Coastal erosion has often been monitored by using remote sensing techniques such as satellite imagery, LiDAR and aerial photographs (Monitoring Oceans and Coasts, n.d.). The Geology Coastal Group has compiled historical records of aerial photographs to depict the ways in which Hawaii’s coastlines have changed over time (Hawaii Coastal Erosion, 2016). This provides helpful data in identifying coastal areas that have experienced excessive erosion. There may be unanticipated long term effects by relying solely on this type of data to make mitigation decisions without taking a closer look at what is happening at a larger scale. Therefore, a more detailed analysis, involving multiple spatial variables, is required to determine the most sustainable way of addressing coastal erosion in areas that need it most.
Research has shown that a vulnerability index is the most popular way to identify areas of concern in regards to coastal erosion (Hegde & Reju, 2007; Mujabar & Chandrasekar, 2013). As described by McLaughlin and Cooper (2010), a vulnerability index should simplify multiple complex and interacting parameters, represented by diverse data types, to form a more readily understood analysis. It makes the most sense to develop a criteria index model for this project because it allows the focus to be placed on various human activities that have been found to contribute to coastal erosion. All variables that were defined in objective 1 are placed into a scoring system where they are assigned a vulnerability level from level 3 being the highest risk to level 0 being no risk in increasing a coastline’s vulnerability to erosion. As seen in Table 1, each chosen variable can be placed within a ranking system where lower values represent a low risk in causing coastal erosion and higher values represent higher risks. The assigned values used in this research project are based on literature that describes the effects of human activity on coasts. A map can then be developed from the criteria index that identifies areas where the highest amounts of these activities occur close to the coast.
Each variable is ranked in the index model according to its type (i.e. agricultural or urban land use) or magnitude (i.e. large harbors or small harbors). The rivers with dams upstream, however, must be reclassified before being placed into the index model. Some dams may disturb a river’s sediment supply to coastlines more than others, so the level of disruption that each dam causes must be determined. The use of digital elevation models (DEMs) can aid in stream channel network analysis (Tarboton et al., 1991). Water will flow and accumulate according to the slope of the land, and therefore, by determining how and where water flows, the transportation of sediment can also be analyzed (Tarboton, 1997). Using flow accumulation models with GIS can depict fluxes of sediment loads which is useful in determining the level of disruption that dams can cause (Schauble et al., 2008). Dams will be classified in this project based on the amount of sediment transportation that they interfere with, and the size of the watershed that they are affecting. A dam is classified as high or low risk based on its influence on the accumulation of flow to the coast as well as the significance of the watershed that it is affecting. This is achieved by using Jenk's Natural Breaks to sort the resulting flow accumulation values for each dam (ranging from 1,089 to 43,180,463) into the three index values. This will be discussed in further detail in objective three.
Stream networks and flow accumulation show which river mouths are affected by dams located upstream. The impact of dams on coastal erosion is conveyed through the watershed areas of Oahu and Kauai. Ranking the watersheds that are affected by dams will show which coastal areas experience the most sediment interference by dams. The resulting watershed vulnerability is a map in and of itself and is used to compare to the coastal vulnerability map of other human variables that aggravate coastal erosion.