Spatially long wavelength noise which contaminates synthetic aperture radar (SAR) interferograms is still a poorly solved problem for precise measurement of ground surface displacements. In this report, three kinds of noise reduction approaches using displacement data measured by continuous Global Navigation Satellite System (GNSS) observation are proposed; a bilinear curve approach, an interpolation approach, and a low pass filter approach. Results of case studies show that long wavelength noises can be dramatically reduced by these approaches. However, in some cases, these approaches fail, i.e., insufficient noise reduction or undesirably vanishing true displacements. The efficacy of these noise reduction approaches depends on distribution of available GNSS points, ground surface displacement to be detected, and other conditions. An optimal approach should be selected depending on the situation.