About CNRFC - Hydrometeorological Forecasting
| Topics: |
According to National Weather Service Instruction 10-911 - River Forecast Center Operations, "The Hydrometeorological Analysis and Support (HAS) forecasters perform the basic hydrometeorological analysis and support functions at the RFC. These basic HAS shift functionsinclude analysis and quality assurance of observed data (e.g., precipitation, temperatures), analysis of forecast data, such as quantitative precipitation forecasts (QPF), for use in thehydrologic modeling process; production of hydrometeorological discussions and other coordination products; and coordination activities with the Service Hydologist (SH), Hydro Focal Point (HFP) or forecasters in associated Weather Forecast Offices (WFOs), neighboring RFCs as applicable, Weather Prediction Center (WPC), and core partners within the water resources community."
While the majority of River Forecast Centers cross-utilize the Hydrometeorological Analysis and Support (HAS) duties with other hydrologic functions, the CNRFC dedicates forecasters to this particular role given the speed at which rivers respond (many within 24 hours) and the extremely diverse meteorological regimes resulting in highly variable spatial distribution of the weather over short distances across the CNRFC area of responsibility. As an example, the highest (Mount Whitney at elevation 14,505 feet above sea level) and lowest (Death Valley - Badwater Basin at 282 feet below sea level) locations in the "lower 48" of the United States are approximately 85 miles apart as the crow flies, and these two points show dramatic meteorological differences. The HAS forecasters at the CNRFC hold a degree in meteorology and are responsible for providing the meteorological inputs to the hydrologic model. These weather parameters, highlighted in the table below, are precipitation, freezing levels, and temperatures.
- Precipitation Forecasts
- 6-Hour time steps (ending 12, 18, 00, 06Z) for 6 Days
- Gridded forecast is converted to Mean Areal Precipitation for each sub-basin
- Used to calculate runoff and account for soil moisture
- Freezing Level Forecasts
- Instantaneous (12, 18, 00, 06Z) for 6 Days
- Gridded forecast is converted to Mean Areal FreeZing Level for each sub-basin
- Used to determine rain-snow elevation
- Each basin has a specified offset to determine snow level
- Temperature Forecasts
- Instantaneous (12, 18, 00, 06Z) for 10 Days, derived from maximum and minimum temperatures
- Gridded forecast is converted to Mean Areal Temperatures for each sub-basin
- Used to balance energy (melting snow)
Numerous tools are available to the HAS forecaster in order to make an accurate set of forecasts, and are represented in the general Operational HAS Function schematic below (Figure 1). There are several observing systems available for analysis, including but not limited to: satellite imagery, doppler radar (NEXRAD WSR-88D), upper air weather balloons (RAOB), vertically-pointed snow level radars, and several surface observation networks from other federal, state, and local agencies (e.g., ASOS, ALERT, GOES, GPS and SNOTEL). The data from these observing systems are used as input to initialize the numerical weather models (e.g., GFS, GEFS, NAM, SREF, HRRR, ECMWF, and GEM), which are a vital component in the forecast process. Finally, a national quantitative precipitation forecast (QPF) is provided by the Weather Prediction Center (WPC) utilizing these same observing systems and numerical weather models. This national QPF is used as guidance by the HAS forecaster in the overall process of generating the meteorological inputs to the hydrologic model.
To further understand how these meteorological inputs generated by the HAS forecaster are used in the hydrologic models by the Hydrologists, see the River Flood Forecasting section for more information.
Figure 1 - Operational Hydrometerological Analysis and Support
The CNRFC HAS forecasters utilize the nationally supported Advanced Weather Interactive Processing System (AWIPS) to view many of the observing systems, numerical weather models, and WPC forecasts described above in the forecast process to generate the meteorological inputs to the hydrologic model. The Display-2 Dimensions (D2D) perspective is part of the Common AWIPS Visualization Environment (CAVE), and includes user interfaces to view both text products and graphic displays, as well as disseminate text products to the world. An example showing the D2D perspective is below (Figure 2), with the interface displaying satellite imagery (water vapor and infrared), doppler radar, and a numeric weather model.
Figure 2 - CAVE / D2D Interface (click for larger image)
The Graphical Forecast Editor (GFE) is another perspective within the CAVE and allows the user to derive gridded forecasts for various meteorological parameters. The CNRFC HAS forecasters interact with the GFE to create the meteorological inputs to the hydrologic models. The GFE perspective contains two primary displays and a toolbar used for editing. The first display is the Spatial Editor, which depicts gridded data and allows the user to modify individual grid cells using a number of tools built-in to the GFE and local applications written by users. The second display is the Grid Manager, which provides an inventory of available grids, including: local forecasts, numerical weather model elements, and forecasts from other RFCs and WFOs surrounding and within the local office's area of responsibility. The graphic below (Figure 3) shows the GFE perspective with a precipitation forecast being displayed along with the local application used at the CNRFC to generate a precipitation forecast.
Figure 3 - CAVE / GFE Interface (click for larger image)
In general, the methodology of creating a precipitation forecast is a two-step process, points to grids and grids to basins. In order to account for impacts of complex terrain on the distribution of precipitation, the CNRFC utilizes an annual precipitation climatology grid as an underlying dataset. This precipitation climatology was generated at Oregon State University and is known as PRISM. To generate a gridded QPF from the point data, the process uses the PRISM normals and an inverse-distance squared weighting with the five nearest "neighbors" (QPF points) going into the calculation. The second step to obtain basin QPF from gridded QPF is to simply take the precipitation sum of the grid boxes within the basin, and then divide by the number of grid boxes. Figures 4a, 4b, and 4c below show the QPF in point, grid, and basin formats.
 Figure 4a - Point QPF | >>> |  Figure 4b - Gridded QPF | >>> |  Figure 4c - Basin QPF |
Freezing level forecasts heavily rely on the numeric weather models (e.g., GFS, ECMWF, GEM, and NAM). One technique commonly used is to drill down from the top level of the numeric weather model until the freezing level is found. This is known at the "top down" method. Other techniques use empirical formulas from parameters in the numeric weather models. Three of the more frequently used ones are listed below, utilizing temperature and height (calculated thickness) values.
- 700-mb Temperatures
- 850-mb to 700-mb Thickness
- 1000-mb to 500-mb Thickness
Another key dataset used in the near-term is the network of vertically-pointed snow-level radars scattered throughout California. These instruments were designed, built, and deployed by NOAA / Earth System Research Laboratory / Physical Sciences Division for a project involving the California Department of Water Resources to address water resource and flood control issues across the state. Some key locations where these vertically-pointed radars have been installed are Shasta, Oroville, Colfax, New Exchequer, Pine Flat, and Kernville. An example of the output available on the NOAA / Earth System Research Laboratory / Physical Sciences Division web site is depicted below (Figure 5).
Figure 5 - Vertically-Pointed Snow Level Radar Output
The technique to generate a temperature forecast starts with the expected maximum and minimum temperatures. These forecast high and low temperatures are then converted to instantaneous 6-hour temperatures (12, 18, 00, 06Z) using a fixed diurnal curve. The primary sources for these temperatures are the official maximum and minimum forecast temperatures issued by the 12 National Weather Service Forecast Offices located within the CNRFC area of responsibility (days 1 through 7) and numerical weather model guidance (days 8 through 10).
