Interpretation of fault system in the Tana Sag, Kenya, using edge recognition techniques and Euler deconvolution

The recognition of fault structures in gravity anomaly data is the most important step in the interpretation of geological and geophysical data. Edge detection and edge enhancement are commonly used for researching the edge locations of geological bodies. The edge recognition techniques applied in this paper are the tilt angle (TA) and its horizontal derivative (TA-THDR), as well as the normalized vertical derivative of the total horizontal derivative (NVDR-THDR), in which higher order derivative was involved. The results of TA and its total horizontal derivative provide detailed information to reflect the deeper sources boundaries more accurately and effectively. The information outside the geological body edges obtained from the NVDR-THDR results is eliminated by using a threshold value greater than 0, which improves the horizontal resolution to detect the fault structures with smaller scale or deeper depth. However, tilt angle and its horizontal derivative are found to be sensitive to noise, as the higher order derivative is. The noise effect is reduced by using upward continuation during the processing of the measure gravity data.