In TL model, the power index ranges from 2/3 for h up to O(103) km, and 5/3 for h smaller than 1 km. Hanssen (1998, 2001) also estimated the index close to 1 from the differential tandem InSAR pairs. Emard-son et al. (2003) estimated that the value is close to 1 too. The power index in our estimation is 0.88, which is close to their studies. However, the nugget effect is neglected in the former two studies. Emardson et al.(2003) regarded the additional variance is caused by the height difference. In Tokyo area, the peak-to-peak height difference is less than 800 m, so we do not attribute the nugget effect to the height difference. In our model, the nugget effect is 35.2 mm2,which should be caused by the WZD observation variance, ranging 25–81 mm2 (Figs. 5 and 6).