The ring laser gyroscope (RLG) technique has been investigated for over 20 years as a potential complement to space geodetic techniques in measuring Earth rotation. However, RLGs are also sensitive to changes in their terrestrial orientation. Therefore in this paper, we review how the high-frequency band (i.e. signals shorter than 0.5 cycle per day) of the known phenomena causing site deformation contribute to the RLG observable, the Sagnac frequency. We study the impact of solid Earth tides, ocean tidal loading and non-tidal loading phenomena (atmospheric pressure loading and continental hydrosphere loading). Also, we evaluate the differences between available models of the phenomena and the importance of the Love numbers used in modeling the impact of solid Earth tides. Finally, we compare modeled variations in the instrument orientation with the ones observed with a tiltmeter. Our results prove that at the present accuracy of the RLG technique, solid Earth tides and ocean tidal loading effects have significant effect on RLG measurements, and continental hydrosphere loading can be actually neglected. Regarding the atmospheric loading model, its application might introduce some undesired signals. We also show that discrepancies arising from the use of different models can be neglected, and there is almost no impact arising from the use of different Love numbers. Finally, we discuss differences between data reduced with tiltmeter observations and these reduced with modeled signal, and potential causes of this discrepancies.