Source Document: Marco Pregnolato, Marcello Petitta, Christian Iasio, Giedrius Kaveckis, Stefan Schneiderbauer, Lydia Pedoth. 31-05-2015 Disaster impact and land use data analysis in the context of a resilience relevant footprint. Deliverable 3.2
The “Disaster footprints and maps” report aims to contribute to a better understanding of how coarse scale quantitative data can prove useful in a study on resilience. The main goal of the work is the exploitation of large datasets in search of indicators and information valuable in one or the other way for resilience research. A quick overview of the different understandings of the concept of footprint is presented. A particular matter of interest is the possibility to check whether the land and its use could be related to dimensions relevant for resilience through the use and elaboration of CORINE Land Cover data.
The investigation, set up in order to analyse these “signs on a map”, is based on the position that a footprint is multi-dimensional, possibly bridging quantitative measurements and qualitative indicators. The application of such a methodology on a supra-national scale analysis could provide indications useful in the definition of the domains to be examined within the perspective of a resilience study at a local scale investigation.
The concept of ‘footprint’ could be explained as a twofold approach: a first part in which the cause-effect mechanisms in a chain leading to an impact and following backwards the chain of effects of a process has to be identified. A second part in which a model is designed to project and estimate potential future impacts. The methodology proposed in the full document follows this structure.
For the landslide we recognize in LULC one possible (not the only one) cause for the activation and we design a model to monitoring, assessing and projecting the potential LULC impact for landslide. We finally presented a methodology to use LULC as a footprint of landslide.
In about 81.6% of the landslides the LULC does not show any change. When we consider natural areas, such as forest, mountains, etc., which after a landslide event do not present any change in LULC classification, we can relate this stability to the capacity of the territory to recover and to return to the previous conditions (restoring the vegetation, etc.), which could be related to a kind of territorial or natural resilience.
Conversely, not being able to observe any change in LULC after a landslide within an area associated to human activities (agriculture, urban settlements, roads, etc.) could also be seen as the return to a previous risky and hence non-favourable condition. It could be argued that rebuilding a street, or an urban settlement, or an industrial structure in a landslide prone area after the occurrence of an event means that the consideration of possible future events in the planning process has not taken place appropriately. Hence, an important aspect of resilience, the capacity of learning and adapting to certain situations has apparently not taken place. However, in practice this would need a case-by-case assessment since it could be possible that measures to protect from future damage have been taken but that those measures did not lead to a change of land use class.
The added value of this methodology is in introducing an objective and reproducible index, the HTI. As presented in the report, the HTI definition is based on state of the art literature review. The index takes into account on one hand the LULC to characterize the territory and on the other hand, the other physical variables associated to the territory. The introduction of an objective index for landslide classification and monitoring represents a powerful instrument, which can be exported to other regions and countries.
Concluding, the LULC analysis when used in combination with other data-sources, can provide relevant information, in particular when high resolution LULC are considered. This study indeed shows the importance of the spatial scale when disaster footprint are studied. Considering the results obtained in the Part B of the document, in which regional scales and high-resolution LULC dataset have been used, we can conclude that LULC is a relatively useful instrument when local scales are considered. Conversely, for national scales, only general interpretations can be drawn.
From the results obtained at regional scale we can confirm that this methodology has potential and this potential will be more and more useful for an assessment in practice with increasing quality and higher resolution of the datasets in the future (namely LULC data, landslide event data and data on damages). Reason is that landslides are mostly triggered and influenced by factors that are spatially explicit at local scale and that spatial datasets of high quality improve footprint assessments.
The applied methodology introduced here can be replicated in other areas providing an integrated approach for disaster footprint identification.