Speaker
Description
Keywords: asteroid, impact risk, modeling uncertainty
Determining the potential damage and risk from an asteroid impact involves many sources of uncertainty. Limited observational data leads to uncertainties in the asteroid properties, orbital uncertainties affect the potential impact location, and the entry and damage models all have some inherent modeling uncertainty. When each of these sources of uncertainty are combined, they can result in wide ranges for damage and risk estimates. As more information is gained about the object from additional observations or reconnaissance missions, uncertainties in asteroid properties and impact location are expected to reduce. A focus is therefore often placed on getting as much information as possible about a potential impactor. Modeling parameters, however, will remain uncertain. Therefore, from a risk perspective the question should be, what information refinements add meaningful value to making the best available decisions.
The compounding effects of uncertainty on hazard and risk calculations are highlighted by the range of results in the PDC25 impact exercise scenario. Even with help from JWST to constrain the size and type in the Epoch 1 assessment, asteroid property uncertainties lead to mass and energy ranges spanning multiple orders of magnitude. The Epoch 1 risk swath cuts through multiple regions including all of Africa, the Mediterranean, and Eastern Europe when accounting for orbital uncertainties. Together with modeling uncertainties, this leads to potential ground damage sizes ranging from zero or near zero up to 250 km and affected population estimates up to millions of people.
In this study, we use results from the Asteroid Threat Assessment Project (ATAP) Probabilistic Asteroid Impact Risk (PAIR) assessment to investigate which properties and parameters are most valuable to refine from a risk perspective for the PDC25 impact exercise scenario. We evaluate how different levels of precision in the inputs affect the model output precision along the swath, highlighting the variations in risk uncertainty reductions. Specifically, we will compare model output uncertainties assuming several different refinements (e.g. JWST, reconnaissance missions, etc.) and assess how our risk knowledge is affected by these additional sources of information. These results aim to start a conversation about what information makes the most difference from a risk perspective in determining when and how to act. Understanding the limits of what is most useful to evaluating potential damage and risk in this hypothetical scenario frees up additional resources to focus on factors that improve estimates in other areas, which may be vital in a real scenario where limited resources should be used most efficiently.