The MFDFA technique applied to two simulated time series. The time and amplitude values are in arbitrary units. The green data is antipersistent, akin to Poisson noise present in X-ray detectors, and is defined as fractional Brownian motion (fBm) with a Hurst exponent H = 0.1 (B. B. Mandelbrot & J. W. V. Ness 1968). The blue data is a persistent time series, which is more akin to thermal X-ray emission, defined as fBm with H = 0.8. The stars on the right-hand plot indicate the ground truth time series Hurst exponents. An intuitive way to interpret the right-hand plot is as a probability density of different fractal behavior across the time series. The left-hand time series are snapshots of infinite processes; in the infinite-time limit, the multifractal spectra on the right would converge to the stars. The gray line indicates the persistence/antipersistence boundary h = 0.5.