GSA Annual Meeting in Indianapolis, Indiana, USA - 2018

Paper No. 171-12
Presentation Time: 9:00 AM-6:30 PM

EVENT-SPECIFIC AND ANNUAL PRECIPITATION CONTROLS ON MASS WASTING SITES IN PUERTO RICO AFTER THE PASSAGE OF HURRICANE MARIA


BAYOUTH GARCIA, Desiree1, RODRÍGUEZ FELICIANO, César A.1 and HUGHES, K. Stephen2, (1)UPRM Department of Geology, University of Puerto Rico Mayaguez, Call Box 9000, University of Puerto Rico Mayaguez, Mayaguez, PR 00681, (2)Department of Geology, University of Puerto Rico at Mayagüez, Call Box 9000, Mayagüez, PR 00681-9000

In 2017, Hurricanes Irma (Sep. 6) and María (Sep. 20) brought intense rainfall to Puerto Rico. The passing of these hurricanes left the island devastated, causing flooding, an island wide blackout - where some communities were left without power for several months -, bridge and road collapses, causing several communities to be isolated for long periods of time. Aerial and satellite imagery from FEMA, NOAA, and DigitalGlobe were used in ArcMap in order to catalog a landslide inventory after Hurricane María, in which more than 40,000 landslides were identified. During Hurricane María, the WSR 88D NEXRAD Doppler Radar experienced failure, leading to inconsistency in precipitation estimate data products. Nevertheless, precipitation raster estimates for Hurricane María (from NASA Global Precipitation Measurement, National Weather Service Quantitative Precipitation Estimate, and the National Hurricane Center), Hurricane Irma (from National Weather Service Quantitative Precipitation Estimate), and Mean Annual Precipitation Data (from the WorldClim 2.0 dataset) were used to explore any correlation between the new inventory and precipitation estimates using the frequency ratio method.

Hurricane Irma’s precipitation estimate shows an overall positive correlation pattern in ratios, possibly indicating that the passing of Hurricane Irma saturated soils in the island, before the landslides being ultimately triggered during Hurricane María. Hurricane María’s precipitation estimates shows a positive correlation in a range of precipitation values in obtained ratios, but the results are not clearly interpretable, likely due to the failure of the Doppler Radar system, that is an essential tool to calculate local rainfall amounts. The Mean Annual Precipitation data shows a consistently positive correlation when comparing rainfall and landslide occurrence. This difference in results led to the conclusion that precipitation was sufficient enough during Hurricane María to trigger landslides across all of Puerto Rico, but the chemical weathering pattern controlled by Mean Annual Precipitation was one of the primary limiting landslide inducing factors. Results from this study will contribute to a much needed update of the official Landslide Susceptibility Map of Puerto Rico.