Multi-hazard risk assessment in Adana, Turkey
Summary
The province of Adana, located in southern Mediterranean Turkey, has been exposed to multiple types
of natural hazards in the past. The most prominent hazards present in this area are earthquakes, floods
and landslides. In this area, studies have mainly focused on individual events or on specific locations.
An overarching framework of the multi hazards and their feedbacks, which is crucial in minimalizing
losses and city planning, is missing. This research is aimed to estimate earthquake, flood and landslide
hazard and risk. The estimation is done by calculating hazard separately for the three hazard types,
including their feedbacks and interaction with other hazards. After these calculations, risk was
estimated by applying the calculated hazard to a damage equation, resulting in three maps
representing monetary values of loss.
The results for earthquake hazard show that the southern part of the study area is dominated by high
peak ground acceleration (PGA) values, while a decrease in value occurs when going northwards.
Increasing the magnitude fades the circular buffer zones around historical hypocentra with high PGA
values in the south. The flood hazard map indicates that the steep and elevated north and south east
contain little to no hazard, while the flat and fertile south contains most hazard. The same steep slopes,
which were also noticeable in the flood hazard map, can be seen in the north and eastern part in the
landslide hazard map, since these account for most of the unstable pixels. All risk maps have a very
similar pattern based on the land use distribution of the study area. Differences between the hazards
are the spatial distribution of high value pixels within the urban environments and the fact that
landslides are less dependent on land use distribution patterns.
The methodology and data used for this research is excellent for assessing risk on a relatively large
scale. Land use can be obtained globally on a high resolution. Combining the calculated hazard with
the widely available and accessible land cover data is a great alternative for when the study area is
large or when data is scarce. However, availability and accessibility have a negative relationship with
accuracy, especially with landslide risk assessment. Studies making use of building inventories or other
intensive data to estimate risk show more accurate results on a higher resolution. Many risk
assessments are conducted making use of this methodology, but few use a multi-hazard approach or
are executed on a large scale. Using a multi-hazard approach on this scale will provide new insights in
land use planning, hazard adaptation, hazard prevention and hazard mitigation.