The precipitation interpolation map VS our new model:
We will use the average annual precipitation interpolation map to compare with our new model, since our
new model aims to describe a general precipitation pattern, and does not take into account any seasonal
effects. In addition, due to the lack of data from the valley system layer, the new model excluded the
northwestern tip of B.C., the southern end of Vancouver Island (including Victoria), and the southwestern
tip of B.C. mainland (south of Vancouver). The new model is generally very similar to the interpolation map.
They both have a dark blue band along mid to south coast, on Queen Charlotte Island, and on Vancouver Island,
showing the highest precipitation in those areas. The Okanagan Valley, the Interior Plateau, and the Great
Plains which experience the least amount of precipitation, as depicted on the interpolation map, are also
modelled as the driest areas in the new model. They are the lightest blue regions on both the interpolation
map and the new model. The Okanagan Valley, Interior Plateau receives minimal precipitation due to their
rain-shadow location behind the Coast and Cascade Mountains. Whereas the Great Plains in northeastern B.C.
lies in the rain-shadow region east of the Rockies. A relatively high precipitation is observed on both the
interpolation map and the new model in southeastern B.C., which corresponds to the Columbia Mountains and the
Southern Rockies. This region has relatively high precipitation due to orographic uplift of the westerly winds
travelling up the mountain slopes. On the interpolation map, it shows up as a bluish band east of the Interior
Plateau. And in the new model, the region shows up to the east of Interior Plateau in a distinct blue colour
outlining the Southern Rockies. Some differences exist between the interpolation map and the new model.
The northern central part and the northwestern tip of B.C. appear to have a higher precipitation in the new
model than in the interpolation map. This discrepancy could be due to the fact that we only had very few
precipitation data points in those areas, thus giving a pattern not representative of the area. The new model
describes the northern central B.C. to have moderate precipitation that increases from west to east, as can be
seen in the faint increase in intensity of the blue colour towards the east to the Northern Rockies. This trend
is well documented in climate literature (Schaefer, 2001), which further proves the validity of our new model.
Errors and Uncertainty:
- Limitation of the data: Due to the limitation of the data, we did not include all of the possible
variables in the precipitation function in the model. The variables examined in our project are more toward
the continental impacts on the precipitation; however, precipitation can also be influenced by many other
factors, such as meteorological factors (humidity, temperature, cyclones etc.).
- Selection of threshold criteria: Since we could not find any previous research or documents about weight
assignment of the variables according to their influence on precipitation, we have assigned relative weightings
to the four variables (aspect, distance from coast, elevation, valley system) in our new model based on our knowledge
gained from the physical geography courses. Therefore, the weightings may not truly reflect reality. But given
that the weight assignments were relative, the overall precipitation pattern generated is still reliable in a
- Distance from coast: The coastlines drawn in for B.C. mainland and Queen Charlotte Island were just
approximate. The buffer zones were generated from this approximate coastline, therefore the boundaries of
the buffer polygons are fuzzy. Uncertainty arises for the areas adjacent to the fuzzy boundaries since one
cannot determine which buffer polygon the areas should belong to. Moreover, the buffer distance of 200km was
picked so that it would generate a manageable number of classes in describing the continentality effect. Since
the buffer distance was selected without proper reasoning, it might not describe continentality correctly.
- Valley System using BEC Zones: One of the hardest problem in our selection was Valley system. It was
quite difficult to obtain accurate area considered to be valley system since the characteristic of each valley differ,
such as high/low elevation valleys, or small/large scale valleys. The use of BEC Zones assumes that the region is
a valley because of the species composition due to the climatic regime. The assumption can be fatally flaw if other
factors contribute to the growth of a particular species and is not precipitation dependant.