Forecasting Inundation and pre-event Mitigation: Can we expand our IDSS toolbox post Harvey? (145)

Lance Wood, NWS Houston/Galveston, Dickinson, TX

Katie Landry-Guyton, NWS Houston/Galveston

Derek Giardino, West Gulf RFC

Fernando Salas, Lyncos Technologies



The greatest forecasted rainfall amounts for a landfalling U.S. tropical cyclone occurred ahead of the record rainfall produced by Hurricane Harvey. Prior to landfall along the coastal bend of Texas, the Weather Prediction Center (WPC), in collaboration with the NWS Houston forecast office, predicted up to 40 inches of rainfall across southeast Texas. These amounts, when compared to flash flood guidance, yielded the forecast of an extreme threat of flash flooding across portions of southeast Texas for all three forecast periods (Day 1-3), including a first ever Day 3 forecast of an extreme flash flood threat in the Excessive Rainfall Outlook.

While providing Harvey briefings to decision makers, many questions arose concerning the expected impacts from the extreme flash flood threat. For example: How many houses are likely to be flooded in Houston? How high will water rise in neighborhoods adjacent to creeks and bayous? Many of my friends and family asked me if their home would flood, and if they should evacuate. Although we knew structure flooding would likely occur prior to the August 27, 2017 flash flood emergency, we did not have reliable high resolution inundation forecasts to aid with answering these questions.

This study utilizes hourly precipitation forecasts from the High Resolution Rapid Refresh (HRRR) model from August 26, 2017, to produce hydrographs for points across the Houston metropolitan basin ahead of the devastating flash flood event. From these hydrographs, inundation and flood extent maps are produced utilizing a variety of techniques, including but not limited to the Height Above Nearest Drainage (HAND) method from the National Water Center. These maps are then evaluated for accuracy, with the best available presented to area decision makers. Our main goal is to assess whether the state of the science (modeling both rainfall intensity and location, as well as the resulting inundation) is sufficient to attempt resource staging and targeted evacuations ahead of significant inland rainfall threats.