was r 2 = 0.767.

Meteorological fire danger indices and remote sensing

Meteorological fire danger indices with a specific focus on large fire danger rating, and their potential integration with satellite data to improve spatial and temporal resolutions of the estimates are the themes of this chapter.