STATISTICAL RELATIONSHIPS BETWEEN TOPOGRAPHY AND PRECIPITATION IN A MOUNTAINOUS AREA

This study contributes to a goal of the Mackenzie GEWEX (Global Energy and Water Cycle Experiment) Study to model the water and energy balan ces of the Canadian Arctic Basin on spatial scales of 100 kilometres a nd time scales of one month. Accurate estimates of the spatial distrib ution of precipitation at these scales are critically important inputs into hydrologic models used to simulate these processes over complex terrain. The development of a suitable precipitation model for the Mac kenzie basin is seen as an important Step in our understanding of clim ate variability and climate change under conditions of increasing conc entrations of atmospheric greenhouse gases. In this study, univariate linear regression is employed to determine the relationships between m onthly precipitation and orographic parameters including elevation, as pect and slope in the mountainous terrain of Williston Basin located i n north central British Columbia. Slopes, aspects and elevations are c alculated using a 5-minute digital elevation model (DEM). The study fi nds several statistically significant (p < .05) correlations between e levation and monthly precipitation for terrain having either north or east aspect based on individual Julys between 1986 and 1990. Terrain h aving south or west aspect displays very low correlations for those sa me months. For winter conditions, the relationship between precipitati on and elevation is generally weak for all aspects for individual Janu arys between 1986 and 1990, The lack of correlation with wintertime pr ecipitation may very well be due to the poor quality data related to s now undercatch which is estimated to be roughly 50 percent. The qualit y and quantity of the data generally is also suspect owing to the volu me of missing records and the short period of record. Station elevatio ns and DEM elevations correlate equally well with monthly precipitatio n, contrary to the findings of several recent studies which show bette r correlations using coarser resolution. The study found no consistent relationship between monthly precipitation and slope in summer or win ter.