Enhancing Disaster Resilience through the Reproducibility of Social Vulnerability Models

Social vulnerability models are a type of analysis that assesses a population’s potential for loss based on a myriad of socioeconomic indicators that either increase or decrease a demographic’s resilience against disasters (Cutter et al. 2003). They are becoming increasingly important as climate change worsens consequently intensifying the frequency and unpredictability of disasters. In order to be able to respond to these heightened disasters, social vulnerability models are a necessary tool to highlight which areas will face the greatest impacts and require the most support and resources in order to overcome them. Due to their position of being at the intersection of quantitative and qualitative analysis, social vulnerability models face the challenge of consistency across diverse places. There is no single set of socioeconomic indicators of vulnerability that will work for every model for each disaster. Due to the diversity of conditions that surround uniquely diverse populations and their vulnerability to disasters, social vulnerability models that adequately represent their specific circumstances have a lot of subjectivity in the research design process in the steps of how indicators are chosen, how the indicator data is normalized, how it is weighted, how it is aggregated, and finally how it is visualized. These steps that have a level of subjectivity leave a significant room for uncertainty. When a spatial analysis such as this has such a great political sway when it comes to policymaking and disaster response, while internally holding so much of this potential for uncertainty it then comes upon the community of of GIScientists to investigate the validity of these models. This is where the reproducibility and replicability of social vulnerability studies is of utmost importance. It is the responsibility of all GIScientists to ensure that others can reproduce their results and verify their findings to enhance the scientific validity of all geospatial research. This, however, has implications beyond just the reproducibility of the scientific community. GIScientists moreso owe it to the people in the vulnerable areas being assessed to ensure reproducibility so that it can be properly tested for validity, and peer-reviewed to best inform the decision making of politicians.

This discussion is at the forefront of current relevance in the Geography community where the ability to both understand and conduct reproducible research literally has the power to save lives. The teaching and development of reproduction studies of social vulnerability models such as our reproduction of Malcomb et al. 2014 is a critical step in spreading awareness about the power of reproducibility and well structured research compendiums.