Building Data

Building Data

A custom dataset has been developed by Technosylva to support the impact analysis and risk modeling within FireRisk and FireSim. This dataset has been prepared using two different sources:

1. Microsoft building source

2. Others local buildings sources existing in some counties or cities

This data was developed for the entire State of California.

A description about how the custom Technosylva enhanced building dataset was prepared is provided below.

Microsoft Building Data Source

The Microsoft buildings data source includes building footprints for the State of California (10,988,525 buildings). This data is publicly available having been published by Microsoft. No attributes exist about the buildings, it is simply a polygon footprint.

The Microsoft buildings information has some area where buildings data does not exist, although imagery shows buildings (i.e. holes in the data). To improve this source data, Technosylva supplemented the buildings data with other sources. 

Integrating Other Local Building Data Sources

Additional building data was obtained from a variety of local data sources, such as counties, cities, etc.

Description of building loss factor/estimated buildings destroyed:

Building Loss Factor (BLF) is a new individual building attribute that Technosylva developed by conducting analysis on 13-years of DINS data. It represents a probability of severe damage or destruction for that building. It is derived by analyzing landscape characteristics, such as aspect, slope, landform, surrounding fuels and building density to identify common characteristics that reflect DINS data. We used these common characteristics to define a loss factor for each building that can be applied during a moderate-high-extreme weather event.

BLF is primarily intended to be used for fire spread predictions (simulations) to allow us to provide an estimate of the buildings destroyed from the total amount of buildings threatened (impacted). We do this by averaging the BLF for all the buildings impacted. For example, if 100 buildings are threatened for a simulation, we take all the individual BLF values (one per building) and average them. The total is the estimated number of buildings destroyed.