Precipitation Data

Precipitation data for the GVRD was obtained from Environment Canada. The following sections describe the acquisition of precipitation data, the process of sorting and selecting the data, and the exploratory data analysis undertaken in creating the precipitation model.

The precipitation data was acquired from Environment Canada’s web site. Historical data up to and including December 1999 was available for download in the form of a CD image. The 2001 CDCD West CD zip file was downloaded and burned onto a CD-ROM in order to access the climatic data. The climate data CD lacks data from 2000 to 2003. A request to the data steward was made in hope of obtaining the data for this time period. As a result of no response from the data steward, the data was manually obtained from Environment Canada’s web site utilizing a cut-and-paste method. The final compiled data contains a single table of 43 stations along with the monthly average precipitation in mm based on data available from 1980 to 2003.

From the climate data CD-ROM, the precipitation data for all weather stations within the following latitudes and longitudes was obtained: 49° 00' N to 49° 30' N latitude and from 121° 30' W to 124° 00' W longitude. This produced 81 weather stations within the GVRD. By sorting through these stations it was found that the length of record varied for each station. Start dates, end dates, and gaps in the series all varied. It was decided that useable stations would be selected based on two characteristics: containing 15 years of data and the 15 years falling within the period of 1980 – 2003. This produced 34 stations with the appropriate data and 9 stations that fell just outside of the qualifications. It was decided that these 9 stations would be included because they were often located in areas that lacked data points.

Simple descriptive statistics on the data set were investigated initially to examine any apparent trend. It has been found that there is a great variation in the daily mean value for a given station based on historical data. For example, mean precipitation for January 1st at Simon Fraser University on Burnaby Mountain is 9.79 mm based on data from 1984 to 1999. However, the standard error is 4.16 mm. This indicates that a confidence interval of 95% for the mean value lies between 18.11 mm to 1.47 mm. The implication of this will be discussed later in the report.

Linear regression analysis on the elevation and monthly mean precipitation data was performed by SPSS, a statistical software package that investigates any apparent relationships. The results show a very weak linear relationship between elevation of station sites and the precipitation amount. Table 4 gives an overview on the R-squared values for all of the mean monthly values as well as mean annual precipitation values. figure 5 shows the scatter plot between elevation and mean annual precipitation for each of the 42 sites as an example of how the dataset looks.