We present a near global statistics on the correlation properties of daily temperature records. Data from terrestrial meteorological stations in the Global Daily Climatology Network are analyzed by means of detrended fluctuation analysis. Long-range temporal correlations extending up to several years are detected for each station. In order to reveal nonlinearity, we evaluated the magnitude of daily temperature changes (volatility) by the same method. The results clearly indicate the presence of nonlinearities in temperature time series, furthemore the geographic distribution of correlation exponents exhibits well defined clustering.
