Abstract
Using rotated principal component analysis (PCA), unique, orthogonal spatial patterns of daily and monthlyprecipitation on a well-instrumented, mountainous watershed in Idaho are examined for their relationship totopography, geographic location, and atmospheric variability. Precipitation pattern and homogeneous precipitationregion differences between daily and monthly timescales and between winter and summer Seasons were identifiedusing the rotated PCA procedure. In general, monthly data produced regional boundaries more closely alignedwith topography, reflecting the integration of many storm events on monthly timescales. Spatial fields, derivedfrom mapping rotated component loadings at 46 precipitation stations on a 234-kmz watershed, were found tobe highly correlated with topography and geographic location. The eight-year time series of the components forspecific watershed regions were found to be moderately related to linear combinations of meteorological variablesderived from a single radiosonde station approximately 50 km from the waterhed. This would indicate thepotential usefulness of data from a single location, such as a general circulation model grid point, to provideclues about spatial pattern changes and regional precipitation fluctuations even on a small watershed, if sufficientinformation about local climate (i.e., topographic influences) is first established.
