Crowdsourcing Disaster Relief

When Hurricane Katrina made landfall in August, 2005, few imagined it would affect 92,000 sqmi, nearly destroy a major US city, take the lives of 1,800 people, and leave over 400,000 people homeless. Many reports into the response of Katrina indicate that there was a lack of situational awareness at every level, up to the incident commander, that hindered command and control within the Emergency Operations Centers (EOC) of the affected counties. While several mistakes were made during the most critical hours of the storm, lessons were learned that would make emergency response of future storms more efficient.

In the initial hours of a large-scale disaster, first responders scramble to establish command and communications within an EOC. One of the primary objectives for the EOC is to gain a sight picture for situational awareness about the size and scale of the disaster, was well as the extent of the damage. Depending on the size of the disaster, this can often take days to weeks to obtain. This information determines how the incident commander deploys his assets. The information is also used in requests for federal funding through FEMA regional offices. The earlier an incident commander can receive accurate information, the faster help can be sent to those in need.

“The most critical and coveted, yet situationally difficult category of information to capture immediately following a disaster is accurate and timely intelligence about the scope, extent, and impact of the event.”

FEMA 9321.1, 1999


Fig.1- Crowdrank output from a disaster relief discovery campaign

To gain this situational awareness early on, FEMA currently funnels ground reports of damage and injuries to their GIS team in the EOC. GIS technicians begin plotting damage as soon as they set up operations. Additionally, aerial imagery and surveys may be ordered to survey the damage, as well as satellite imagery when available. But there's a problem with the way FEMA currently gathers geospatial data following the storm; too few personnel to analyze the many sources of data that flood the EOC. Without adequate personnel to help plot the damage, and relay those maps to strategic-level response elements, precious time is wasted.

This is where it makes perfect sense to introduce crowdsourced satellite imagery products into the National Response Framework, well before the next major disaster. Our GeoHIVE Community has the ability to tag features in up to 100,000 sqkm of high-resolution satellite imagery each day. DigitalGlobe has two teams that work together to crowdsource imagery for first response agencies. Our FirstLook team monitors disaster events around the globe, obtaining pre and post-event imagery from our constellation as soon as it is practical. GeoHIVE can use that imagery to generate a Discovery campaign, where users within the GeoHIVE Community search for features of interest, such as: flooded roads, damaged buildings, heaps of debris, and damaged vehicles.

When major disasters strike, our GeoHIVE Community is standing by, ready to quickly map out the damage.