ABSTRACT
A
model of Environmental Sensitivity Analysis (ESA) was designed in GIS to identify the
sensitivity levels in four environmental components facing a potential oil
spill, along the shoreline of
The four environmental components along
the shoreline areas were delimited according to the size of the study area, and
identified as: Coastal mangrove habitats, coastal estuarine, sandy- beach coastal, and urban coastal. A survey was designed to
know the opinion of local experts on the sensitivity levels of such ecosystems, assigning a
numerical value to each type of coastal,
according to its high or low influence
on shoreline sensitivity. The distance to the
shoreline was also considered as a determining factor on sensitivity values.
The
fuzzy logic theory was applied for two purposes: first, for the implementation of model-based
decision rules, on the use of the concept of fuzzy numbers and linguistic variables;
second, as a proposal, for supervised classification of satellite images. It was necessary to gather and automatize the required information to be included in the
model, using digital techniques and visual interpretation of satellite
images. Bands were selected according to
Principal Component Analysis (PCA) technique. Land covers generated from maps (1:100.000) through digital
cartography show: routes and towns, hydrography.
Normalized Difference Vegetation Index (NDVI)
was used to generate land cover of dense and flooded vegetation.
Results obtained
revealed the different sensitivity values of each of the four environmental components
along the shoreline. Comparing the four models developed, it is concluded that
each one proved to be better than the preceding, beginning from Simple hierarchyzation
up to FAWM. Application of FAWM generated four final sensitivity maps (a, b, c, d) of
land covers or scenarios. Resulted land cover (d) is selected because it shows
the highest values, whose sensitivities are organized as following: (1)
Mangrove ecosystem, within
Comparing
the models’ results, the best one resulted to be FAWM, complemented with multicriteria
evaluation (MCE) techniques, which proved to be the most robust and accurate,
for its theoretical basis and easy-to- use model, its application under GIS,
and its flexibility in application under the concept of fuzzy logic, applied to
decision rules, which let to the construction of scenarios.
It
is concluded that the application of Multicriteria
Evaluation (MCE) techniques together with fuzzy logic, let to the development
of the best model in a satisfactory way, when modelling
environmental information with GIS. The use of these techniques permit to a
better decision-making process for the diagnosis and management of natural resources,
as well as the design of protective and mitigating measures against the risk of
oil spills.
Key
words:
Environmental Sensitivity Analysis, oil
spills, multicriteria
evaluation (MCE) techniques, fuzzy
logic, GIS, Remote sensing,
Se diseñó un modelo de Análisis de Sensibilidad
Ambiental (ASA) bajo Sistemas de Información Geográfica (SIG), para la detraminación de diferentes niveles de sensibilidad de
componentes ambientales, ante un derrame petrolero, ubicados a lo largo de la
línea de costa del Lago de Maracaibo, Venezuela. Se desarrollan cuatro métodos,
(1) ordenación simple, (2) adición simple coberturas con dos criterios, (3)
adición simple complementado con técnicas de Evaluación Multicriterio
(EMC) para la generación de pesos mediante la comparación por pares, y (4)
Método de Ponderación Aditiva Difusa (MPAD) utilizando
Se identificaron los diferentes componentes
ambientales que caracterizan la línea de costa, teniendo en cuenta el tamaño
del área de estudio, resultando los siguientes tipos: costas de manglares,
costas estuarinas, costas de playa arenosa y costa
urbana. Se diseñó encuesta para solicitar la opinión de expertos en cuanto a su
consideración de los niveles de sensibilidad de tales componentes, y asignarles
los pesos correspondientes de los tipos de costa que ellos consideran tienen influencia en la sensibilidad de la línea de costa.
Igualmente, se utilizó el criterio distancia a la orilla, en la determinación
de las sensibilidades.
Se utilizó
Los resultados obtenidos
están dados por las diferentes sensibilidades de los componentes de la línea de
costa. Al comparar los cuatro modelos se establece que cada uno es una mejora
del anterior, comenzando por la jerarquización simple
hasta el MPAD. Con el MPAD se generan cuatro coberturas o mapas finales de
sensibilidades (a, b, c, y d), o escenarios. Se escoge la cobertura con los
valores más altos (d), cuyas sensibilidades se organizan en el siguiente orden:
el tipo de costa más sensible es el manglar, en sus primeros
Al comparar los modelos, el
producido con el MPAD, que aprovecha los pesos generados con
Se concluye que la aplicación de las TEMC y
Palabras claves: Análisis de Sensibilidad Ambiental,
Derrames petroleros, Técnicas de Evaluación Multicriterio,
Lógica Difusa, Sistemas de Información Geográfica, Sensores Remotos, Lago de
Maracaibo, Venezuela.