Flood Forecasting TechnologyAECbytes Feature (September 14, 2017)

All the way away back in 2005, in the wake of Hurricane Katrina and the horrific devastation it brought to the city of New Orleans, I wrote the article “Hurricanes and their Aftermath: How Can Technology Help?” to explore how technology was being used at that time in the relief, rehabilitation, and reconstruction efforts along the Gulf Coast, as well as how it could play a play a more critical role in the future to prevent cities and their inhabitants from suffering the same fate New Orleans did.

Here we are, 12 years later, reeling from two back-to-back hurricanes, Harvey and Irma, which have battered Texas, Florida, Cuba, the Caribbean Islands, and several other coastal regions of the US. While the loss of life this time has thankfully been a lot less compared to that caused by Hurricane Katrina, the damage to buildings and infrastructure has been just as devastating. According to a preliminary estimate from Moody's Analytics, the total damage to Texas and Florida alone caused by Hurricanes Harvey and Irma is between $150 billion and $200 billion, which is comparable to the costs from Hurricane Katrina in New Orleans in 2005.

This brings us back to the same question I asked in 2005: how can technology help? After all, given the level of technological advancements in almost every aspect of our lives, it seems like a no-brainer that we should be apply our technological knowhow and prowess to also reducing the impact of natural disasters like hurricanes on life and property. While we can certainly shore up our efforts to design better hurricane-resistant structures—and this is where AEC technology can help with improved analysis tools that structural engineers could use—improving our ability to forecast areas of flooding before a hurricane actually hits can lead to better preparation and improved evacuation.   

So what does flood forecasting actually entail from a technological standpoint? What is the current state of the art? And where does AEC technology come in, if at all?

I set about to explore these issues, and my findings are captured in this AECbytes article.

Current Forecasting Technology

In order to be able to predict areas of flooding caused by hurricanes, we need to be able to predict hurricanes in the first place, and this is something that is already well in place, thanks to advances in meteorological science (Figure 1). Most of us are familiar with the weather maps used by meteorologists to capture current conditions and provide forecasts of rain, storms, tornadoes, and hurricanes. The science behind meteorology is constantly advancing and weather patterns can now be predicted several days in advance. While there is still some measure of unpredictability—such as the exact path of a hurricane, whether it will gain or lose intensity, where it will make landfall, and so on—the overall predictions are still overwhelmingly accurate.

Figure 1. The 2016 North Atlantic hurricane tracking chart. (Image courtesy: weathertiger.com.)

While the prediction of weather events is a well-established science that is rapidly progressing, the prediction of the flooding caused by these weather elements on the ground is relatively not yet as advanced. That is not for lack of trying, however. Given the extensive damage caused by flooding to both life and property, efforts to find ways to minimize the loss through improved forecasting are underway at several government organizations in the US such as the National Oceanic and Atmospheric Administration, U.S. Geological Survey, American Geosciences Institute, and NASA. By and large, however, most of these forecasts do not take into account aspects such as the terrain of the land, whether it is inhabited and if so to what extent, the location of buildings and infrastructure, and other physical characteristics. That is why most of the forecasting done by these organizations is displayed in the form of 2D maps, such as WaterWatch, which shows flood and high flow conditions for the US (Figure 2); the Significant River Flood Outlook map (Figure 3); and detailed flooding forecasts for different regions of the US (Figure 4). All these maps are interactive and regularly updated to show the most current flood conditions and forecasts.

Figure 2. The WatchWatch interactive map by the U.S. Geological Survey, which is regularly updated to show flood and high flow conditions for the U.S.

Figure 3. The Significant River Flood Outlook interactive map by the National Oceanic and Atmospheric Administration, showing the flood likelihood for different states.

Figure 4. Detailed flood forecasting map for the south-eastern region of the US published by the National Weather Service, an agency within the National Oceanic and Atmospheric Administration.

Flood forecasting that actually takes into account the conditions on the ground—topography, development, population, and so on—is even less common. The few examples I could find include an initiative called CI-FLOW (which stands “Coastal and Inland Flooding Observation and Warning”), a demonstration project that predicts the combined effects of coastal and inland floods for coastal North Carolina taking into account GIS data including topography, soil type, streams and rivers, slope of the land and vegetation patterns (Figure 5); GFMS (Global Flood Monitoring System), a NASA-funded experimental system being developed at the University of Maryland which maps flood conditions worldwide using precipitation data from NASA satellites in conjunction with land surface model data (Figure 6); and an older Floodpath project by the U.S. Geological Survey, which created flood forecast maps using weather data, GIS, and LIDAR (Figure 7).

Figure 5. CI-FLOW, developed by the National Severe Storms Laboratory (NSSL), predicted water levels above sea level at the time of Hurricane Irene’s landfall in August 2011.

Figure 6. The GFMS online flood monitoring system funded by NASA.

Figure 7. The test implementation of the USGS’s Floodpath flood forecast mapping project for the Snoqualmie River Basin in Washington in 2009.

In addition to government agencies which do the bulk of the work related to flood forecasting, there is some work related to floods being done by private firms as well. However, this is more of an “after the fact” event in the form of “catastrophe modeling,” which is used to estimate the losses that could be sustained due to a catastrophic event such as a hurricane or earthquake, primarily for the insurance industry. See, for example, the images from a report prepared by the firm, KCC, estimating insurance losses from Hurricane Matthew last October (Figure 8). The top image shows the flooding footprint for North Carolina using high resolution elevation and stream gauge data, highlighting the hardest hit areas, while the lower image shows the ZIP codes that were estimated by the firm to incur over $25 million in flooding damages from the hurricane. KCC estimated that insurers would pay a total of over $7 billion for damages in the US combined as a result of Hurricane Matthew

Figure 8. An extract from a report estimating the insurance losses in North Carolina from Hurricane Matthew last year from the catastrophe modeling firm, KCC.

Potential Technology Solutions from AEC

Solutions like the ones discussed above, even the few that take GIS data into account in their flood forecasting technology, do not incorporate data related to buildings, infrastructure, and population. This is where AEC technology comes in. In my 2005 article, I had expressed this hope:

In time, we could also find a smart way of integrating the BIM models of individual buildings within the city's CIM [city information model], so that we have a highly accurate and detailed digital replica of a city which can be subjected to sophisticated analysis and simulations. We could then predict the impact of a hurricane, earthquake, tsunami, gas leak, bioterrorist hazard, or any other kind of conceivable disaster not only on the city as a whole but on individual buildings and neighborhoods within the city as well.

We are not quite there in 2017, but we are getting closer. CIM or “city information modeling” is now an upcoming technology, a good example of which is Cityzenith Smart World, reviewed in AECbytes earlier this summer. While Smart World does not yet have the analysis capabilities, as I pointed out in my review, to predict “the areas that would be impacted in the event of a disaster such as an earthquake or hurricane and make the necessary plans to mitigate it or determine how emergency and evacuation services would be provided,” it is coming out with a new version next year that has an underlying math and geometry model with individually selectable meshes that make this type of simulation work possible. Users could layer any contextual terrain, buildings, infrastructure, population, services, and additional data (static and real-time) into the model and then leverage the computational layer and underlying math architecture to run any required simulation.

Turning to established AEC technology vendors, leading firms such as Autodesk and Bentley which also develop infrastructure design solutions in addition to BIM solutions (see the article, Extending BIM to Infrastructure), have some tools that can be used for flood analysis, not at the level of a city but at a smaller scale. For example, Autodesk’s infrastructure design solution, InfraWorks, supports flood simulation through a third-part plug-in called RiverFlow2D developed by Hydronia, LLC (Figure 9).

Figure 9. The animation of a flood simulation in Autodesk InfraWorks created using the RiverFlow2D plug-in.

Bentley, on the other hand, has a large number of water-related applications in its vast infrastructure portfolio including FlowMaster, StormCAD, CivilStorm, WaterGEMS, and many more, intended to be used by engineers for land development projects, design of levees, and optimizing the operations of water distribution, wastewater, and stormwater systems. These applications can also be used for water-related simulation, as in the case of the SimTejo organization in Lisbon, Portugal, which used Bentley’s SewerGEMS product and WaterObjects.NET technology to accurately predict sewer overflows in the city’s water network (Figure 10).

Figure 10. Predicting sewer overflows in Lisbon’s water network by SimTejo using Bentley’s hydraulics and hydrology software.

I was pleasantly surprised to find that even Vectorworks, a smaller AEC technology vendor that does not have an infrastructure design portfolio as such, has examples of its solutions being used for flood-related work. In addition to Vectorworks ARCHITECT for BIM, the company also develops LANDMARK for landscape design, and one of its users, the Australian landscape architecture and urban design firm McGregor Coxall, has used these products for an urban design project that required modeling and analysis of the flood zone of a river running through it (Figure 11). Another example is an urban design proposal by the firm, LEVENBETTS, in the wetlands region of New Orleans taking into account the flooding caused by Hurricane Katrina, created using Vectorworks (Figure 12). And finally, the plug-in for pedestrian simulation in Vectorworks, SimTread, which was originally developed to understand crowd behavior during building evacuations has been expanded by its Japanese developer to simulate whole city evacuations during natural disasters like tsunamis and earthquakes such as the large one that struck Japan in 2011 (Figure 13).

Figure 11. The Parramatta River Vision Plan project by the Australian landscape architecture and urban design firm McGregor Coxall, which used Vectoworks for water simulation in addition to the conceptual design. (Image courtesy: McGregor Coxall)

Figure 12. Wetland City, an urban design proposal by LEVENBETTS in New Orleans.

Figure 13. A SimTread software simulation of a city’s evacuation time.


In contrast to Hurricane Katrina 12 years ago which resulted in hundreds of deaths, Hurricanes Harvey and Irma have had a smaller death toll, which is a promising indication that flood forecasting and evacuation measures have improved. As shown in this article, technological enhancements have definitely been made, not just in predicting hurricanes in the area of meteorological science, but also in actual flood forecasting using GIS and other ground data. The next step is to also include building and infrastructure data in flood forecasting, which is where AEC technology comes in. All the individual pieces of the technology are there, they just need to be brought together in an effective way. In this regard, CIM seems a really promising technology, although BIM applications from vendors like Autodesk, Bentley, and Vectorworks are also contributing to the effort.

Of course, any kind of flood forecasting improvements must be done in conjunction with actual on-the-ground solutions such as overhauling infrastructure, designing flood-resistant structures, strengthening building codes, and improved zoning that avoids development in low-lying areas prone to flooding.

It is also important to keep in mind that disasters like floods are not just an issue in the US, but all over the world—in fact, thousands of people die each year from floods in developing countries. The US has always been at the forefront of technological innovation, and hopefully it will do the same with flood forecasting technology as well.

About the Author

Lachmi Khemlani is founder and editor of AECbytes. She has a Ph.D. in Architecture from UC Berkeley, specializing in intelligent building modeling, and consults and writes on AEC technology. She can be reached at lachmi@aecbytes.com.

AECbytes content should not be reproduced on any other website, blog, print publication, or newsletter without permission.

Related Archive Articles

  • Hurricanes and their Aftermath: How Can Technology Help?
  • In the wake of Hurricane Katrina in 2005, this article explores how technology is being used in the relief, rehabilitation, and reconstruction efforts along the Gulf Coast, and what technologies we might need in the future for cities prone to natural disasters to combat them more effectively.
  • Extending BIM to Infrastructure
  • Now that BIM has conquered the building industry and is here to stay, this article explores the application of the same "information modeling" concept to infrastructure.
  • Bentley's Year in Infrastructure 2014 Conference
  • Details on Bentley's CONNECT generation of software and a guest keynote on the new High Speed Rail project proposed for London on which the use of BIM is being seen as the "lifeblood" of the project.
  • Around the World with BIM
  • This article attempts to capture the key developments in BIM implementation in different countries around the world, so that we have a better global perspective on BIM.