An IDF Curve is a tool that characterizes an area’s rainfall pattern. By analyzing past rainfall events, statistics about rainfall re-occurrence can be determined for various standard return periods, for example, the size rainfall event that statistically occurs every 10 years. Typically 2-, 5-, 10-, 25-, 50- and 100-year return periods are shown on IDF curves. These curves are used in design with the assumption that past rainfall statistics continue to represent rainfall statistics into the future.
IDF stands for Intensity-Duration-Frequency. Intensity, duration and frequency are the parameters that make up the axes of the graph of IDF curves. The IDF curve is created with rainfall records collected at a specific monitoring location. For example, the Victoria International Airport IDF curve shown below was created with records gathered at a rainfall monitoring station located at the Airport, operated by Environment Canada (Meteorological Services of Canada, a Division of Envirinment Canada, formerly known as Atmospheric Environment Service) .
Rainfall intensity in the IDF Curve is the average rainfall depth that falls per time increment. Simplified, high rainfall intensity indicates that it’s raining hard and low intensity that it’s raining lightly. Typically the rainfall intensity is stated in mm/hr in Canada and in inches/hour in the United States (Chow, Maidment and Mays 455).
Lines on the IDF Curve graph represent probability, for example, the 10-year line would represent rainfall events that have a probability of occurring once every 10 years. Another way to put it is that the probability of a 10-year magnitude storm (or greater) occurring in any given year is 1/10 or 10%, and of a 50-year storm occurring 1/50 or 2%. Remember that the information presented in the graph is based on statistical analysis, not a prediction of actual storms (Bedient and Huber 385).
Each plotted line in the graph represents rainfall events with the same probability of occurrence, in a range of durations. A 10-year storm can therefore be of any duration, for example a 10-year 30-min storm, a 10-year 2-hour storm or a 10-year 12-hour storm. So, there can be many 10-year storms, and the rainfall intensity is higher for the short storms and lower for long duration storms.
We also know from observation that during long storms there are periods with light rain and periods with heavy rain. The average rainfall intensity for the storm is less than the intensity of the storm when it’s raining the hardest. For shorter storms it is more likely that the average rainfall intensity is similar to the heaviest rain period of the storm. This is why the Intensity (y-axis of the graph) decreases as the storm gets longer.
Typically statistical storms are reported for standard durations between 5 minutes and 24 hours (http://www.criacc.qc.ca/climat/idf_e.html). In regions with really long strong storms (e.g. the west coast of Canada), statistical storms are sometimes also reported in longer durations: 48 hours (2 days) and 5 days.
IDF curves are most often used for design. Municipalities and other approval agencies typically set out standards for design of infrastructure that includes minimum capacity in terms of rainfall return periods. For example, in Canada storm sewers are typical designed to carry a minimum of the 5-year storm. Simply put, this means that all the runoff from a 5-year storm from the area upstream of the sewer system must fit in the sewer without overflowing onto the road.
For a specific location and set of site specific characteristics, a particular storm duration will produce the greatest rain effect, usually the highest peak runoff flow or greatest rainfall volume. Circumstances can include whether an area is forest or urban or whether the location is in the mountains or in the middle of a prairie. Testing the various storm durations will determine which statistical storm will produce the greatest effect (governing storm duration). Knowing which storm duration is governing is important when designing storm water management facilities or estimating flood elevations in order to make sure the worst case is being used for design.
IDF curves are created by analyzing years of rainfall records. The longer and more complete the record, the better the quality of the statistical analysis. Long records of rainfall data are also less likely to represent a short-term rainfall anomaly, for example, a decade of high precipitation that is not representative of the long-term rainfall pattern of the region. The analysis of the rainfall record begins with Cumulative Frequency Analysis. During this process the interval of events with a certain magnitude are determined for a range of metrics within the rainfall record, e.g. the 15-minute storm with the greatest quantity of rainfall that occurs at a 10-year interval. A cumulative frequency distribution is typically used for this analysis. There are approximately a dozen methods for carrying out this analysis (Chow, Maidment and Mays 363). The most common distribution methods correspond to the methods used by the government agencies that typically provide IDF curves. Environment Canada uses the Gumbel - Method of Moments to create their curves, which are widely used in Canada.
Frequency Analysis can be carried out using statistically software, which is often available as a free download, but can be cumbersome as many of the programs were designed many years ago. On the Useful Links page of this site, you can find links to free CFA software.
IDF curves are available for many locations in Canada, produced by the Meteorological Service of Canada Meteorological Service of Canada, a division of Environment Canada, formerly Atmospheric Environment Service. They are updated periodically, many as recently as 2005, and are available for free download from their FTP Site . A different system of illustrating rainfall statistics exists in the United States. Areas of same statistical rainfall depths are mapped for specific return periods and storm durations. There iso-hyetal maps are called TP40s (Technical Paper No. 40) (Tecnical Paper No. 40) and can be found for free download on the US National Weather Service website . These links are great starting points for understanding rainfall behaviour in different areas of the US or Canada. Local or specialized agencies may have their own rainfall statistics that more closely emulate local conditions and experience, or particular uses for the rainfall information. Examples of these agencies include provincial (or state) transportation (or highways) departments, municipalities, environmental protection agencies, etc.