Assessing flood damage using satellite imagery
Avatar photo Nash Prado

Assessing flood damage using satellite imagery

Artificial Intelligence | Infrastructure | Monitoring

Satellite imagery is used to assess flood damage immediately after a flood before the lives of emergency response crews are put at risk. They can tell disaster response teams what roads to avoid, what houses are in the most distress, and other dangers that may be present. In this article, we explore the ways that Proimagery uses satellite imagery to augment existing disaster response procedures.

Tropical Storm Imelda – September 2019

On September 18, 2019, Tropical Storm Imelda made landfall over Texas, soaking some areas with more than 10 inches of rain and bringing the threat of flash floods across southeast Texas. In just a few short days, Imelda left an estimated $5 billion dollars of damage in its wake.

Processed Synthetic Aperture Radar (SAR) imagery by ProImagery shows the extent of flooding over the state of Texas one day after the onset of Tropical Storm Imelda

Through satellite imagery, the extent of flooding can be calculated and the most hard-hit areas can be accurately pinpointed. This information is critical to government, city planners, disaster response teams, and insurance companies to accurately understanding flood risk at the municipal, state, and national levels.

Using the Normalized Difference Water Index on high-resolution satellite images taken over Plum Grove on September 20, 2019, the extent of flooding can be observed. Synthetic Aperture Radar Imagery captured September 19, 2020, shows clear indications of flooding.

previous arrow
next arrow

Extracted flood extent from high-resolution satellite imagery over Plum Grove, Texas

Routes are often blocked by water or flood damage. Processed satellite imagery is used to detect which roads are affected by flooding. Processed flood extent information from Synthetic Aperture Radar and Optical satellite imagery can be overlaid on a pre-flood basemap and road network map to determine which roads are most affected by the flood. This critical information helps disaster response teams plan their travel routes and prioritize the most affected areas.

Through the use of machine learning, building footprints can also be detected from very high-resolution satellite images. Building footprints are overlaid and processed together with flood extent maps to determine which houses and roads were affected by flooding.


previous arrow
next arrow

The images show which properties were affected (houses and Roads) by the flood extent over Plum Grove, Texas



The use of satellite imagery to augment disaster response teams has been a relatively recent innovation due to improved image qualities, falling imagery prices, and improved processing algorithms. Areas at risk of flooding should be monitored in advance of any tropical storm or hurricane. If you would like to know more about how Proimagery can help disaster response planning, please feel free to reach out.