DETECTING THE SOLAR CYCLE FROM GROUND-BASED MEASUREMENTS

Detection of trends and long-term patterns such as the solar cycle using the ground-based measurements is problematic due to the high degree of variability on daily, seasonal and annual time scales. We have approached this problem using thermal diffusion of the temperature time series in a conductive heat flow model. Thermal diffusion filters the input signals so that the amplitude diminishes exponentially with depth as a function of the period of variation and the thermal diffusivity of the medium. The result is that the detection limit of a signal varies according the relationship between signal period and depth (10 m). We have applied this filter to daily mean values for air, soil and downward radiation during the 1981-2003 period from a network of automated meteorological stations within North Dakota, South Dakota, Nebraska, and Kansas that are maintained by the HPRCC. Trends of the air and soil temperatures, and downward solar radiation at the surface during the study period correlate well with each other. More important, these trends are nearly the same as the variation of phase-shifted solar irradiance at the top of atmosphere measured by satellites. Preliminary results from this study have shown that the variations of the surface air and soil temperatures within a solar cycle range from 0.25 to 0.4 °C at the Northern Great Plains. The solar irradiance at TOA during the past two solar cycles varied about 0.1% (~1.4 Wm^-2), producing a climate forcing of 0.24 Wm^-2 and resulting in a surface temperature change of ~0.18 °C (Natural variability). Thus, the anthropogenic variation during a solar cycle will approximately be within a range of 0.07 to 0.22 °C, and a more detailed study is warranted.