The Colorado-New Mexico Regional Extreme Precipitation Study (CO-NM REPS) was a multi-year effort to update extreme precipitation information for the Dam Safety offices of Colorado (http://water.state.co.us/damsafety/dams.asp)  and New Mexico (http://www.ose.state.nm.us/DS/dsInex.php). MetStat is proud to have been part of that study to generate updated extreme precipitation-frequency estimates in collaboration with MGS Engineering delivered through a web-based interface called MetPortal™The CO-NM REPS study documentation is available on the Colorado Division of Water Resources and Colorado Dam Safety web-site and through the following Google Drive link: https://drive.google.com/open?id=1BwS5VLnQ6_5N_NWjfgbLUPV42Ietex4n.

While final Rules for Colorado will be promulgated later in 2019, the precipitation-frequency estimates are slated to be used in dam safety as a means for examining the adequacy of spillways across the state at low- to high-hazard dams (Figure 1). This differs from previous dam safety rules which were based on fractions of probable maximum precipitation (PMP) and illustrate a shift in the hydrologic community towards more risk or probabilistic-based approaches. Contrary to probabilistic-based approaches, PMP is a deterministic approach that attempts to estimate the theoretical upper limit of precipitation depths that is physically possible for a given duration, location, and time of year. For New Mexico, Charles Thompson, Bureau Chief at New Mexico Office of the State Engineer, said in May 2018 “We use NOAA Atlas 14 for low hazard potential dam spillway analysis at the 1/100 AEP. We have not applied the 1/1000 AEP to significant hazard at this point (which is a FEMA recommendation that came out after our regulations were promulgated). Our regulations require a fraction of the PMP for significant hazard dams depending on size. In the future, once we have developed and accepted the risk based tool then we may see things differently.” 

CO rules table 2

Figure 1. Draft rules for selecting Intensity-Duration-Frequency requirements to assess spillway sizing requirements for low, significant, high, and extreme hazard dams.

MetStat provided hydrometeorological products in a frequency-based format that can support both standards-based and Risk-Informed Decision-Making (RIDM) for Dam Safety programs. The hydrometeorological information allows for engineers to conduct detailed stochastic flood analyses and develop hydrologic hazard curves for complex projects that warrant very detailed flood analyses related to the potential consequences downstream if a dam were to fail.

To provide the hydrometeorological information to dam-owners across the states of Colorado and New Mexico, we have developed a tool called MetPortal™ which is our web-based user-interface. In the future, versions of MetPortal™ will be used to deliver various meteorological or hydrometeorological information for specified projects. The CO-NM REPS MetPortal™ Precipitation Frequency tool is available at https://conm-reps-gui.shinyapps.io/metportal/. The CO-NM MetPortal™ Precipitation Frequency (PF) tool (Figure 2) delivers point and watershed PF estimates based on user input of a point or a watershed GIS shapefile.

MetPortal point page

Figure 2. Screenshot of the CO-NM REPS MetPortal™ tool on the Point PF Interface tab.

One of the key goals of the precipitation-frequency project was to provide increased accuracy in the estimation of extreme precipitation for use in hydrologic modeling of extreme floods, in particular annual exceedance probabilities (AEPs) well beyond 1 in 1000. Through MetPortal™, estimates and 90% uncertainty bounds are provided for AEPs from 1 in 10 (or 10-1) through 1 in 100,000,000 (10-7) (Figure 3). Providing AEPs to such rarity is necessary for dam safety and provides perspective to PMP.

To increase accuracy in the estimation of extreme precipitation for use in hydrologic modeling of extreme floods, PF estimates are available by storm type. The storm types include Local Storms (LS), Mesoscale Storms with Embedded Convection (MEC), Mid-Latitude Cyclones (MLC), and Tropical Storm Remnants (TSR). Each of these storm types has different spatial, temporal, and seasonal characteristics which should be maintained for realistic modeling of floods. Conducting separate PF analyses for each storm type reduced uncertainties caused by mixed-populations of storm types.

MetPortal PFcurve2

Figure 3. Example point PF results for the 2-hour Local Storms at a location in Colorado.

Both PF estimates at a point and over a watershed are available through the MetPortal™. Point PF estimates are used when a basin size is 10 square miles or less. Basins larger than 10 square miles require the use of watershed PF estimates that were developed specifically for basins in Colorado and New Mexico and specifically for each storm type. Each of these types of PF estimates are available on separate tabs of the MetPortal™ where a shapefile can be uploaded and watershed values automatically calculated (Figure 4).

MetPortal watershed page

Figure 4. Watershed PF estimates (AEP of 10-5) for the Mid-Latitude Cyclone (MLC) storm type after a watershed shapefile has been uploaded in the CO-NM REPS MetPortal™ on the Watershed PF tab.

Lastly, the MetPortal™ provides access to representative temporal (Figure 5) and spatial patterns as scalable historical and synthetic temporal patterns to use in hydrologic modeling simulations. Synthetic storm temporal patterns provide general information on storm characteristics such as: the time to peak of the maximum precipitation rate, the order of precipitation magnitudes surrounding the maximum precipitation rate, tercile of the storm duration with the maximum precipitation volume, and a mass curve/incremental precipitation from depth-area relationships from a suite of storms. The user can also select appropriate representative historical storms based on local geography, climatology, and dam safety considerations – high-intensity, short duration events for peak-runoff-sensitive locations or longer duration events for volumetric concerns, for example.

temporal plot

Figure 5. Example of a representative storm pattern for the MLC storm type in Colorado that is available via the MetPortal™.

MetPortal™ easily delivers all the pieces necessary to maintain compatibility between watershed PF relationships, spatial and temporal storm patterns, and storm seasonality. Uncertainty bounds for a selected AEP can be useful in accounting for precipitation-related uncertainties in hydrologic modeling of floods. The data are intended for use by consultants and dam safety staff for detailed flood analyses and allow the development of unbiased estimates of spillway performance. The data provide for a standards-based approach and support the development of hydrologic hazard curves for complex projects that warrant very detailed flood analyses for use in Risk-Informed Decision-Making (RIDM).

MetStat is actively working on future MetPortal™ platforms to serve other types of precipitation data, analytics, mapping, alerting, and monitoring features. Stay tuned for news about these new platforms!