Jake Johnson 11/14/2018
One prediction of climate change is that the intensity and rainfall rates of cyclones will increase[1]. For example, in the case of hurricane Harvey consider the following:
… record high ocean heat values not only increased the fuel available to sustain and intensify Harvey but also increased its flooding rains on land. Harvey could not have produced so much rain without human-induced climate change.[2]
Similar flooding was observed following other recent hurricanes such as Florence, Maria, and Michael.
Also expected to increase are the damages and costs of such storms. These devastating effects are illustrated starkly via the use of satellite imagery both before and after they occur. For example, this Washington Post graphic shows the destruction brought by Super Typhoon Yutu. These pictures are indeed worth a thousand words, but they are also documenting potentially many thousands of dollars in damages.
Recently, efforts have been made to use satellite imagery as means for automated damage detection in storm-effected areas[3][4]. I propose using and building on these existing methodologies to assess storm-damaged areas and linking this information to estimates of monetary damages. This would provide a novel approach for quantifying the costs of disasters and perhaps predicting the costs of future events. Not only would this be useful in terms of disaster response, it would also serve as important information for any cost-benefit analysis of policy interventions seeking to address climate change.
-
- Code Base for analysis in [4]
- Indicative of the scope of this proposal in terms of data size
- NOAA GeoTIFF imagery ~60GB
- DigitalGlobe GeoTIFF imagery ~3TB
-
Trenberth, K. E., Cheng, L., Jacobs, P., Zhang, Y., & Fasullo, J. (2018). Hurricane Harvey Links to Ocean Heat Content and Climate Change Adaptation. Earth’s Future, 6(5), 730–744. http://doi.org/10.1029/2018EF000825
-
Barnes, C. F., Fritz, H., & Yoo, J. (2007). Hurricane Disaster Assessments With Image-Driven Data Mining in High-Resolution Satellite Imagery. IEEE Transactions on Geoscience and Remote Sensing, 45(6), 1631–1640. http://doi.org/10.1109/TGRS.2007.890808
-
Cao, Q. D., & Choe, Y. (2018). Deep Learning Based Damage Detection on Post-Hurricane Satellite Imagery. arXiv. https://arxiv.org/abs/1807.01688