Study results based on actual sensor data comparisons in Carberry for hundreds of hourly observations for the month of February and May, 2011 and July, 2010 clearly illustrate that there is little statistical difference in the measured variables between the Environment Canada station (control) versus the WeatherFarm station (measured) for temperature, dew point, precipitation and, in most cases, relative humidity, as well as wind speed (February 2011). Additionally, precipitation measurements were also compared between all three stations – WeatherFarm, MAFRI and Environment Canada – and were found to be significantly correlated in all cases.
Precipitation measurements for the liquid (rainy) season were also compared between the three station rain gauges, and little statistical difference was found in tipping buckets in relationship to weighing gauge measurements. Further an extreme rainfall event in Saskatchewan is analyzed providing proof that tipping bucket gauges do not underestimate heavy rainfall events. The WeatherFarm network provided highly localized weather data during the storm to capture convection precipitation. This study demonstrates that real-time weather information from the WeatherFarm station sensors is complementary to the Environment Canada stations data, and provides weather information with the granularity necessary for real-time weather event management and decision making relative to flood prediction, monitoring and forecasting. This data has also been used to increase the timeliness and accuracy of watches and warnings and special weather statements issued by Environment Canada.