Estimación del contenido de humedad de vegetación mediterránea a partir de imágenes MODIS

摘要: espanolLa presente tesis doctoral se centra en la estimacion operativa del contenido de humedad combustibles vivos (LFMC) regiones clima mediterraneo utilizando informacion obtenida imagenes captadas por el sensor Terra-MODIS, ambito condiciones peligro incendio. En concreto, plantea como alternativa a los tradicionales modelos empiricos o indices meteorologicos, uso simulacion reflectividad (RTM), puesto que estos ultimos poseen una mayor capacidad generalizacion y rigor fisico. Asi, exploran diferentes formas adaptacion parametros entrada dos tipos mediterraneos (pastos matorral) varias tecnicas inversion, con proposito final conseguir metodologia operativa, precisa, robusta generalizable LFMC. Las estimaciones LFMC RTM son siempre contrastadas las obtenidas mediante calibrados datos tomados campo, objetivo verificar si primeros mas idoneos segundos. Finalmente, analizan relaciones entre pastizales matorrales probabilidad ocurrencia incendio (numero incendios, total superficie quemada grandes incendios) resaltar interes esta variable indicadora incendios zonas mediterraneas. Con demuestra tienen potencial e meteorologicos para gran escala, ya que, correctamente parametrizados, conducen robustas validacion. Todas conclusiones estan respaldadas once anos trabajo llevado cabo sistematicamente bajo protocolos estandarizados dominadas vegetacion mediterranea. EnglishThis thesis is focussed on the operational estimation of Live Fuel Moisture Content Mediterranean vegetation species from Terra-MODIS imagery in context fire risk assessment. The organised around seven papers. first paper (in Spanish) an exploratory analysis potential meteorological data and to assess by examining correlations between those datasets field demonstrates that among danger satellite derived are examined, latter produce better with and, therefore, potentially more suitable for studied. second English) deals comparing interest empirical relations statistical models based simulated reflectances radiative transfer (RTM). both good results, but present greater as they robust estimations, consequently expected have generalization power. This hypothesis tested third Spanish), ground measurements obtained at other areas dominated grasslands Central Spain Australia. confirmed provide similarly accurate results analogous types vegetation, than when applied structure composition different calibration site. Once demonstrated their interest, remains difficulty adequately parameterise these models. fourth English), fifth sixth papers propose use ecological rules avoid simulating unrealistic spectra, where parameters values never met simultaneously field. woodlands, addressing problem fixing demonstrating choice co-occurring observed provides results. analyses how used co-evolve shrublands which leads designation reduced realistic co-evolution ranges. approach proves its usefulness significantly decreasing residual error compared run unrestricted In paper, parameterized using some criteria explained two merit function compare spectra (relative minimum quadratic distance spectral angle) explored. concludes angle consistent measure similarity estimations also model calibrated (empirical RTM) measured four locations Spain., concluding been highly improved therefore global scale. last examines predictor occurrence ecosystem. It number fires, burned large fires. Three logistic regression then built predict three dimensions occurrence. discussion proposed significant do not sufficiently predictions all periods low or high since, expected, depends many factors. After there extra chapter (chapter 8) methodology should be followed generate cartography mean this presents examples. As conclusion, methods (meteorological data, considered operatively over reasonably area, advantages disadvantages compared. way improve parameterisation validated, making governing relationships imposing restrictions co-variations. ends confirming parameter prediction. All work supported work, systematically carried out along periods, validation steps.