Variogram Time Series Analysis Applied to the Spatial Structure of Snow Accumulation

Abstract The correlation of earth system properties is important for assessing monitoring strategies, determining scales of modeling, and improving forecasting capabilities. We present a new method to examine the spatial scale of inter‐annual patterns from time series data. The variability in annual patterns between stations is computed using daily data from a network of stations. This variability is used to compute the semi‐variance for intervals of distance and plotted in the form of a variogram. Variograms are used to identify the correlation distances for a specific process. Here, the method is applied to 90 stations of daily snow water equivalent accumulation and precipitation data over the Southern Rocky Mountains of the Western USA for a 40‐year period (1981–2020). At 5‐, 10‐, or 20‐km lag distances, snow accumulation patterns are very similar to 90 or 100 km. Snow accumulation patterns are less correlated up to about 380 km; beyond there is no quantifiable spatial correlation. Summer precipitation patterns are correlated up to about 60 km while winter precipitation patterns are spatially consistent for 100 km and likely to more than 300 km. Subsets of the accumulation and precipitation data to explore differences due to geographic location, land cover type, and the Oceanic Niño Index yielded similar results.