A machine learning-supported rapid classification of building damage follo…

Research papers, The University of Auckland Library

The rapid classification of building damage states or placards after an earthquake is vital for enabling an efficient emergency response and informed decision-making for rehabilitation and recovery purposes. Traditional methods rely heavily on inspector-led on-site surveys, which are often time-consuming, resource-intensive, and susceptible to human error. This study introduces a machine learning-supported surrogate model designed to streamline the assessment of building damage, focusing on the automated assignment of damage placards within the context of New Zealand's post-earthquake evaluation frameworks. The study evaluates two key safety evaluation protocols—Rapid Building Assessment (RBA) and Detailed Damage Evaluation (DDE)—and integrates corresponding databases derived from the 2010–2011 Canterbury Earthquake Sequence (CES) in Christchurch. Six ML classifiers—Multilayer Perceptron (MLP), Random Forest (RF), Support Vector Machine (SVM), K-Nearest Neighbours (KNN), Gradient Boosting Classifier (GBC), and Gradient Bagging (GBag)—were rigorously tested across both databases. The results indicate that the RF-based surrogate model outperforms the other classifiers across both RBA and DDE protocols. Two distinct sets of critical predictors have been further identified for each protocol, allowing for the rapid retrieval of essential data for future on-site surveys, while retaining the RF model's predictive accuracy. The developed surrogate model provides a pragmatic tool for practising engineers to rapidly assign placards to damaged structures and for policymakers and building owners to make informed recovery decisions for earthquake-affected buildings.

The regional consequence of a disaster: Assessing employment transition du…

Research papers, Lincoln University

Orientation: Large-scale events such as disasters, wars and pandemics disrupt the economy by diverging resource allocation, which could alter employment growth within the economy during recovery. Research purpose: The literature on the disaster–economic nexus predominantly considers the aggregate performance of the economy, including the stimulus injection. This research assesses the employment transition following a disaster by removing this stimulus injection and evaluating the economy’s performance during recovery. Motivation for the study: The underlying economy’s performance without the stimulus’ benefit remains primarily unanswered. A single disaster event is used to assess the employment transition to guide future stimulus response for disasters. Research approach/design and method: Canterbury, New Zealand, was affected by a series of earthquakes in 2010–2011 and is used as a single case study. Applying the historical construction–economic relationship, a counterfactual level of economic activity is quantified and compared with official results. Using an input–output model to remove the economy-wide impact from the elevated activity reveals the performance of the underlying economy and employment transition during recovery. Main findings: The results indicate a return to a demand-driven level of building activity 10 years after the disaster. Employment transition is characterised by two distinct periods. The first 5 years are stimulus-driven, while the 5 years that follow are demand-driven from the underlying economy. After the initial period of elevated building activity, construction repositioned to its long-term level near 5% of value add. Practical/managerial implications: The level of building activity could be used to confidently assess the performance of regional economies following a destructive disaster. The study results argue for an incentive to redevelop the affected area as quickly as possible to mitigate the negative effect of the destruction and provide a stimulus for the economy. Contribution/value-add: This study contributes to a growing stream of regional disaster economics research that assesses the economic effect using a single case study.