Example synthetic data (top two rows) generated from the measured bispectrum and example LCT data (bottom two rows). The different columns are the (synthetic or real) data (left), contours of the estimated quadratic nonlinear component overplotted over the data (right), the residual after subtracting the estimated nonlinear component from the data (middle column, two bottom rows), and the Gaussian field from which the synthetic data were generated (middle column, top two rows). Contour lines in the right column indicate levels of ±1σ_XNL (dashed) and ±2σ_XNL (solid) of the nonlinear component (positive: red, negative: blue). In addition, the skewness is given in the corresponding histogram; both were computed from the entire dataset. See Fig. F.5 for the corresponding TD results.