Hydrologic Modeling and Support Services
MetStat’s Hydrologic Modeling Experience
MetStat has experience developing deterministic and probabilistic hydrologic hazard analyses (HHAs) for dams and other structures of interest. Historically, deterministic Probable Maximum Floods or Design Floods have been used in dam safety analysis. Probabilistic hydrologic loadings are used in risk-based methods of dam safety analysis. Multiple methods of developing hydrologic loadings are available including simple methods such as peak flood frequency analysis or very complex methods such as the Stochastic Event Flood Model (SEFM) developed by MGS Engineering Consultants. MetStat and MGS have partnered on many SEFM projects across the continental United States. SEFM is used to create a synthetic time series for a watershed, which is used to develop magnitude-frequency relationships for flow and reservoir water surface elevation.
A reservoir elevation frequency curve illustrating the probabilities of various hazard thresholds. Image credit: Mel Schaefer, MGS Engineering.
Stochastic Event Flood Model (SEFM)
The SEFM simulates thousands of years of flood annual maxima by utilizing a calibrated deterministic flood model and treats input parameters as stochastic variables. The variables are chosen from ranges of values specified by the user based on historical data to model realistic watershed behavior. The goal is to create synthetic simulations of reality, not to create a maximum or worst-case scenario condition.
The precipitation volume is distributed spatially and temporally using storm type templates developed from storm analyses of past events. Initial states for each simulation (e.g., antecedent precipitation, antecedent snowpack, soil moisture deficit) are resampled from long series of simulated values produced by continuous simulation. Additional details on the sampling procedure, which uses Total Probability Theorem, can be found at MGS.
The use of the stochastic approach allows for the development of separate hydrologic hazard curves for flood peak flow, maximum reservoir level, flood runoff volume and any other flood characteristic that can be obtained from the outputs of a watershed model. In addition, dam failure modes such as those related to flood capacity (e.g., reservoir elevation or flood runoff volume), spillway gate operability (e.g., modifications to gate operations), and other failure modes related to flood capacity (e.g., piping failures) can be represented by a combination of 1) chance elements and 2) functional relationships to critical flood characteristics, such as the depth and duration of overtopping.
Sacramento Soil Moisture Accounting Model (SACSMA)
The SACSMA model is the primary basis of the river forecast models applied at the National Weather Service offices across the country. MetStat has expertise in model development, calibration, verification, and development of inputs to this model. In 2015, MetStat assisted RTI International with the development of GIS-based tools for the California-Nevada River Forecast Center (CNRFC) for the generation of watershed boundaries, development of snow water equivalent climatologies, and creation of probability of snow cover and areal extent. We updated the IHABBS and ICAP programs used for the SACSMA model.
MetStat has also applied for funding through FEMA’s Hazard Mitigation Grant Program to support real-time extreme precipitation and flood alert capabilities for the State of South Carolina during heavy precipitation events, such as the atmospheric river associated with Hurricane Joaquin in 2015 and Hurricane Matthew in 2016. The project will involve the development of flood risk categories for small watersheds typically lacking operational hydrologic forecasts and will be implemented as a MetStormLive product.
Hydrologic Modeling Services
- Probabilistic HHA using peak flow frequency analysis
- Probabilistic HHA using SEFM
- Seasonal probabilistic HHA
- Design Hydrographs
- Reservoir routing
- Model calibration, verification, and forecasting
- Training on SEFM and other models (i.e., SACSMA)
Hydrologic Modeling Support Services
In addition to our direct hydrologic modeling services, MetStat is prepared to provide expertise and input data in support of any modeling system:
- High resolution precipitation data via MetStorm
- Historical time series and climatologies for model calibration and resampling techniques
- Observed hydrology for calibration and verification.
Applying SEFM to Forecasting on Rush Creek, CA
SEFM has also recently been used for the first time as a forecasting tool, with great success. Southern California Edison (SCE) approached MGS and MetStat in the spring of 2017 to develop real-time forecasts of inflow to the Rush Creek System of dams. The calibrated SEFM model was run with observed initial snowpack and forecasts of precipitation and temperature to develop 28-day flood forecasts. These practices remain experimental, as the imperfect initialization on Jun 20 demonstrates. However, the results predicted reservoir inflow trends with a high degree of accuracy, and we look forward to refining our techniques even further in future projects. At the conclusion of the forecasting effort, MetStat and MGS were asked to develop design hydrographs to be used in the design of flow control structures to mitigate future snowmelt runoff.