During the summer months the City of Winnipeg experiences severe rainstorm events thatexceed the capacity of the land and sewer drainage systems. There exists a need for a real timemonitoring system to provide prompt and accurate rainfall measurements. This data is critical to informcivic officials of probable areas of flooding and to aid in minimizing Combined Sewer Overflows (CSOs).Currently, the City employs a network of 24 rain gauges for collection and analysis of rainfall over an areaof approximately 700 km2. This density of gauges is required since critical storms tend to be convectiveand thus spatially heterogeneous. The goal of this study is to investigate methods that can supplementthe current rain gauge network using Doppler radar products from Environment Canada.As a preliminary step towards the project goal, an inter-comparison of aerial rainfall from radar and raingauges is being evaluated. This study will concentrate on employing the refurbished Doppler radar forquantitative precipitation estimates. Measurement of rainfall using radar reflectivity is based on theconversion of radar reflectivity (Z) to rainfall rate (R) using the well known Z-R power law relation.Currently four summer seasons of 10 minute data are available to develop a Z-R relationship specificallyfor the City of Winnipeg.Data from 75 storm events during 2000-2003 will be presented. Raw radar reflectivity was collected fromtwo data products: i) the CAPPI (constant-altitude planned position indicator) scan, which is a horizontalradar projection from an altitude of 1.5 kilometers and ii) the lowest Doppler scan 0.3 degree above thepoint of tangency. The radar data for the project was made available from the Environment CanadaWoodlands Radar located north-west of Winnipeg. This radar has a spatial resolution of approximately 1km X 1 km over the City. From the entire 300 km radar scan, a data window of 51 radials by 46 points perradial was extracted. Freely available software known as the Gridded Analysis and Display System(GrADS) was used for all data handling and analysis.Initial analysis of the data set has revealed that the Z-R power-law coefficients currently used byEnvironment Canada tends to over predict rainfall intensity for large summer storm events. Varioustemporal and spatial averaging techniques will be utilized of overcome problems associated with radarsampling area (～1 km2) and rain gauge area (200 mm diameter), and both evaporation and travel beneaththe radar beam. This averaging will include interpolation of gauge data using Cressman objectiveanalysis.
展开▼
