We model long run dynamics of precipitation at five transportation hubs in Atlantic Canada. In doing so, we designed methodology and applied it to twelve precipitation time series – six rainfall series and six snowfall series. Eleven out of twelve time series were estimated with OLS, and only Saint John rainfall series was estimated with GLM to deal with non-normal distribution of error terms. Based on our methodology and estimation, we detected that four precipitation series do not contain significant time trends over the latest period: Three of them are rainfall series with one snowfall series. Six out of remaining eight precipitation series do contain significant time trends that started in 1950s-1960s or 1980s. Moreover, we have similar trends of -3 mm per year in three snowfall series. All of them have started in 1950s-1960s. This result directly points towards potential climate change in the region and supports the conclusion of many climatologists of less snow in Atlantic Canada due to climate change. In general, our estimation showed that snowfall series are more deterministic, and they exhibit negative trends. Rainfall series are less predictable exhibiting either negative or positive trend.
Zircon - This is a contributing Drupal Theme Design by WeebPal.
