Subseasonal climate variability for North Carolina, United States

Subseasonal trends in climate variability for maximum temperature (Tmax), minimum temperature (Tmin) and precipitation were evaluated for 249 ground-based stations in North Carolina for 1950–2009. The magnitude and significance of the trends at all stations were determined using the non-parametric Theil–Sen Approach (TSA) and the Mann–Kendall (MK) test, respectively. The Sequential Mann–Kendall (SQMK) test was also applied to find the initiation of abrupt trend changes. The lag-1 serial correlation and double mass curve were employed to address the data independency and homogeneity. Using the MK trend test, statistically significant (confidence level ≥ 95% in two-tailed test) decreasing (increasing) trends by 44% (45%) of stations were found in May (June). In general, trends were decreased in Tmax and increased in Tmin data series in subseasonal scale. Using the TSA method, the magnitude of lowest (highest) decreasing (increasing) trend in Tmax is − 0.050 °C/year (+ 0.052 °C/year) in the monthly series for May (March) and for Tmin is − 0.055 °C/year (+ 0.075 °C/year) in February (December). For the precipitation time series using the TSA method, it was found that the highest (lowest) magnitude of 1.00 mm/year (− 1.20 mm/year) is in September (February). The overall trends in precipitation data series were not significant at the 95% confidence level except that 17% of stations were found to have significant (confidence level ≥ 95% in two-tailed test) decreasing trends in February. The statistically significant trend test results were used to develop a spatial distribution of trends: May for Tmax, June for Tmin, and February for precipitation. A correlative analysis of significant temperature and precipitation trend results was examined with respect to large scale circulation modes (North Atlantic Oscillation (NAO) and Southern Oscillation Index (SOI)). A negative NAO index (positive-El Niño Southern Oscillation (ENSO) index) was found to be associated with the decreasing precipitation in February during 1960–1980 (2000–2009). The incremental trend in Tmin in the inter-seasonal (April–October) time scale can be associated with the positive NAO index during 1970–2000.