PENETRATIVE STRAIN ESTIMATION USING NORMALIZED FRY, RF-Φ AND SHAPE MATRIX EIGENVECTOR METHODS: A COMPARATIVE STUDY FROM EASTERN HIMALAYAN FOLD THRUST BELT

Penetrative strain generally decreases from internal to frontal thrust sheets in fold thrust belts (FTB). The normalized Fry and the R_f-φ are two of the most commonly used methods to estimate R (axial ratio) and φ (orientation) of the strain ellipse. In this study, we compare the accuracy of normalized Fry, R_f-φ, and shape matrix eigenvector (SME) methods as a function of lithology, texture, and structural position of the thrust sheets in the FTB. We conduct the study on seven major thrust sheets from two regional transects of Sikkim and far-eastern Arunachal Himalayan FTB. We analyzed 293 thin sections that were cut from 104 quartz-rich sandstone, quartzite, phyllite, schist, and gneiss. 80-120 quartz grains were digitized per thin section using the EllipseFit 3.8.0 software. The central vacancy in the normalized Fry plot was objectively fitted using the enhanced, continuous function method (CFM), and the point-count density (PCD) methods. In each thin section and for every method, we calculate 2σ (95%) uncertainties for R and φ using bootstrap error analysis to understand the goodness of the best-fit strain ellipse. In the internal thrust sheets, R >1.5 (R ~1.6-4.43), and R_f-φ method provides an accurate estimation of the penetrative strain. Results from the SME method also remain consistent with the R_f-φ method. As the grain-to-grain contacts between quartz were low in the studied gneiss, schist, and phyllite of the internal thrusts, the normalized Fry method did not provide a proper central void in most of the samples. The frontal thrust sheets dominantly comprising sandstone and quartzite, and record R <1.5 (R ~1.1-1.3). The R_f-φ plots show high fluctuation (>160°) in the φ values suggesting that initial grain ellipticities were higher than the penetrative strain. At such low strains, the R_f-φ plot strongly distorts the data distribution resulting in unreliable R and φ values. The 2σ uncertainties indicate that the SME and PCD methods provide the lowest errors in the low strain rocks as compared to the enhanced and CFM. SME method assumes initial grain shapes to be ellipsoidal with random orientations, whereas the normalized Fry method assumes initial spherical grains of different sizes. Thus, for low strained rocks, the SME method yields a better estimate of the penetrative strain in the studied rocks.