GeoSpatial Data Processing to Support Seismic Emergency Management

Developing a WebGIS Service to Support Locational Choices of Shelters

1. Problem Statement:

Earthquakes pose a significant threat to urban areas, often leading to catastrophic loss of life and property. A critical challenge in earthquake-prone regions is the rapid identification and allocation of safe emergency shelters. Traditional methods for selecting shelter locations often rely on subjective criteria and lack real-time adaptability. This limitation is further exacerbated by the dynamic nature of urban landscapes and evolving demographic patterns. During the course of "GEOSPATIAL DATA PROCESSING TO SUPPORT SEISMIC EMERGENCY MANAGEMENT", student of the Civil Engineering for Risk Mitigation (CERM) program in Polytechnic University of Milan (PoliMi), try to address this challenge by leveraging advanced geospatial data processing techniques. The aim is to develop a comprehensive, data-driven approach for identifying optimal shelter locations in Amatrice, Italy, that are safe, accessible, and well-equipped to meet the needs of affected populations in the event of an earthquake. This innovative approach that we took during our project, promises to enhance emergency preparedness and response, potentially saving lives and reducing the impact of seismic events.

2. Background:

The impact of natural disasters goes beyond the loss of human lives. These events can cause significant economic damages, displacing people from their homes and communities, disrupting livelihoods, and affecting infrastructure and public services (Boustan et al., 2020). According to a report by the United Nations Office for Disaster Risk Reduction, the economic losses associated with natural disasters have been increasing in recent decades, with an estimated $2.97 trillion in damages reported between 2000 and 2019 (United Nations Office for Disaster Risk Reduction, 2020). It is therefore essential to take proactive measures to reduce the risk and impact of natural disasters through disaster risk reduction strategies, preparedness plans, and effective response mechanisms (Tansey et al., 2018).

Based on these historical records, it is evident that there has been a considerable decrease in the number of fatalities caused by floods and draughts, due to improved prediction methods, more resilient infrastructures, and better emergency preparedness and response system. However, low- frequency and high-impact events such as earthquakes are neither predictable nor preventable.

Nonetheless, the extent of damage caused by earthquakes can be mitigated through effective disaster management planning (Bello & Aina, 2014). While it is impossible to prevent or predict some natural hazards like earthquakes, their adverse impacts can be minimized through a variety of actions and strategies. Implementing robust mitigation and preparedness measures can reduce communities' exposure to hazards and enhance their resilience to disasters(Lam & Kuipers, 2019).

3. Data

The data that we have used to do the analysis in order to find the best locational choices for shelters for both emergency, and recovery phases, is mostly in GIS-software-readable formats, such as shapefiles, or rasters, as explained below

3.1. DTM layer: it came in TIFF format, and was downloaded from Lazio, Regional Geoportal

3.2. Slope layer: it was computed with QGIS from the Digital Terrain Model (DTM) using the slope function on raster layers which calculates the rate of change in the elevation for each cell in the DTM

3.3. Population: it came in the shapefile format and was downloaded from Lazio, Regional Geoportal

3.4. Soil Type: it came in shapefile format and was found from the International Hydrogeological Map of Europe HME1500 consists of a series of general hydrogeological maps at a scale of 1:1,500,000, covering nearly the whole European continent.

3.5. Land Cover: it came in shapefile format from the Copernicus platform

3.6. Floods and Faults: both of these two layers also were downloaded from Lazio, Regional Geoportal in shapefile format.

3.7. Hospitals, Police Stations, Power Grid Network, and Fuel Station were also found from Open Street Maps

4. Approach and Framework:

we developed a 4-phase framework in which the first two phases aimed to filter out unsuitable areas of land based on their geological characteristics and susceptibility to geohazards. To summarize our workflow, during the first two phase, we eliminated portions of land unsuitable due to geological factors such as slope, soil type, and landcover as well as the areas susceptible to geohazards, specifically floods, landslides, and presence of fault lines.

In the second phase, we further eliminated areas prone to the natural disasters, as well as those near fuel stations or electric grid network, due to the risk of fire or explosions and collapse of these grids. By the end of this phase, we had identified potential suitable areas in the event of an earthquake or any geohazards. Nevertheless, there are additional factors that can influence the suitability of an area for sheltering purposes. These factors fall into the categories of services and lifelines.

In the third phase, the level of the suitability of shelter areas was ranked by applying an increasing factor to the outcome of the previous phases. Considering the proximity to medical centres (hospitals) for healthcare access, proximity to police stations for security considerations, accessibility to water resources, and accessibility to the city's primary and secondary road network as well as the proximity to the high- density population zones.

Furthermore, in the fourth phase, further categorization was performed to the ranked suitable areas obtained from the third phase to highlight the most favourable ones with respect to: medical needs, road accessibility, and distance to the city centre. Identification of the areas for medical needs during emergency would help the user to efficiently distribute the elder and those who need medical assistance. The most accessible areas will help to accommodate those who don’t have access to vehicles or with disabilities. Nonetheless, the areas in the proximity to the city centre were considered more favourable for the SAR teams for the reasons mentioned in.

A WebGIS application using ArcGIS online has been designed. The portal contains the crucial maps needed during the emergency namely, Ranked Suitable Areas, Classified Areas based on Needs and Suggested Areas for Search and Rescue (SAR) teams. In addition, a dashboard has been developed to make use of the updated data during the emergency phase of the event. For instance, number of people to be evacuated, total evacuated and shelters’ capacity. Since the ArcGIS online is a commercial product and need licencing, we have also published our results on the GeoNode platform as well, which acts a redundancy for the ArcGIS Online.

5. Methodology

5.1. Phase 1: Studying Geological Criteria to Find Suitable Land

5.1.1. Suitability of Land based on Land Slope: One of the crucial factors to consider when determining the suitability of a land area for choosing shelter location is the land slope. Extensive research conducted by (Liu et al., 2011) has identified that areas with slopes exceeding 10 degrees are deemed unsuitable for sheltering purposes due to the possibility of soil erosion. Consequently, these areas have been assigned a weight of 0 in our classification system to ensure their exclusion in subsequent stages. Conversely, to ensure effective drainage (Camp Site Planning Minimum Standards | UNHCR, n.d.).

5.1.2. Suitability of Land based on Soil Type: Soil permeability is a crucial factor to consider when determining suitable locations for shelters. The permeability of soil is influenced by various factors such as soil porosity, infiltration capacity, and soil saturation. Clay-rich soils, for example, have low permeability and infiltration, leading to high surface runoff. Consequently, it would not be advisable to construct shelters on such soil. Conversely, soils predominantly composed of sandstone exhibit higher permeability compared to clay soils, resulting in reduced surface runoff.

5.1.3. Suitability of Land based on Land cover: The type of land cover is another crucial factor to consider when selecting suitable locations for shelters. It is evident that areas densely covered by the trees and vegetation are unsuitable for sheltering purposes. Conversely, areas with moderate vegetation can serve as suitable sites for establishing shelters. These areas can effectively reduce dust and erosion, and compared to forested areas, they pose a lower risk of fire hazards.

5.1.4. Phase One Outcome:

5.2. Phase Two: Studying Hazard Susceptibility Criteria to Find Suitable Land

5.2.1. Susceptibility to Flood: Flooding events can cause significant losses of lives, extensive damage to infrastructure and disruption of essential services (Dall’Osso et al., 2009). For this reason, the flooding susceptibility layer was considered. The areas that are prone to flooding have been disregarded to highlight those areas that poses the opposite definition

5.2.2. Susceptibility to Seismic Hazards: Studying the area for seismic hazards is an utmost importance. The damage caused by such hazard will be fatal to the city’s infrastructure, buildings, and lifelines. Due to the unavailability of Earthquake Hazard Assessment layer for the city of Amatrice, we have used a layer that show the active faults in the area and disregarded the areas that lie around them (as a second alternative, shake map of a previous event could have been used to classify the areas for seismic susceptibility based on their recorded PGA)

5.2.3. Susceptibility to Man-Made Hazards: Man-Made hazards can vary from infrastructure collapses to industrial accidents. Considering these areas that happen to have a potential hazardous effect in case of damage due to natural disaster (e.g., earthquake, flooding) is crucial. To avoid secondary damage to evacuees, potential shelter areas that fall within 500 meters of earthquake faults or high-risk areas such as gas stations and powerplants to be excluded from the suitable areas for shelters (Sustainability 2017, 9, 2098). Also, shelters that fall within 250 meters of power towers are to be excluded.

5.2.4. Susceptibility to Landslides: Landslides can have a significant impact on the overall hazard and subsequent damage caused by the earthquake. The assessment of landslide’s susceptibility will allow for a comprehensive understanding of the suitability of the area in the second phase. The areas that have been disregarded are those that pose a high and very high susceptibility.

5.2.5. Phase Two Outcome:

5.2.6. Phase One and Two Combination:

5.3. Phase Three: Ranking Areas Based on Serviceability and Lifelines

As explained before, during the first two phases the approach was to eliminate unsuitable areas due to the geological characteristics of the land, and its susceptibility to different hazards by assigning (multiplying) weights to them accordingly. Consequently, at the end of second phase, the areas of the land which are suitable for sheltering purposes remains, irrespective of their accessibility to services and lifelines.

In the third phase of this case study, and in order to take into account the proximity of the suitable lands to different services and lifelines, the relative-weights from the outcome of phase one and two were multiplied by increasing factors according to their closeness to the services. The closer the area to a service the higher factor it will be multiplied to.

5.3.1. Ranking the Areas based on Accessibility: The accessibility has been assessed based on the distance to the roads from the suitable areas developed in the outcome of phase 1 & 2. Buffering zones with different distances around the road network were analysed. The areas overlapping these zones were assigned different weights. Moreover, the suitable areas that were not accessible by the road network were excluded in this phase.

5.3.2. Ranking the Areas based on Proximity to Medical Services: Medical service access is a very important aspect to consider when locating a shelter. In the city of Amatrice, we have 2 medical services centres. Buffering circles around these two areas have been made and the suitable areas from the previous phase outcome is being overlapped and given weights.

5.3.3. Ranking the Areas based on Proximity to Water Sources: Water resources expressed in the available river and water streams can be beneficial for the evacuees in order to find a sustainable water source, especially during the early phases of constructing the emergency shelter. Nonetheless, the assigned factor was given the lowest weight compared to the other factors.

5.3.4. Ranking the Areas based on Proximity to Police Stations: Placing shelters close to police stations would provide a level of security to the places placed around them. Buffering circles with different radii have been performed and each class has been given a weight to it.

5.3.5. Ranking the Areas based on Proximity to Population Densities: The suitable areas have been overlapped over the population areas. The areas have been classified based on the population density number corresponding to the overlapped suitable areas, after removing the areas that overlap with actual residential buildings

5.3.6. Phase Three Outcome:

6. Summary and Final Notes:

As you may have noticed, this project was done in 4 phases. But here the first 3 phases are presented so far. the 4th phase, was the development of a WebGIS tool to present the same results. For this purpose, we developed a web map with ArcGIS online for the decision-makers, as well as an open-source GeoNode Server to be accessible to the public.

Other than that, our team developed a dashboard along with the web map, in which the decision-makers could monitor the shelters capacity, and its accessibility in a real-time manner.

Since the web pages are not accessible anymore, I have provided some screenshots of it in the images below.