Rainfall Basics: DDF Curves and Design Storms

It’s August in Florida. An afternoon storm rolls in as you make your way home from work. The clouds darken, the sea breeze picks up, and the skies open to release a heavy downpour. When you arrive the next morning, your coworkers are talking about the rain. How would you describe the storm?

---------------------------------------------------------------------------------------------------------

Characterizing storms is not just interesting small talk. Rigorous characterization of storms over the long-term results in design storms that guide key choices involved in increasing resilience to flooding. 

As planners and engineers design cities and infrastructure, an important consideration is how much precipitation might fall over a project area. Depth-duration-frequency curves, or DDF curves, use statistics to characterize observed rainfall and create design storms that are incorporated into standards for resilient development.  

Much like how you describe a summer storm to your coworker, DDF curves characterize precipitation events by answering three questions: how much rain will fall given how long a storm lasts, and how frequently is a storm with those characteristics expected to occur?  

How much? 

"That was a pretty heavy downpour — there were ten inches of water in my rain gauge.” 

Depth, or the total amount of precipitation accumulated, is depicted on the y-axis of a DDF graph. Depth is usually specified in millimeters or inches. 

How long? 

"It rained for over two hours straight!” 

Duration refers to the length of the precipitation event — how long the storm lasted. This variable is depicted on the x-axis of a DDF graph.  

How rare? 

“We haven’t had this much rain since the storm that flooded First Avenue ten years ago.” 

Frequency refers to how often a storm is expected to occur – its rarity. Frequency is expressed as a return period, or average recurrence interval, which describes the average amount of time between events of a certain intensity.  

Return periods are calculated by ranking historical data on the amount of rain delivered by storms, then calculating the likelihood of events that deliver as much or more rain. This is known as the annual exceedance probability, or AEP. The inverse of this exceedance probability is the return period. For example, a storm with a 10 percent chance of occurring in any given year, a 10 percent annual exceedance probability, will have a return period of 10 years. Notably, more intense storms occur less frequently and have longer return periods.  

Design storms 

Depth-duration-frequency curves depict anticipated accumulation of rain at a given location for storms of varying duration and frequency. Data from DDF curves are used to create “design storms” — hypothetical precipitation events that are used as standards to which infrastructure projects can be built. 

Hydraulic infrastructure, such as stormwater systems, structures to prevent flooding, and other systems to control the flow of water, are built to cope with certain design storms. Design storms support an evaluation of tradeoffs between protection from floods and costs grounded in historical data.  

Selecting an appropriate design storm depends on assessing risk. Infrastructure to protect critical, expensive, and long-lasting projects will be built to higher standards designed to cope with rarer and more powerful storms.  

For example, storms with 2-, 5-, and 10-year return intervals are best used when the consequences of flooding are minimal, such as a public park or a parking lot. Assessments of risk for residential areas and major roadways often consider 25- and 50-year storms, respectively, while 100-year storms are the industry standard for mapping riverine floodplains and designing infrastructure to protect key assets. When planning protection for critical and high-risk infrastructure, like nuclear power plants, engineers might consider 200- or 500-year storms. 

Integrating future changes in rainfall into design of infrastructure 

Return periods are calculated using historical data, emphasizing the importance of accurate records of rainfall to support proper design of infrastructure. However, research has shown that the depth, duration, and frequency of rainfall events are changing, meaning historical data may not depict future rainfall accurately.  

Researchers in the Flood Hub’s Rainfall Workgroup have developed change factors that allow practitioners across Florida to account for these anticipated changes in rainfall. Change factors are developed from outputs of global climate models through statistical downscaling, and they act as multipliers when applied to DDF curves and design storms.  

To access the Flood Hub’s change factor dataset, visit our Downloads page.  

Current DDF curves for 242 stations around Florida are provided by the National Oceanographic and Atmospheric Administration (NOAA) through the Atlas 14 program.