We analyze the temporal variation of the diurnal anisotropy of sub-TeV cosmic ray intensity observed with the Matsushiro ( Japan ) underground muon detector over two full solar activity cycles in 1985-2008 .
We find an anisotropy component in the solar diurnal anisotropy superimposed on the Compton-Getting anisotropy due to the earth ’ s orbital motion around the sun .
The phase of this additional anisotropy is almost constant at \sim 15:00 local solar time corresponding to the direction perpendicular to the average interplanetary magnetic field at the earth ’ s orbit , while the amplitude varies between a maximum ( 0.043 \pm 0.002 \% ) and minimum ( \sim 0.008 \pm 0.002 \% ) in a clear correlation with the solar activity .
We find a significant time lag between the temporal variations of the amplitude and the sunspot number and obtain the best correlation coefficient of +0.74 with the sunspot number delayed for 26 months .
We suggest that this anisotropy might be interpreted in terms of the energy change due to the solar-wind induced electric field expected for GCRs crossing the wavy neutral sheet .
The average amplitude of the sidereal diurnal variation over the entire period is 0.034 \pm 0.003 \% , which is roughly one third of the amplitude reported from AS and deep-underground muon experiments monitoring multi-TeV GCR intensity suggesting a significant attenuation of the anisotropy due to the solar modulation .
We find , on the other hand , only a weak positive correlation between the sidereal diurnal anisotropy and the solar activity cycle , in which the amplitude in the “ active ” solar activity epoch is about twice the amplitude in the “ quiet ” solar activity epoch .
This implies that only one fourth of the total attenuation varies in correlation with the solar activity cycle and/or the solar magnetic cycle .
We finally examine the temporal variation of the “ single-band valley depth ” ( SBVD ) quoted by the Milagro experiment and , by contrast with recent Milagro ’ s report , we find no steady increase in the Matsushiro observations in a 7-year period between 2000 and 2007 .
We suggest , therefore , that the steady increase of the SBVD reported by the Milagro experiment is not caused by the decreasing solar modulation in the declining phase of the 23rd solar activity cycle .