City of Central showcases data-driven flood model, EarlyFlows forecast system and grant strategy

Mayor Wade Evans of the City of Central told the Amite River Basin Commission that his city has deployed a data-driven approach to stormwater management, including a living hydraulic model, lidar mapping, targeted maintenance and a public flood-forecasting tool called EarlyFlows.

“Once you did everything, the last step would be to implement a flood warning system where you could utilize the model and put in the forecasted rainfall from NOAA,” Mayor Wade Evans said, describing the city’s EarlyFlows system and its public interface.

Evans said Central used grant and ARPA funding to buy high-resolution lidar, surveying equipment and hydrology monitoring stations and to develop an integrated model and forecasting service. He told the commission the city has invested roughly $1 million over the past nine years to develop and refine its hydraulic model, and he listed other costs the city incurred: about $42,000 to commission new one-foot lidar, $85,000 for Leica survey equipment and consultant-based work to secure multiple grants for the same projects. Evans said the city had identified roughly $78 million in projects and had about $30 million in local money on hand when he took office; the city has also recently secured more than $60 million in grants, he said.

Stoker Brown of CSRS, who participated in the presentation, described how the forecasting system runs multiple forecast and hindcast model ensembles hourly, ingests MRMS (radar-based) rainfall estimates and measured gauge data, and produces a range of projected stream-level responses. “Forecasted rainfall comes out every hour. We run that forecasted rainfall. We also run a hindcast of what rain happened in the last hour,” Stoker Brown said. He said the system ties multiple forecast models and observed data together so both residents and emergency managers can see likely flood depth ranges at specific locations.

Evans said the city’s EarlyFlows system is live after a months-long testing period and that the city has calibrated it against measured gauge data. He said the public can log in, select a location and see a hydrograph and a range of forecasted peak depths; central officials use the same information for local response decisions. Evans said the EarlyFlows tool is expected to help the city reduce fear, improve preparedness and could help reduce flood-insurance rates: Central’s FEMA Community Rating System (CRS) class improved from a higher level to class 5 through local regulations and programs, and Evans said the city expects to reapply for class 4 after implementing the forecasting system (class 5 is roughly a 25% premium discount; class 4 is about 30%).

Evans described how the city has changed development regulations to account for conveyance zones, requires off-site drainage assessments for any proposed development and has pursued extensive community engagement. He said Central contracts much of its maintenance work, conducts planned, whole-neighborhood ditch and culvert repairs (rather than piecemeal fixes), and uses survey equipment to ensure proper fall in culverts before excavation. “We intentionally manage our stormwater system,” Evans said.

On technical scale and scalability, Stoker Brown told commissioners the approach is scalable but requires model domain tailoring: “One model is not going to be able to do a forecast for the entire state…we’ve downscaled detailed models into a very streamlined model that runs in 15 minutes” for city-scale forecasting, while larger HUC-scale or basin-scale models would run separately.

Evans also described pilot work on aquifer recharge and local wetland mitigation banks: the city is exploring using development and landscaping codes to improve infiltration to shallow sands and the Southern Hills aquifer; he said Central applied for related funding but was turned down and continues to refine modeling and geologic work. He asked the commission for letters of support for regional mitigation and recharge efforts where appropriate.

Commissioners asked technical questions about how the model ingests forecast versus observed rainfall and whether the forecasting system “learns” from errors. Stoker Brown explained that forecasts are run hourly from ensemble meteorological products, hindcasts use MRMS gauge-corrected radar rainfall to represent what already fell, and model outputs are compared to measured gauges for calibration after events.

Mayor Evans credited the city’s grant strategy and consultant team for enabling multiple funding sources to be combined for the same projects and said that approach allowed Central to move work forward without relying solely on local taxes. He closed by inviting commission members to review the city’s approach and consider scalability to other basins or regional modeling efforts.