Interpolation of 1961-97 daily temperature and precipitation data onto Alberta polygons of ecodistrict and soil landscapes of Canada

Soil quality models developed for ecodistrict polygons (EDP) and the polygo ns of the soil landscapes of Canada (SLC) to monitor the concentration of s oil organic matter require daily climate data as an important input. The ob jectives of this paper are (i) to provide a method that interpolates the da ily station data onto the 894 SLC polygons and 150 EDP in the province of A lberta, Canada, so that the interpolated data fit not only climate mean but also climate variability, especially for the precipitation field, and henc e can be used as realistic climate input to soil quality models and (ii) to understand the variability of the Alberta daily climate, such as precipita tion frequency. The procedure interpolates the station data onto a dense ne twork of grid points and then averages the gridpoint values inside polygons . The procedure and results for maximum temperature, minimum temperature, a nd precipitation are reported in detail. The interpolation uses the observe d daily data for the period 1 January 1961-31 December 1997 (13 514 days) w ithin the latitude-longitude box (45 degrees -64 degreesN, 116 degrees -124 degreesW). Because the precipitation field can have a short spatial correl ation length scale and large variability, a hybrid of the methods of invers e-distance weight and nearest-station assignment is developed for interpola ting the precipitation data. This method can reliably calculate not only th e number of precipitation days per month, but also the precipitation amount for a day. The temperature field has a long spatial correlation scale, and its data are interpolated by the inverse-distance-weight method. Cross-val idation shows that the interpolated results on polygons are accurate and ap propriate for soil quality models. The computing algorithm uses all the dai ly observed climate data; despite that, some stations have a very short tim e record or only summer records.