Delos Team Highlight: Spatial Informatics Group

Delos is fortunate to have some of the world’s most outstanding talent on our team, and we’re excited to highlight them in a series of posts. Today, we’re highlighting not just one individual, but a whole team of incredible scientists: Spatial Informatics Group.

Spatial Informatics Group (SIG) is an interdisciplinary think tank specializing in complex environmental geospatial problems. Their team includes dozens of PhD level academic researchers, data scientists, and geospatial analysts. These analysts are also specialists in fields such as landscape ecology and forestry, and are experts in using tools such as complex statistical analysis and data science, machine learning, Geographic Information Systems (GIS), remote sensing, spatial data mining, and dynamic spatial modeling.

Delos is partnered with SIG and works closely with their principal scientists to create our disaster models. These principal scientists are truly world-beating in their fields– they’re university professors who are also founders and directors of university research labs. They’ve published in the most prestigious academic journals and have have authored a long list of research papers that are frequently cited by the news media. They’ve been featured at UN Climate Change summits and conferences, and created models as a joint venture with NASA.

Besides such impressive academic achievements, these scientists also have a proven history of applying their research to industry. SIG has created models of wildfire and hurricane disaster risks for state, national, and international government agencies, NGO’s, private companies, and utilities. Among their many relevant projects is the creation of wildfire risk maps for Cal Fire and the California Public Utilities Commission.

Partnering with SIG is an enormous privilege, and their extensive experience has allowed Delos to hit the ground running in creating our models. The rest of the Delos team has been floored by the brilliant work that they’ve done and continue to do, and we feel incredibly lucky to have them on our team!


Understanding last year’s wildfires: Kevin’s talk at the Insuretech Fusion conference

Even though the devastating and tragic wildfires of last year are no longer making news headlines, the insurance industry is continuing to pay close attention to how the disaster unfolded, and what it means for future wildfire seasons. Delos CEO Kevin Stein was invited to speak about the California wildfires at the InsurTech FUSION conference in Silicon Valley, put on by the Silicon Valley Insurance Accelerator. We often get questions about the wildfires of late 2017, so we were eager to share what we’ve learned, and how we’re building technology that we believe will make a difference in the impact from future wildfire disasters.

You can watch Kevin’s talk here: (The link jumps to the point in the video when Kevin starts his talk, which is approximately 9 minutes long.)

In his talk, he touches on the following points:

The wildfire season of 2017 led to record losses, particularly the Tubbs Fire in Northern California. Was this fire season an anomaly? Or does it represent a new normal? In order to answer this, we examined the following data:

Acres Burned & Buildings Burned graphs

  • The left chart illustrates the number of acres burned per year, which is increasing gradually over the last few decades, including this year.
  • We then looked at the data on the right, which illustrates the trend in the number of structures burned per year in CA, which occasionally spikes dramatically. But if the number of acres burned has only been gradual increasing over decades, why are we surprised by spikes in losses?

These two trends reveal a discrepancy in expectations. We then looked at where the Tubbs Fire occurred, as overlayed on the California State wildfire risk map, as shown here:

Napa Risk Map with Fire Overlay

Many of the hardest hit areas were thought to be in lower risk locations.

We concluded that this discrepancy in expectations is because of the modeling. This isn’t too surprising when considering that disaster modeling is a fairly nascent science, and that wildfire modeling is even newer than hurricane modeling.

Developing better wildfire risk models will have a very real impact on how communities experience wildfire losses. By knowing where to take proactive measures to harden properties against natural disasters, we can tremendously reduce the risk of loss. Delos is excited to be advancing modeling technology in order to make this kind of difference in communities’ resilience in the face of increasing wildfire dangers.