While flood modeling and flood mapping allow for long-term planning, flood forecasting enables evacuation of impacted areas ahead of a flood event. Flash floods are particularly difficult to predict because they aren’t always caused solely by meteorological phenomena. In addition, while floods in larger river systems can be predicted up to a few days in advance, rivers with “flashier” hydrographs might provide little lead time for a warning. 

Factors that impact flood forecasting include [53]:

• how quickly the storm is moving
• rainfall intensity
• recurrence of storm systems in one area
• soil type and quality
• land surface type (i.e., how much impervious surface there is in a given area)
• vegetation type and quality
• topography (especially terrain slope)
• impacts of recent wildfire
• depth and condition of snowpack
• any existing infrastructure (like bridges, levees, or culverts)

Data regarding precipitation, expected precipitation, current river stage and flow are sent to river flood forecast centers, where computer models are used to determine the speed and extent of potential flooding. Recent advances in both data acquisition and modeling techniques have improved forecast rates and accuracies, though research is ongoing. New models are calibrated and validated using flash flood observational data from USGS automated discharge measurements, trained spotter reports from the National Weather Service, and databases compiled by the National Severe Storms Laboratory (NSSL) [53]. 

In 2004, the NSSL launched the Multi-Radar Multi-Sensor (MRMS) system, a network that integrates data from radar networks, surface and ground observations, lightning detection systems, satellite data, and computer weather prediction models. The MRMS provides high-resolution, rapidly updating data on precipitation locations, amounts, and types. MRMS data is input into NSSL’s Flooded Locations and Simulated Hydrographs (FLASH) project. FLASH was launched in 2012 and is the world’s first continental-scale flash-flood forecast system; its hydrologic models are updated every ten minutes and encompass a grid of 11 million data points scattered across the country [54]. 

[53] NOAA National Severe Storms Laboratory, “Severe Weather 101: Flood Forecasting,” [Online]. Available: https://www.nssl.noaa.gov/education/svrwx101/floods/forecasting/. [Accessed June 6, 2022].

[54] NOAA National Severe Storms Laboratory, “FLASH: Improving the Science Behind Flash Flood Prediction,” [Online]. Available: https://inside.nssl.noaa.gov/flash/. [Accessed July 18, 2022].