New Zealand is situated on the boundary of the Australian and the Pacific tectonic plates. Off the east coast of the North Island, the Pacific Plate subducts beneath the Australian plate (Figure 2-1). Off the south west coast of the South Island, this feature is reversed with the Australian Plate subducting under the Pacific plate. The transition between these subduction zones occurs through the top east of the South Island and down the west coast in the form of the transpressional Alpine Fault (Bain, 2014). This fault has an estimated rupture reoccurrence interval of ~330 years and has had up to ~480 km of movement in the form of multiple magnitude ~8 earthquakes (Castelltort, et al., 2012).
Tectonic setting of New Zealand adapted from (Bain, 2014)
At the northern end of the Alpine Fault, the system begins to transition into the Hikurangi Trench subduction zone (Figure 2-1 and) as multiple faults splay off creating the Marlborough Fault System (MFS) (Wilson, Jones, Molnar, Sheehan, & Boyd, 2004). The four major splay faults include the Wairau, Awatere, Calrence, Kekerengu and Hope faults and each have a moment magnitude potential of up to 7 Mw or possibly greater (Robinson & Davies, 2013).
At 12.02 am on 14 November 2016, a complex sequence of 21 fault ruptures in the MFS, including a section of the Hope and Kekerengu faults. The ruptures produced a series of earthquakes with a combined magnitude of 7.8 Mw. Their collective energy spread 250km northward at 7,200 kilometres per hour with the seismic shaking lasting up to two minutes. Strong shaking was reported throughout New Zealand’s North and South Islands (GNS, 2016).
At Kaikoura, a massive shoreline shelf was thrust upwards, while parts of the South Island were shunted more than 5m closer to the North Island and other parts raised by up to 8 m (T+TI, 2017). The resultant tsunami waves reached a peak height of about 2 m at Kaikoura, 7 m at Goose Bay and 5 m high at Little Pigeon Bay on Banks Peninsula which knocked a cottage off its foundations (T+TI, 2017). Tsunami waves were also recorded in Wellington Harbour, Castlepoint, Lyttleton and the Chatham Islands. However, there was no significant damage to infrastructure (T+TI, 2017). State Highway One (SH1) and the main trunk rail line running to the north and south of Kaikoura were blocked by a few of the expected 80,000 to 100,00 landslides that occurred after the shaking (GNS, 2016). The secondary inland Kaikoura road had also been blocked. With road and rail links severed, Kaikoura township - home to approximately 3,700 people – and its wide rural catchment scattered with many farms and tiny, remote communities, were only accessible by air (T+TI, 2017). Massive changes to the sea shore and seabed had rendered the town’s port commercially inoperable until dredging of the harbour could occur. At least 150 landslides had blocked rivers creating landslide dams (GNS, 2016). Dam outbreak floods created a high risk for townships down stream with some towns needing to be evacuated during periods of higher rainfall during the days and weeks after the initial earthquake (GNS, 2016).
Active faults in the Marborough Fault System adapted from the Kaikoura Earthquake Viewer (T+TI, 2016)
Rapid Damage Mapping
In a natural disaster, such as an earthquake, accurate and locationally precise information on the damage caused is vital for prioritised rescue and relief operations, to mobilise resources for repair and recovery (Saito & Spence, 2004). An early response in case of strong and destructive event is very important to support and to manage the rescue activities (Dell’Acqua, Bignami, & Chini, 2011). Conventional methods of information gathering that often rely on teams studying the damage on the spot are slow and incompatible with immediate relief (Robinson & Davies, 2013). There is now considerable interest in the acquisition, interpretation and use of information from remote sensing such as satellite imagery or unmanned aerial vehicles (Saito & Spence, 2004). This is because of the opportunity it presents to gather information over a wide area quickly, consistently, and independently of the situation on the ground (Saito & Spence, 2004). Both qualitative and quantitative methods can rapidly create a damage map that visualises the distribution of damage and types of damages observed on buildings (Saito & Spence, 2004). Depending on the quality of the images obtained, the accuracy is typically suitable to support early emergency and rescue planning in areas which need the most immediate help (Saito & Spence, 2004). To create a rapid damage map, the following four processes are required:
1. Obtaining information on the building stock, e.g. footprints, function of the building, height, area size, material of building.
2. Assessment of the damage to buildings, such as total collapse or heavy damage to the structure.
3. The creation of damage maps. The maps could take the shape of grid-based damage maps or building-by-building damage maps, when possible.
4. Road network, access and damage location maps (Saito & Spence, 2004)
During the Kaikoura earthquake, IRDR’s Disaster Loss DATA project and the CODATA Task Group LODGD worked together with T+TI to provide TripleSat satellite images of the affected Hurunui District. Geo-spatial information was developed for the New Zealand Earthquake Commission (EQC) on the damage caused, and was made available through a web-based viewer to all government agencies, response and recovery agencies, engineers and researchers, thereby informing ﬁrst response and mitigation measure (Fakhruddin, Murray, & Maini, 2017).
Loss Data Collection for Natural Disasters
Damage and loss estimation is often difficult immediately after a natural disaster since data and information are not easily obtainable (Fakhrudden, Murray, & Maini, 2017). When human, monetary or environmental losses occur because of a disaster, extensive loss data is often collected and stored by different organizations, but the thoroughness and accuracy of the data can vary among local entities (Fakhruddin, Murray, & Maini, 2017). The data collection is pivotal to the comprehensive assessment of disaster impacts and can be hard to obtain due to the differences in the form it is presented in by different agencies. Centralised disaster loss databases are crucial to producing and acting upon risk information (Fakhruddin, Reinen-Hamill, & Murray, 2017). It also allows data to be shared across all agencies to help ensure a more efficient and prioritised response. Once the response is over, risk interpretation, with standardized loss data, can also provide loss-forecasting data in referencing historical loss modelling (Fakhruddin, Reinen-Hamill, & Murray, 2017).