METHODOLOGY

 

Landslides need to be located and defined before analysis of their effects on the landscape and the consequences in the human societies and physical factors can be examined.

 

Three main steps will be used to analyse landslides across the Canary Islands. Firstly, an investigation will be conducted based on previous work in the literature review and the articles about the landslides in the Canary Islands where several authors do a location of the landslides. This will be achieved by visiting areas known to have landslides and a geomorphological recognition of the landslides areas looking for cliffs, valleys or any geomorphological footprint in the relief caused by landslides. Secondly, with this information of the different authors and the geomorphological recognition, the location and the landslide areas will be defined and digitalisation will be done in the different areas affected by landslides in Tenerife. Lastly, ArcGIS software will be used to analyse and study the landslides areas. Alongside these main steps, statistical analysis of data will also be undertaken.

 

As previously mentioned in the introduction, Tenerife will be fully investigated and compared with El Hierro. Due to limitations in the field work, El Hierro and the rest of the islands will not be visited for ground truthing. Observation in Tenerife will be compared with El Hierro when enough information has been generated to correctly analysis.

 

i. Geomorphological recognition

 

The proposed area will be ground trothed via three routes based on the distribution of the landslides in Tenerife according to research by Carracedo et al., (2009); Masson et al., (2002) and Seisdedos et al., (2012).

 

Anaga, Güímar and La Orotava are locations know to have landslides. Field trip 1 intends to join these locations together enabling the investigation of evidence of landslides including cliffs and steep slopes that outlines the boundaries of the landslides. Güímar and La Orotava are known landslides and it is likely to find sufficient evidences. However, as stated by Romero (1986), insensitive erosion processes may affect the analysis of indicative characteristics of the landslides in the older part of the island such as Anaga. This route has also been chosen for the analysis of land use in landslide areas and has been outlined as on the objectives.

 

Icod, Roques de García and La Orotava are landslides in the north of the island. Field trip 2 has been chosen to expand upon field trip 1. Due to post-volcanic activity having once very strong in this area of the island the edge of the different landslides may be hard to locate. Overlapping of the landslides may also affect the cliffs and steep slopes.

 

Vegetation distribution in the north of the island could be considered important because this area is affected by trade winds (Dorta, 1996) and the most important landslides are in this area.

 

Field trip 3 will be in the west of the island, where landslides of Teno and Bandas del Sur can be located. Teno is an old part of the island similar to Anaga and may consist with erosion processes that affect the signals of the landslides. Volcanic activity has increased Teno’s land area into the ocean, which in turn has increased the landslide area (Yanes et al., 1988).

 

Bandas del Sur may be difficult to identify, due to a lack of literature and evidence being unavailable.

 

ii. Location, definition and digitalisation of landslide areas

 

Evidence derived from the geomorphological recognition and orthophotos (aerial photographs) of Tenerife will be used to make maps with locations of the landslides. Orthophotos are in the public geographic maps website of the Government of Canary Islands, GRAFCAN.

 

Photos and 3D vision in GRAFCAN will be used to define the landslides around Tenerife. Moreover, information from different authors about landslides in Tenerife (Alonso et al., 2005; Masson, et al. 2002; Anguita et al., 1999 and Seisdedos et al., 2012) and geological data from GRAFCAN will be used to delimit the landslides properly observing scarp slopes or changes in the composition of the soil.

 

El Hierro and the rest of the islands will be located and defined without the ground trothing.

 

Finally, landslides will be digitized creating a polygon layer over the maps and relief of the islands in GRAFCAN. This layer will be the base map for the rest of our methodology.

 

Objetive 1 and 2 will be done with the location, definition and digitalisation.

 

iii. Geoprocessing of landslides areas with ArcGIS

 

The Landslides layer will form the base of this part in the methodology and will be used in ArcGIS software. This layer will be utilized to obtain the shape of the landslides in the Canary Islands. This shape will be polygons and it will be used to determine the areas affected by landslides in the different maps (slope, streams networking and land occupation mainly).

 

Shape will give information about areas in Km2 using the attribute table, also the areas of the landslides could be compared and determinate the affect area of the landslides in the different islands. This observation about the areas affected could give very relevant information about the influence of the landslides in the different islands and if it is an important factor in the landscape in geomorphological terms.

 

Maps used in the introduction of this dissertation about the age of the landslides and islands or the frequency of this process will be produced with this information because, new data will be added to the attribute table as age of the islands, age of the landslides or volume of land affected in each island.

 

Shapes of polygons of land occupation of Tenerife and El Hierro will be obtained in GRAFCAN because it will be very complex and the aim of this dissertation will be to analyse the effects of the landslides in the landscape. Also, the DTM map (black and white basic map) of Tenerife and El Hierro will be obtained from the U.S. Geological Survey which will be used to obtain the slope map and hydrographic network map and in general, all relief variables.

 

Slope, network hydrographic and a reclassified land occupation maps will be the main result of the slapes of polygons and DTM. Maps as streams or contours could be used to understand more correctly the study area and analysis properly.

 

The slope map will be in percent and it will show the different break in the slopes around the landslides. DTM map will be added in ArcGIS and the file will be projected in UTM coordinate system. In the ArcToolBox > Spatial Analyst Tool > Surface > Slope. Input raster will be the DTM map and the output measurement will be percent. In the same route, we could do the counter map.

 

Stream networks will need the DTM map as well. Fill, Flow accumulation and flow direction will be necessary to do this map and these options will be find in the ArcToolBox > Spatial Analyst Tool > Hydrology.

 

- Fill map: This function fills the sinks in a grid. If cells with higher elevation surround a cell, the water is trapped in the cell and cannot flow. This function eliminates this problem. (raster map)

- Flow direction: Using the fill map, a new map will be obtained with the different directions of the flows. (raster map)

- Flow accumulation: the map of flow direction will be used to compute the accumulated number of cells that are draining to any particular cell in the MDT.

 

The flow accumulation will give the number of cells that drain to a particular cell, it can be used to define a stream networks.

 

Reclassified land use maps will be done with the land use map. Originally there are twenty six different types of land use and a map with a new classification with five different types will be done. The new groups will be Urban, agriculture and livestock, with vegetation, without vegetation and industrial.

 

- Urban: Roads, infrastructures, dams and towns.

- With vegetation: seven different types of vegetation.

- Without vegetation: areas with vegetation and volcanic areas where the vegetation is sparse.

- Industrial: Industrial.

- Agriculture and livestock: ten different types of crops and two different types of livestock.

 

After these processes, all maps will be general for Tenerife and El Hierro so we will need to get these maps for the landslides areas using a tool in ArcGIS called Clip (ArcToolBox > Analysis Tools > Extract > Clip. Land occupation and physical variables in the different landslide areas will be obtained after this process using the option merge (ArcToolBox > Data Management Tools > General) and dissolve (ArcToolBox > Data Management Tools > Generalization).

 

All maps must be projected in UTM using (ArcToolBox > Data Management Tools > Projections and Transformations > Raster > Define Projection.

 

Finally, the contrast between an area with landslide and without landslide will be the last process because this information will be relevant to do a valuation of the impacts of the landslides in the landscape, and if there are differences between the areas affected by the landslides and the areas without this process. Güímar town area will be used to do this because is a town where city is between both areas.

 

The objectives 2, 3 and 4 and the aims 1 and 2 will be studied with these techniques.

 

iv. Statistical analysis

 

Microsoft Exel will be used for the statistical analysis in this part of the methodology. Basic information will be obtained from the different articles and the attribute tables that maps will generate. This information will be useful to understand the process and quantify the different volumes and areas of the landslides. Secondly, comparison between landslides and areas with and without landslides will be easier with these data. Graphics can be a useful tool to better interpret all data.

 

This information will help to define the answer to the objectives and aims in general, but it is the cornerstone to explain the objective 5 and the last aim of this dissertation.