What is Seattle RainWatch?
Seattle RainWatch is a real-time weather system that
provides rain accumulation totals for the past 1- to 48-hours and
forecasts rain accumulation for the next hour for the Seattle metropolitan region. It uses rainfall estimates derived
from radar data that are calibrated with local rain gauge networks
to improve accuracy over other radar only indicated precipitation
estimate products. The forecasts are made using radar echo motion
vectors over the past hour and are extrapolated outward
temporally and spatially.
How does it work?
Radar reflectivities (dBZ) from the level II data-stream provided by the KATX (Camino Island) WSR-88D Doppler Radar are ingested in real-time by the Seattle RainWatch system. The reflectivities are converted into rainfall rates using a standard dBZ to rainfall rate conversion and are then calibrated by real-time data provided by on-the-ground local rain gauge networks. These calibrated rainfall estimates are then integrated into rainfall accumulation totals for 1-,6-,12-,24- and 48-hours
Who is involved with Seattle RainWatch?
The Seattle RainWatch project is lead by Prof. Cliff Mass. The real-time system was developed by Phil Regulski and David Carey provides web design and graphics. It is a product of the Mesoscale Analysis and Forecasting Group at the University of Washington's Department of Atmospheric Sciences. Funding and one of rain gauge networks is supplied by Seattle Public Utilities.
Please send us an e-mail if you would like to get involved or expand coverage to your region. email@example.com
What benefits does Seattle RainWatch provide users?
Traditional rainfall estimates calculated by radar reflectivity are prone to a number of problems that can create incorrect rainfall accumulation estimates (See: What are common problems with the radar reflectivity to rainfall rate conversion?). Without calibration from on-the-ground rain gauges these errors can propagate and intensify as successive rainfall estimates are combined to calculate accumulations over longer time-scales. Seattle RainWatch removes these radar biases by calibrating the radar-indicated rainfall rates to rain gauge networks scattered throughout the region. Therefore Seattle RainWatch provides a more accurate picture of the current hydrological state of the region.
What are common problems with the radar reflectivity to rainfall rate conversion?
There are many different reasons why the radar reflectivity to rainfall rate conversion can produce errors. WSR-88D Radar Rainfall Estimation: Capabilities, Limitations and Potential Improvements by S.M. Hunter provides a good introduction to this subject. Below is a summary of that paper.
Known problems with radar reflectivity:
Z-R relationship issues:
Many studies have shown there are variations in the particle size distribution which the Z-R relationship is based. In order to account for different distributions people have developed different Z-R relationships for different types of rain (stratiform, convective), types of precipitation (drizzle, rain, hail, snow) (Joss et al. 1970), geographic locations (maritime, continental) (Carins et al. 1998; Quinlan and Sinsabaugh 1999), etc. . The most commonly used is Z = 300R^1.4 but errors will be introduced based on the difference between the true particle size distribution and the assumed.
- Radar calibration: Biases in the measurement of reflectivity can arise from incorrect hardware calibration at the radar site.
- Attenuation: Small (~1dB) errors due to wet radome.
- Frozen hydrometeors and the melting layer: The weather radar equation assumes scatterers are spherical in shape, liquid and evenly distributed drops distributed evenly throughout the volume scan region. There will be a difference between the true distribution and the assumed introducing errors.
- Anomalous propagation: Refractivity lapse rates, created by large gradients in temperature and/or water vapor, produce superrefraction and subrefraction of the radar beam and inaccurate calculations of the beam height. This will alter return reflectivities from the true state of the atmosphere.
- Beam blockage: Terrain blocks radar pulses and result in areas of no coverage.
- Range effects: Earth curvature and resulting beam overshoot and beam spreading can all alter the radar reflectivities.
Each of the following products have three different levels of zoom capability overlayed with terrain features:
- Full domain: The entire region covered by the KATX (Camano Island) radar
- Greater metro: Centered around the greater Seattle metropolitan area
- Local metro: Centered around Seattle City limits
||The current radar
reflectivity. NOTE: When radar is in Clear-Air Mode the reflectivities are filtered to reduce clutter.
||The current rainfall rate in in/hr
calculated by the Z-R relationship (radar reflectivity to rainfall rate relationship). This product is NOT
calibrated with the rain gauge networks.
|1 hr precip
||The accumulated precipitation (inches) in the last hour calibrated with the rain gauge networks
|6 hr precip
||The accumulated precipitation (inches) in the last 6-hours calibrated with the rain gauge networks.
|12 hr precip
||The accumulated precipitation (inches) in the last 12-hours calibrated with the rain gauge networks.
|24 hr precip
||The accumulated precipitation (inches) in the last 24-hours calibrated with the rain gauge networks.
|48 hr precip
||The accumulated precipitation (inches) in the last 48-hours calibrated with the rain gauge networks.
|1 hr precip FCST
||A 1-hour forecast of
precipitation accumulation (inches) using the current
radar motion vectors calibrated to the rain gauge networks.
||Difference between the radar-calibrated accumulation versus the rain gauge observations.
|Calculated versus obs
||Plot of the observations and radar-calibrated rainfall accumulations.
Carins, M.A., Huggins, and S. Vasiloff, 1998: Precipitation algorithm improvements in the Eastern Sierra. NWS Western Region Technical Attachment, No. 98-08, Salt Lake City, UT, 5 pp.
Federal Meteorological Handbook No. 11: Doppler Radar Meteorological Observations (WSR-88D). Parts A-D.
Hunter, S.M. WSR-88D Radar Rainfall Estimation: Capabilities, Limitations and Potential Improvements.
Joss, J.K., K. Schram, J.D. Thams, and A. Waldvogel, 1970: /On the quantitative determination of precipatation by radar./ Wissenschaffliche Mineilungen Nr. 63, Eidgenossisichen Komission Zum Studium der Hagelbildung und der Hagelawehr, 38 pp.
Marshall, J. S., and W. M. Palmer: The distribution of raindrops with size. /J. Appl. Meteor./, 5, 165-166.
Quinlan, J.S. and E.J. Sinsabaugh, 1999: An evaluation of the performance of snow algorithm at NWFO Albany, NY during the 1997-98 winter season. Preprints, 29th Int. Conf. on Radar Meteorology., Montreal, Quebec, CA, Amer. Meteor. Soc., 794-979.
Wilson, J.W., 1979: Radar Measurement of Rainfall - A Summary. Bulletin American Meteorological Society, Vol 60, No. 9, 108-1058.
WSR-88D Operations Course Handbook. Operations Training Branch Operational Support Facility. Norman, Oklahoma.
© 2009,2010,2011,2012,2013,2014,2015,2016,2017,2018 University of Washington Dept. of Atmospherics Sciences