Abstract
The ever increasing number of traces in land 3D surveys has prompted important developments in the field of near surface velocity modeling. One development that stands out is replacement of semi‐interactively picked refraction arrivals with batch picked refraction arrivals (Tim Keho, personal communication). The objective of this study is to design and implement a robust methodology to quality control batch refraction arrival picks prior to their use in near surface inversion algorithms. Examples of new quality control tools presented here include: (1) decomposed relative arrival time versus offset and azimuth with optional overlay on shot records, (2) slowness maps which can be converted to velocity grids for potential input to near surface modeling packages and for comparison with hard data such as uphole velocity surveys, and (3) maps indicating the location of anomalous jumps in first arrival time picks. These anomalies, referred to as shingling, are usually associated with velocity inversions. Knowing where near surface velocity inversions occur is critical to identifying challenging data zones that may require additional near surface analysis or noise suppression processing. A key requirement of our approach is the ability to execute in a batch mode on large data volumes within a reasonable amount of wall clock time.